JP4114376B2 - Coordinate determination method for colorimetry - Google Patents

Description

尚、座標ｐiは離散的な座標であるから座標間の距離を積算しても厳密には曲線の長さにはならないが、階調数が２５６であって多数存在するため、ほぼ曲線の長さとみなすことができる。以下、近似値ｌを曲線の長さと同視して説明する。また、式（１）において右辺は各座標間の距離を示している。
【００８３】
隣り合う座標間の距離の最大値Ｌmaxは、曲線ＷＣ上に各座標間距離が均等になるように配置した場合の当該距離の定数倍（本実施形態においては２．５）であり、以下の式（２）にて算出することができる。
【数２】

尚、式（２）においてｎは測色用座標数であり本実施形態においてはｎ＝１０である。
【００８４】
このようにしてＬmaxを制限することにより、曲線ＷＣ上に座標を均等配置した場合を基準として、その定数倍より遠くに隣り合う座標が離れないようにすることができる。従って、曲線ＷＣ上の一部において座標が密になることを防止することができる。ここで、式（２）における２．５は印刷メディア等の条件や実験結果によって好ましい値を算出し、適宜変更することができるが、概ね２．０〜３．０の範囲内が好ましい。Ｌmaxは、Ｌａｂ色空間中の曲線の長さに依存しているので、ＣＭＹ色空間の各稜線毎に異なった値となり、それぞれ計算する。
【００８５】
一方、本実施形態において隣り合う座標間の距離の最小値Ｌminは、上記各稜線について定数値０．７を採用している。すなわち、Ｌａｂ色空間は均等色空間でありその距離は色差であるから、座標間の最小距離を規定することによって当該座標の色差の最小値を規定することができる。人間の目による色の認識限度は概ね１．０程度であるから、Ｌminをこの程度の値にすることによって座標が近づきすぎることを防止するとともに、少なくとも人間の目に有意な色の変化が感じられるような測色用座標間の距離を確保することができる。むろん、視覚による認識限度は相対的なものであり、上記定数値０．７はその値の一例であるが、概ね０．５〜１．０程度が好ましい。
【００８６】
ＣＭＹ色空間中の稜線に対応したＬａｂ色空間中の曲線上のある座標に対して、どのような座標位置であれば隣り合わせることが可能であるかを判別するには、上記式（１）（２）を計算することによって判別することができるが、具体的な座標位置は各座標について他の座標との距離を実際に計算してみる必要がある。本発明における試行時には距離制限を加味しているが、試行工程はコンピュータ演算によって自動で繰り返すことが好ましいので、試行工程において距離制限の条件判別を逐一実行するのは計算速度上好ましくない。
【００８７】
そこで、本実施形態においては、予め距離制限をフラグ化した隣接行列を用いて試行工程の計算速度向上を図っている。図１２は当該隣接行列の説明図である。隣接行列の要素はａijであり、ｉ，ｊはそれぞれ０〜２５５の整数値をとるとともに要素ａijの実際の値は０もしくは１である。ａijの値は以下の条件式（３）によって規定されている。
【数３】

【００８８】
すなわち、Ｌａｂ色空間中の座標ｐiに関し、その隣の座標として座標ｐjを想定したときに上記距離制限を満たす座標ｐjが存在しうる場合に要素ａijが”１”になる。
【００８９】
このような隣接行列を規定しておくと、ＣＭＹ色空間中で仮決定される測色用座標列Ｘを端から順番にｘk（ｋは０〜９，ｘkは０〜２５５であってＣＭＹいずれかの座標値）としたとき、ｋ番目の座標については隣接行列の要素が以下の条件式（４）を満たすときに上記距離制限内であると判定することができる。
【数４】

すなわち、隣接行列においては注目している座標ｐiより番号の大きな座標ｐj（式（３）による条件ｉ＜ｊ）を比較対象としているので、座標の一方側で隣り合う座標についての条件をフラグ化しているが、番号ｋ−１と番号ｋ＋１を考えることにより番号ｋの座標について両隣の距離制限を加味することができる。
【００９０】
図１３は、この隣接行列および上記条件式（４）によって距離制限の判別が実施される様子を説明する説明図である。同図においては簡略のため隣接行列の一部を示しており、その要素値も説明のための例として図のように決めてある。試行工程においては、測色用座標列Ｘのうちの一つが移動され、逐次上記ずれが算出される。同図に示す例では１番目の座標を移動させることとしてｘ_１’として示している。また、ｘ_０の値は０，ｘ_２の値は８であるとして説明する。
【００９１】
ｘ_１’の距離制限を上記条件式（４）で考える場合、上述のように当該ｘ_１’のみの移動を考えるため、ｘk-1とｘk+1とは固定であってこの例ではそれぞれ”０”と”８”である。そこで、まずａ_0xk＝１を考える。このとき、図１３に示す隣接行列の１行目が”１”であるか否かを考えればよいので、ａ₀₂，ａ₀₃，ａ₀₄，ａ₀₅，ａ₀₆が該当することが分かる。一方、ａ_xk8＝１を考える際には、図１３に示す隣接行列の８列目が”１”であるか否かを考えればよいので、ａ₁₈，ａ₂₈，ａ₃₈が該当することが分かる。
【００９２】
従って、条件式（４）を満たすｘkは２と３であり、ｘ_１’が取りうる座標値は２もしくは３である。そこで、ｘ_１については２と３を代入して上記試行用対応関係を規定し、上記ずれを算出する。このような隣接行列による条件判別と試行用対応関係の規定、ずれの算出を測色用座標列Ｘの総ての番号ｋについて実施すれば、距離制限内で座標を移動させた測色用座標の総ての組みついてずれを算出することができ、従って、ずれの極小値を把握することができる。この作業をＣＭＹ色空間中の各稜線について実施することにより、各稜線上で測色用座標を特定することができる。
【００９３】
（４）第１色空間内での測色用座標の拡張：
以上のようにして各稜線上で測色用座標を特定した後には、ＣＭＹ色空間中の全域に渡って測色用座標を拡張する必要がある。図１４は、稜線上の測色用座標からＣＭＹ色空間中の測色用座標を決定する様子を説明する説明図である。同図は、ＣＭＹ色空間中に形成される立方体を示すとともに、その稜線上で特定された測色用座標を○で示している。ＣＭＹ色空間の外周面においては、ＣＭＹ色空間中で平行な稜線上の測色用座標を直線で結んだときの交点を測色用座標位置とする。
【００９４】
例えば、測色用座標のそれぞれにおいて座標値が小さい方から番号を付していき、同じ番号の測色用座標同士を直線で結ぶと交点を決定することができる。図１４では、稜線ＷＣ上の測色用座標および稜線ＹＧ上の測色用座標に対して図１４に示すように０〜９番の番号を付し、４番同士を直線で結ぶ。稜線ＷＹ上および稜線ＣＧ上でも同様の作業により測色用座標を直線で結ぶと、これらが交差するので、当該交点を測色用座標ｘnとする。総ての稜線について、総ての番号の測色用座標について同様の作業をすれば、ＣＭＹ色空間の外周面上で測色用座標を決定することができる。
【００９５】
さらに、ＣＭＹ色空間内部の測色用座標については、面の交点を考える。すなわち、上述のようにしてＣＭＹ色空間の外周面上に直線を引いていくと当該直線に囲まれた面が形成される。図１４に示す例では、ある番号の測色用座標同士を結んだ直線に囲まれる面Ｓ１，Ｓ２，Ｓ３が形成される。各稜線上で測色用座標の位置がどのように変化したとしても、面Ｓ１は面ＷＭＲＹと面ＣＢＫＧに挟まれ、面Ｓ２は面ＷＣＢＭと面ＹＧＫＲに挟まれ、面Ｓ３は面ＷＣＧＹと面ＭＢＫＲに挟まれているので、これらの面Ｓ１，Ｓ２，Ｓ３はＣＭＹ色空間内で交わるとともに共通の交点は１点となる。そこで、この交点をＣＭＹ色空間内の測色用座標ｘｍとする。
【００９６】
総ての稜線について、総ての番号の測色用座標について同様の作業をすれば、ＣＭＹ色空間の内部で測色用座標を決定することができる。この結果、稜線および外周を含むＣＭＹ色空間中に１０^３個の測色用座標を決定することができる。これら１０^３個の測色用座標にて印刷したパッチを測色すれば、１０^３個の座標についてＣＭＹ値とＬａｂ値とを正確に対応づけることができる。これらの座標に基づいて任意のＣＭＹ値についてスプライン補間演算を行えば、任意のＣＭＹ値に対するＬａｂ値（補間値）を得ることができるが、上述のようにこの補間値と真の値（例えば実際に測色した値）との色差が２．０以上になる座標は全体の１０％以下である。従って、この測色値によってＣＭＹ色空間とＬａｂ色空間とを正確に対応づけることができ、この対応関係を参照しながら正確なＬＵＴを作成することができる。
【００９７】
また、普通紙について染料インクを使用して出願人が行った実験によると、当該色差の平均値は１．０程度になり、多数のＣＭＹ値に対して上記色差のヒストグラム（縦軸は分布数、横軸は色差）をとるとその標準偏差が０．６程度になる。普通紙について染料インクを使用しつつ測色用座標をＣＭＹ色空間中に均等に配置して検証すると、上記色差が２．０以上になる座標が全体の２０％になり、色差の平均値が１．５程度になり、標準偏差が１．４程度になる。従って、普通紙については特に、本発明にかかる測色用座標の決定により非常に顕著な精度向上効果を奏すると言える。
【００９８】
（５）測色用座標の決定作業：
以下、上記各工程を一連の作業として実施してＣＭＹ色空間中の稜線上の測色用座標を決定する作業をフローチャートに沿って説明する。図１５は、測色用座標決定作業の全体を示すゼネラルフローである。同図において、ステップＳ１００では上記コンピュータ１０にてプリンタ１２を制御し、上記図５に示すようなＣＭＹ色空間中の各稜線上に均等に配置された１６個の基準座標に該当するＣＭＹ値で印刷を実行する。この結果、印刷用紙上には図４に示すような１９２個のパッチＰが印刷される。
【００９９】
ステップＳ１０５では、当該印刷された各パッチＰを測色器２０で測色し、それぞれのＬａｂ色空間中の座標値を取得する。すなわち、ステップＳ１００，Ｓ１０５が上記基準座標取得工程に該当する。ステップＳ１１０では、当該測色値および上記ＣＭＹ色空間中の座標値を参照してスプライン補間関数を決定する。当該決定されるスプライン補間関数が上記基準対応関係である。ステップＳ１１５では、試行工程において測色用座標同士に距離制限を与えるため、ＣＭＹ色空間中の稜線のいずれかについて距離制限を決定する処理を行う。
【０１００】
ステップＳ１２０では、処理中の稜線上について上記図５に示す様に１０個の測色用座標を均等に配置して初期座標とし、初期測色用座標列Ｘとする。ステップＳ１２５では、番号ｋを”１”に初期化し、ステップＳ１３０にてずれ算出処理を行う。ずれ算出処理では、後述するようにｋ番目の測色用座標を上記距離制限内で移動させる測色用座標の仮決定を行い、基準対応関係により算出されるＬａｂ座標と試行用対応関係により算出されるＬａｂ座標とのずれを逐次算出する。ステップＳ１３５では、番号ｋがｎ−２（ｎは１稜線上の測色用座標数。本実施形態ではｎ＝１０）に達しているか否かを判別しｋがｎ−２に達していると判別されなければステップＳ１３７でｋをインクリメントしてステップＳ１３０以降の処理を繰り返す。
【０１０１】
すなわち、ずれ算出処理では番号ｋについてその可動範囲内で座標を移動させ、逐次ずれを算出するとともに番号１〜８（１０−２）の座標について同様の処理を繰り返すことによって稜線上の測色用座標について総ての可動範囲での移動を行ってずれを算出している。尚、番号ｋは測色用座標について端から順に付した番号であり、０〜９の値である。また、本実施形態においては、稜線上の両端の点は固定的な測色用座標であるとしており、上記ステップＳ１２５〜Ｓ１３７では番号１〜８について可動範囲内での移動を行うようになっている。
【０１０２】
上記ステップＳ１３０のずれ算出処理では、所定の記憶領域内にその時点でずれが最小となっている測色用座標列ＸをＸbestとして記録してある。従って、上記ステップＳ１２５〜Ｓ１３７のループが終了した時点でのＸbestは、上記ずれが極小化したときの測色用座標列である。そこで、ステップＳ１４０においては、処理中の稜線について当該Ｘbestを測色用座標として特定する。ステップＳ１４５では、ＣＭＹ色空間中の全稜線について上記測色用座標の特定が終了したか否かを判別する。
【０１０３】
ステップＳ１４５にて全稜線について上記測色用座標の特定が終了していると判別されないときには、ステップＳ１５０にて処理対象となる稜線を変更し、ステップＳ１１５〜ステップＳ１４５の処理を繰り返す。ステップＳ１４５にて全稜線について上記測色用座標の特定が終了していると判別されたときには、当該特定されている稜線上の測色用座標を稜線上の測色用座標として決定（ステップＳ１５５）し、作業を終了する。すなわち、上記ステップＳ１１５〜ステップＳ１５０における処理が上記試行工程に該当する。
【０１０４】
次に、上記ステップＳ１１５における距離制限決定処理を詳細に説明する。図１６は当該距離制限決定処理のフローチャートである。ステップＳ２００では、処理中の稜線上の全階調値、すなわち階調値０〜２５５を上記基準対応関係のスプライン補間関数に代入しそれぞれの補間Ｌａｂ値ｐi（ｉは０〜２５５）を算出する。すなわち、本実施形態では階調値０〜２５５の座標が上記請求項に言う特定の座標に該当する。ステップＳ２１０では、当該算出したＬａｂ値ｐiを使用して上記式（１）を計算し、ＣＭＹ色空間中の稜線に該当するＬａｂ色空間中の曲線の長さｌを算出する。そして、ステップＳ２２０では当該Ｌａｂ色空間中の曲線の長さｌを使用して上記式（２）を計算し、Ｌmaxを求める。本実施形態においてはｎ−１＝９である。
【０１０５】
この結果、処理中の稜線について、隣り合う測色用座標のＬａｂ色空間中で距離の最大値が算出されることとなる。一方、本実施形態において、隣り合う測色用座標のＬａｂ色空間中で距離の最小値Ｌminはデフォルトで０．７と決められている。このようにして、ＬmaxとＬminとが決定すれば、上記隣接行列を算出することができる。すなわち、ステップＳ２３０では上記条件式（３）に基づいて要素ａijを決定し、隣接行列を生成する。上述のように、この隣接行列から各測色用座標の可動範囲を判別することができるので、当該隣接行列の決定によって距離制限が決定されたと言える。
【０１０６】
次に、上記ステップＳ１３０におけるずれ算出処理を詳細に説明する。図１７は当該ずれ算出処理のフローチャートである。ステップＳ３００では、測色用座標列Ｘの各座標値を上記基準対応関係のスプライン補間関数に代入し、これら測色用座標列Ｘに対応したＬａｂ値を算出する。この結果、ＣＭＹ色空間中の測色用座標列ＸとそのＬａｂ値が算出されたことになるので、ステップＳ３１０にてこれらを使用してスプライン補間関数を決定する。当該決定されるスプライン補間関数が上記試行用対応関係である。
【０１０７】
ステップＳ３２０では、処理中の稜線上の全階調値、すなわち階調値０〜２５５を上記試行用対応関係のスプライン補間関数に代入しそれぞれの補間Ｌａｂ値ｑi（ｉは０〜２５５）を算出する。ステップＳ３４０では、座標ｑiと座標ｐiとのずれを評価するために下記式（５）の演算を行って各座標のずれの平均値Ｅを算出する。
【０１０８】
【数５】

むろん、ここでは平均値Ｅを使用することが必須というわけではなく、積算値でも良い。
【０１０９】
この平均値については、その時点での最小値を予め記録することとしてあり、当該最小値をＥbestとしてある。そこで、上記ステップＳ３４０で算出した平均値ＥがＥbestより小さいか否かをステップＳ３５０で判別する。尚、ある稜線についてずれ算出処理が初めて実行されたときには上記式（５）によって算出された値をＥbestに代入する。ステップＳ３５０にて平均値ＥがＥbestより小さいと判別されたときにはステップＳ３６０で上記測色用座標列ＸをＸbestとして記録し、平均値ＥをＥbestとして記録する。
【０１１０】
ステップＳ３５０にて平均値ＥがＥbestより小さいと判別されないときには、上記ステップＳ３６０はスキップする。そして、ステップＳ３７０では、上記条件式（４）を満たすｘkの総てについてステップＳ３００〜Ｓ３６０の処理が終了したか否かを判別し、同ステップＳ３７０にて条件式（４）を満たすｘkの総てについて処理が終了したと判別されないときにはステップＳ３８０でｘkの座標位置を変更しステップＳ３００以降の処理を繰り返す。ステップＳ３７０にてにて条件式（４）を満たすｘkの総てについて処理が終了したと判別されたときには上記図１５に示す処理に復帰する。
【０１１１】
すなわち、ステップＳ３７０における判別により、ｘkの座標位置が変更されると測色用座標列Ｘのうちｋ番目の座標位置が変更され、ｋ番目の座標は可動範囲内で移動されて逐次ずれが算出される。また、図１７に示すずれ算出処理は図１５に示すステップＳ１２５〜Ｓ１３７のループの一部であるので、この処理は１番目からｎ−２番目の総ての測色用座標について実行される。以上のようにステップＳ１２０，Ｓ３８０が上記測色用座標仮決定工程に該当し、ステップＳ３００，３１０が上記試行用対応関係規定工程に該当し、ステップＳ３５０，Ｓ３６０およびステップＳ１４０が上記測色用座標特定工程に該当する。
【０１１２】
（６）測色用座標を使用して作成されたＬＵＴに基づく色変換処理：
以上のようにして稜線上の測色用座標を特定し、ＣＭＹ色空間中の全域に渡る測色用座標を決定すると、上述のようにＬＵＴを作成する際にＣＭＹ座標とＬａｂ座標とを高精度に色変換可能になる。従って、作成されたＬＵＴの色変換精度が非常に高くなり、非常に高精度の色変換を行いつつ印刷を実行することができる。当該作成されたＬＵＴは、汎用的なコンピュータにて汎用的に行われている印刷処理にて使用可能である。図１８は、印刷時に当該ＬＵＴを使用するコンピュータ構成例を示すブロック図である。
【０１１３】
同図において、コンピュータ１４０は汎用的なパーソナルコンピュータであり、プリンタドライバ（ＰＲＴＤＲＶ）２１０と入力機器ドライバ（ＤＲＶ）２２０とディスプレイドライバ（ＤＲＶ）２３０とがＯＳ２００に組み込まれている。ディスプレイＤＲＶ２３０はディスプレイ１８０における画像データ等の表示を制御するドライバであり、入力機器ＤＲＶ２２０はシリアル通信用Ｉ／Ｏ１９０ａを介して入力される上記キーボード３１０やマウス３２０からのコード信号を受信して所定の入力操作を受け付けるドライバである。
【０１１４】
ＡＰＬ２５０は、カラー画像のレタッチ等を実行可能なアプリケーションプログラムであり、利用者は当該ＡＰＬ２５０の実行下において上記操作用入力機器を操作して当該カラー画像を印刷装置４００にて印刷させることができる。このようなカラー画像の印刷時に本発明によって作成されたＬＵＴが参照される。ＡＰＬ２５０にて作成されるカラー画像のカラー画像データ１５０ａはＲＧＢの各色成分を階調表現したドットマトリクス状のデータであり、ｓＲＧＢ規格に準拠したデータであるとともに、ＨＤＤ１５０に保存される。
【０１１５】
上記ＰＲＴＤＲＶ２１０は印刷を実行するために、画像データ取得モジュール２１０ａと色変換モジュール２１０ｂとハーフトーン処理モジュール２１０ｃと印刷データ生成モジュール２１０ｄとを備えている。また、本発明によって決定された測色用座標を使用して作成されたＬＵＴ１５０ｂはＨＤＤ１５０に保存されている。ＡＰＬ２５０実行時に利用者が印刷実行指示を行うと、印刷にかかる画像データ１５０ａが画像データ取得モジュール２１０ａに取得され、画像データ取得モジュール２１０ａは上記色変換モジュール２１０ｂを起動する。色変換モジュール２１０ｂは、ｓＲＧＢ階調値をＣＭＹＫ階調値に変換するモジュールであり、上記１７^３個の参照点から６４^３個の参照点を生成するとともに、これらの参照点を使用して補間演算を行って任意のｓＲＧＢドットデータをＣＭＹＫドットデータに変換する。
【０１１６】
色変換モジュール２１０ｂが色変換を行ってＣＭＹＫの階調データを生成すると、当該ＣＭＹＫの階調データは上記ハーフトーン処理モジュール２１０ｃに受け渡される。ハーフトーン処理モジュール２１０ｃは、各ドットのＣＭＹＫ階調値を変換してインク滴の記録密度で表現するためのハーフトーン処理を行うモジュールであり、変換後の記録密度でインクを付着させるためのヘッド駆動データを生成する。印刷データ生成モジュール２１０ｄはかかるヘッド駆動データを受け取って、印刷装置４００で使用される順番に並べ替えるラスタライズを行う。このラスタライズの後、画像の解像度などの所定の情報を付加して印刷データを生成し、パラレル通信用Ｉ／Ｏ１９０ｂを介して印刷装置４００に出力する。印刷装置４００においては当該印刷データに基づいて上記ディスプレイ１８０に表示された画像を印刷する。
【０１１７】
この印刷処理において、色変換は本発明によって決定された測色用座標を使用して作成されたＬＵＴを参照して行われるので、ディスプレイ１８０および印刷装置４００の色域全域に渡って高精度に色変換を行うことが可能であり、トーンジャンプの無い高画質の印刷を実施することができる。尚、以上の説明はＰＲＴＤＲＶ２１０による非常に汎用的な印刷処理である。従って、本発明にかかる測色用座標決定方法にて測色用座標を決定し、ＬＵＴを作成すれば、従来の印刷処理にて使用されていたＬＵＴを本発明によるＬＵＴに置き換えるだけで多くの印刷装置にてハードウェア構成を全く変えることなく、非常に高画質の印刷を実行可能にすることができる。
【０１１８】
以上説明したように、本発明においては第１色空間における測色用座標より多数の基準座標によって基準対応関係を規定し、測色用座標によって試行用対応関係を規定し、特定の座標について基準対応関係と試行用対応関係とから第２色空間中の座標を算出する。そして、それぞれの第２色空間中の座標のずれを極小化するように試行を繰り返す。従って、必要充分な数の測色用座標によって表色空間の全域に渡って高精度に色の対応関係を規定することが可能となる。
【図面の簡単な説明】
【図１】測色用座標決定方法の工程を概略的に説明する説明図である。
【図２】測色用座標決定方法の各工程の関係を示すブロック図である。
【図３】測色用座標決定装置の構成例を示すブロック図である。
【図４】測色用座標によって印刷するパッチＰの例を説明する説明図である。
【図５】ＣＭＹ色空間に形成される立方体を示す図である。
【図６】Ｌａｂ色空間に形成される立体を示す図である。
【図７】スプライン補間関数の決定を説明するための説明図である。
【図８】スプライン補間関数の決定を説明するための説明図である。
【図９】スプライン補間関数の決定を説明するための説明図である。
【図１０】スプライン補間関数による色変換精度を説明する説明図である。
【図１１】距離制限を説明する説明図である。
【図１２】隣接行列の説明図である。
【図１３】距離制限の判別が実施される様子を説明する説明図である。
【図１４】稜線上の測色用座標からＣＭＹ色空間中の測色用座標を決定する様子を説明する説明図である。
【図１５】測色用座標決定作業の全体を示すゼネラルフローである。
【図１６】距離制限決定処理のフローチャートである。
【図１７】ずれ算出処理のフローチャートである。
【図１８】印刷時にＬＵＴを使用するコンピュータ構成例を示すブロック図である。
【符号の説明】
１０…コンピュータ
１１…ディスプレイ
１２…プリンタ
２０…測色器
３０…基準座標測色値
３１…基準対応関係規定部
３１ａ…スプライン補間演算モジュール
３２…試行部
３２ａ…試行モジュール
３３…測色用座標仮決定部
３３ａ…測色用座標仮決定モジュール
３４…試行用対応関係規定部
３５…ずれ算出部
３５ａ…ずれ算出モジュール
３６…測色用座標特定部
３６ａ…測色用座標決定モジュール
３７…初期座標
３８…隣接行列
３８ａ…隣接行列データ
３９ａ…ＬＵＴ作成モジュール
３９ｃ…測色用座標測色値[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a color measurement coordinate determination method, a color conversion method, a color measurement coordinate determination device, a color conversion device, a medium in which a color measurement coordinate determination program is recorded, a medium in which a color conversion program is recorded, and a color measurement coordinate determination. The present invention relates to a program, a color conversion program, and a color conversion table.
[0002]
[Prior art]
When handling images that are common to multiple image devices such as displays and printers, the colors of both images are converted with reference to a color conversion table, ICC profile, etc., in order to obtain an image with a common color on both image devices. is doing. There are various methods for creating these color conversion tables and ICC profiles, but as an important process at the time of creation, there is a relationship between the output color of an image device based on arbitrary image data and the color in a specific absolute color space. There is a process that defines the correspondence.
[0003]
By accurately defining the correspondence between the output color of the image device and the color in the absolute color space by this correspondence, it contributes to improving the accuracy of the finally created color conversion table and ICC profile. It is important to accurately define this correspondence for the accuracy of the color conversion table and ICC profile. It is preferable in terms of accuracy if the correspondence between arbitrary image data of the image equipment and the color in the absolute color space can be specified sequentially, but if the correspondence between arbitrary image data is specified, the amount of data becomes enormous. It is impossible.
[0004]
Therefore, usually, the output color of a plurality of representative image data (coordinates for colorimetry) of the image device is measured to obtain accurate coordinates in the absolute color space, and for arbitrary image data, these representative image data The coordinates in the absolute color space are calculated by interpolation from As a conventional colorimetric coordinate determination method for determining the colorimetric coordinates, there is a method of extracting coordinates at regular intervals in the color space of the imaging device.
[0005]
[Problems to be solved by the invention]
The conventional color measurement coordinate determination method described above has the following problems.
That is, since the interpolation function is defined as a curve in the absolute color space in the above-described interpolation calculation, whether or not an interpolation function preferable for accuracy can be obtained depends on the coordinate position in the absolute color space to which the colorimetric coordinates correspond. And the coordinates should not be too far away from each other. However, in the conventional colorimetric coordinate determination method, only the color space of the image device is considered when determining the colorimetric coordinates, and the colorimetric coordinates are determined so as to obtain an interpolation function that is preferable in terms of accuracy. I could not.
The present invention has been made in view of the above problems, and a colorimetric coordinate determination method for determining a colorimetric coordinate capable of defining a color correspondence relationship with high accuracy over the entire color space, Providing a color conversion method, a color measurement coordinate determination device, a color conversion device, a medium recording a color measurement coordinate determination program, a medium recording a color conversion program, a color measurement coordinate determination program, a color conversion program, and a color conversion table With the goal.
[0006]
In order to achieve the above object, the invention according to claim 1 is used in predetermined image equipment.First color spaceMeasure the output color at multiple coordinates ofSecond color spaceBy obtaining the coordinates of the first color space andthe aboveIn the color measurement coordinate determination method for determining the plurality of coordinate positions as the color measurement coordinates in defining the correspondence with the second color space, the first color spaceCoordinatesInAbout the second color spaceGet the coordinates ofAnd aboveWith the first color spacethe aboveWith the second color spaceA standard correspondence defining step for defining the correspondence relationship between the first color space and the coordinates of the first color space in the standard correspondence defining stepAt coarser intervalsDetermine the coordinates, and based on the determined coordinates and the reference correspondenceWith the first color spacethe aboveWith the second color spaceA trial correspondence definition process for defining correspondence;the aboveAdjacent of the determined coordinatesAbout coordinatesthe aboveDefined by standard correspondenceIn the second color spaceThe distance between the coordinatesGive limits and aboveThe specific coordinates in the first color space are defined by the above standard correspondence relationship.Second color spaceStipulated by the coordinates of the above and the correspondence for trialthe aboveCalculate the deviation from the coordinates in the second color spaceCalculation process and deviationCoordinates when color is minimized as colorimetric coordinatesDecisionDoProcessWhen,ComprisingIn the calculation step, in the second color space defined by the reference correspondence relationship, all coordinates that can exist on a line connecting the reference coordinates defined by fixing the predetermined color while making the predetermined color variable. The maximum distance between the coordinates in the second color space corresponding to the adjacent coordinates by a value obtained by dividing the reference length obtained by integrating the distances between the coordinates by (number of coordinates for colorimetry-1) and a predetermined constant. AndAs a configuration.
[0007]
In the invention according to claim 1 configured as described above, when a color conversion table, an ICC profile, or the like is created, the output color from a predetermined image device is measured to obtain coordinates in the second color space. In defining the correspondence between the first color space and the predetermined second color space, a plurality of colorimetric coordinates in the first color space are determined by the following steps.In the reference correspondence defining step, the coordinates in the second color space are acquired for the coordinates of the first color space, and the reference correspondence between the first color space and the second color space is defined.
[0008]
Further, in the color measurement coordinate provisional determination step, the color measurement coordinates are provisionally determined for trial at coarser intervals than the reference coordinates, and in the trial correspondence defining step, the provisionally determined color measurement coordinates are determined. The coordinates of the second color space are obtained based on the standard correspondence relationship, and the coordinates of the first color space and the second color space are determined using the coordinates of the second color space and the provisionally determined colorimetric coordinates as reference points. Specify correspondence for trial. That is, the reference correspondence is defined from the reference coordinates and the color measurement results, and the trial correspondence is defined from the colorimetric coordinates smaller than the reference coordinates. Examples of the correspondence relationship based on the reference points include a method of defining a spline interpolation function by these reference points and an interpolation operation such as tetrahedral interpolation. Generally, the correspondence relationship for trial with fewer reference points is more suitable. Less accurate than standard correspondence. In the present invention, since the error is minimized by the trial described later, the correspondence between the first color space and the second color space can be defined with high accuracy by the colorimetric coordinates smaller than the reference coordinates.
[0009]
That is, in the trial step, the colorimetric coordinates are determined while limiting the distance between the coordinates in the second color space defined by the reference correspondence relationship for the adjacent colorimetric coordinates in the colorimetric coordinate provisional determination step. The coordinates in the second color space defined by the reference correspondence with respect to specific coordinates in the first color space are defined by temporarily determining the coordinates and defining the trial correspondence in the trial correspondence defining step. The step of calculating the deviation from the coordinates in the second color space defined by the trial correspondence is repeated. As a result, the colorimetric coordinates are moved within the distance limit between the coordinates in the second color space, and the deviation in each state is calculated. In the color measurement coordinate specifying step, the coordinate when the deviation is minimized in the trial step is specified as the color measurement coordinate.
[0010]
That is, the colorimetric coordinates are determined so that the coordinates in the second color space obtained by the trial correspondence relationship and the coordinates in the second color space obtained by the reference correspondence relationship are closest. The optimum accuracy for defining the correspondence relationship between the first color space and the second color space is calculated in the coordinates for colorimetry having a smaller number of coordinates than the coordinates. By introducing the above steps for determining all or part of the colorimetric coordinates existing in the first color space, it is possible to define the color correspondence relationship with high accuracy over the entire color space. It becomes possible. In addition, since the colorimetric coordinates are moved within the distance limit between the coordinates in the second color space, these coordinates do not approach or leave more than necessary, and the color is defined by the colorimetric coordinates. It is possible to prevent the accuracy from partially deteriorating and to obtain a preferable accuracy in the entire color space.
[0011]
The predetermined image device in the present invention is not particularly limited as long as it is a device that performs color conversion using a color conversion table, an ICC profile, or the like when exchanging image data with another image device. For example, various image devices such as a display and a printer can be employed. Of course, the device does not have to be a separate unit. For example, it can be said that a printer is integrated with a printer in a fax machine, and the present invention can be applied to such a fax machine.
[0012]
In the color measurement coordinate determination method according to the present invention, a plurality of color measurement coordinates are determined by the above-mentioned minimization, but the color measurement coordinates are provisionally determined in a plurality of areas of the first color space, and the respective areas are determined. With the above minimization, color measurement coordinates can be determined in a plurality of areas of the first color space. If the number of color measurement coordinates reaches a certain number, the color measurement coordinate position is changed from the color measurement coordinates to another color measurement coordinate position. The other colorimetric coordinates can be determined without going through the above minimization. That is, it is not essential to define the reference coordinates in the entire area of the first color space and measure the output color, but if the reference coordinates are specified for some areas and the colors are measured, the entire area of the first color space is measured. Existing colorimetric coordinates can be determined. In addition, the coordinates in the color space are usually necessary and sufficient as coordinates composed of three components, but may be coordinates composed of more components. Specifically, the coordinates of the three color values used in the image equipment as components, the coordinates of the component colors that can uniquely identify the colors, and the like correspond to the coordinates in the present invention.
[0013]
In the reference correspondence defining step and the trial correspondence defining step, the correspondence is defined using the coordinates in the first color space and the coordinates in the second color space associated with the first color space as reference points. Various interpolation methods such as spline interpolation can be employed as described above. In spline interpolation, a spline interpolation function is calculated at the time of interpolation calculation, and the coordinates of the second color space can be calculated from the coordinates of the first color space other than the reference point by the spline interpolation function. It can be said that the correspondence with the two-color space is defined. Even if the correspondence is defined, in the present invention, it is necessary to accurately determine the coordinates for color measurement without increasing the number of coordinates for color measurement more than necessary, and it is necessary to measure all areas in the first color space at once. Since it is not necessary to determine the color coordinates, it is only necessary that the correspondence relationship can be defined with high accuracy by a limited number of reference coordinates for a part of the first color space.
[0014]
Therefore, it is preferable to select the reference coordinates so that a highly accurate spline interpolation function can be determined for a part of the region in the first color space, and the reference coordinates are arranged linearly in the color space, or the reference coordinate components It is good to fix a part of. Further, since the above-described reference correspondence and trial correspondence are defined by the interpolation function, the specific coordinates in the first color space used when calculating the deviation in the trial step are the interpolation function. It is preferable to determine so that the coordinates in the second color space can be calculated directly from the above.Further, as an example of a configuration for limiting the distance between coordinates in the second color space, the calculation step is defined by fixing a predetermined color while changing a predetermined color among colors used in the imaging device. The reference length obtained by integrating the distances between the coordinates in the second color space defined by the reference correspondence relationship for all the coordinates that can exist on the line connecting the reference coordinates is (the number of colorimetric coordinates minus 1). A value obtained by multiplying the divided number by a predetermined constant is set as the maximum distance between the coordinates in the second color space corresponding to the adjacent colorimetric coordinates.
That is, in the calculation step, the colorimetric coordinates are moved within the distance limit to calculate the shift, and therefore the colorimetric coordinates are preferably present on a line connecting the reference coordinates. The movable range is on the line. Since the coordinates in the second color space corresponding to the coordinates for colorimetry defined based on the reference correspondence relationship also form a line in the second color space, in order to determine the coordinates for colorimetry preferable for accuracy, The coordinates of this second color space should not be too close or too far apart on this line. The reference length is the length of this line, and when this is divided by (the number of coordinates for colorimetry −1), the value is obtained when the coordinates in the second color space are evenly arranged from end to end of this line. It becomes length. Therefore, the maximum distance is a constant multiple of the distance when the coordinates are evenly arranged on the line formed in the second color space.
As a result, it is possible to determine the colorimetric coordinates so that the coordinates in the second color space are not separated from each other at a distance larger than a constant multiple based on the case of the uniform arrangement. Accordingly, it is possible to prevent the accuracy from being partially lowered while using a limited number of colorimetric coordinates. Here, as a specific example on the line connecting the reference coordinates, for example, when the reference coordinates are selected on the ridge line of the first color space, the ridge line is the line. Further, even if all the coordinates that can exist on the line are considered as described above, in the present invention, there is a limitation that the coordinate value is usually an integer, so the coordinate interval is not arbitrary and the number of coordinates Is finite.
[0015]
According to a second aspect of the present invention, the first color space is a color space composed of predetermined three colors used in an imaging device, and the second color space is a uniform color space. That is, in the trial process, it is necessary to evaluate the deviation between the coordinates in the second color space defined by the reference correspondence relationship and the coordinates in the second color space defined by the trial correspondence relationship. In the color space, since the distance difference in the space corresponds to the color difference, that is, the color difference, the evaluation of the shift in the second color space is equivalent to the evaluation of the color shift. Therefore, by evaluating the deviation in the second color space, it is possible to easily evaluate whether or not the color definition accuracy based on the colorimetric coordinates is sufficiently close to the colorimetric value.
[0016]
Specific examples of the uniform color space include Luv space and Lab space (usually these spaces are L^*u^*v^*Or L^*a^*b^*etc,^*In this specification, it is omitted for simplicity. The same shall apply hereinafter). Further, by making the first color space a color space having three predetermined components used by the image device, the coordinates for color measurement can be determined for the color directly used by the image device. It is possible to generate a color conversion table in consideration of the device dependence of the image device by outputting the color at the color coordinates. Specifically, a color space including components of CMY (cyan, magenta, yellow) used in the printer can be used.
[0017]
Further, as a specific example for defining the value of the maximum distance, in the invention according to claim 3, the predetermined constant is 2 or more and 3 or less. In other words, the color measurement coordinate cannot be moved unless the maximum distance is greater than the distance when the coordinates are evenly arranged as described above, and therefore it is essential that the predetermined constant be 1 or more. As a result of repeated trials by the applicant experimentally under this essential requirement, if the predetermined constant is a value of about 2, a minimal solution can be derived by the calculation process, and if the predetermined constant is a value of 3 or more, It has been found that even when minimization is performed, the accuracy tends to be reduced when creating a color conversion table or ICC profile. Therefore, when the predetermined constant is set to 2 or more and 3 or less, it is preferable in accuracy and trial processing.
[0042]
As described above, according to the first aspect of the present invention, it is possible to define the color correspondence relationship with high accuracy over the entire color space, and the coordinates in the second color space are more than necessary. Provides a colorimetric coordinate determination method that can prevent the specified accuracy of the colorimetric coordinates from partially deteriorating without approaching or moving away from the image, and that can achieve a favorable accuracy in the entire color space. can do.Furthermore, it is possible to prevent the accuracy from being partially lowered while using a limited number of colorimetric coordinates.
  According to the second aspect of the present invention, it is possible to easily evaluate whether or not the color definition accuracy based on the colorimetric coordinates is sufficiently close to the colorimetric value.
Furthermore, according to the third aspect of the invention, it is possible to obtain a predetermined constant value that is suitable in view of accuracy and feasibility of trial processing.
[0048]
(1) Configuration of the present invention:
FIG. 1 is an explanatory diagram for schematically explaining the steps of the colorimetric coordinate determination method according to the present invention, and FIG. 2 is a block diagram showing the relationship between the steps. In FIG. 1, a computer 10 can be realized by a general-purpose personal computer, and a display 11 and a printer 12 are connected to each other. In this embodiment, colorimetric coordinates are determined in the CMY color space used by the printer 12. At this time, by performing recursive calculation, the coordinate for color measurement is specified for each ridge line (12 ridge lines) in the CMY color space, and the outer peripheral surface of the CMY color space is further determined from the specified coordinate for color measurement. Determine the colorimetric coordinates that exist inside.
[0049]
If the number of colorimetric coordinates is set to 10, and a patch is printed by the printer 12 for each coordinate, the color of the patch is measured and the coordinates in the Lab color space are obtained, and an arbitrary interpolation operation based on the result is performed. When the coordinate values in the CMY color space and the coordinate values in the Lab color space are associated with each other, the color in which the color difference between the true Lab coordinate value and the interpolation value is not less than a predetermined value can be suppressed to not more than a predetermined% of the whole. . In the present embodiment, the color having a color difference of 2.0 or more can be suppressed to 10% or less of the whole.
[0050]
In this way, after determining the colorimetric coordinate position that associates the CMY color space with the Lab color space with high accuracy, the CMY color space and the sRGB color space used in the display 11 are obtained using this correspondence. It can be associated with high accuracy, and a high accuracy color conversion table (LUT) can be created. The printer 12 in this embodiment is a four-color printer that uses black (K) ink in addition to CMY ink.³The correspondence between the sRGB space and the CMYK space is defined by a single reference point.
[0051]
The above configuration will be described below with reference to the drawings. In the computer 10 shown in FIG. 1, data of 256 gradations (integer values of 0 to 255) of CMY colors is given to the dot matrix-like pixels to express a desired color, and a color image is printed by the printer 12. be able to. In the present embodiment, ten colorimetric coordinates are specified for each ridge line in the CMY color space, and the correspondence between the CMY color space and the Lab color space can be accurately defined. First, a correspondence relationship as a reference for the recursive calculation is acquired so that the individual colorimetric coordinates can be derived by the recursive calculation.
[0052]
For this purpose, first, 16 reference coordinates that are more than 10 are defined for each ridge line in the CMY space. Then, the printer 10 is controlled by the computer 10 to print the patch P based on these CMY coordinates. The colorimeter 20 can measure the color patch P with high accuracy and obtain the Lab coordinates of the color, and the CMY coordinates and Lab coordinates of each patch P are accurately associated by the color measurement (FIG. 2). Then, the reference coordinate colorimetric value 30). If the spline interpolation function is determined with reference to these 16 sets of coordinates, any CMY coordinate and Lab coordinate on each edge can be associated with each other very accurately. The step of determining the spline interpolation function is a step of defining the reference correspondence, and the reference correspondence defining unit 31 corresponds to FIG. Here, the number of reference coordinates “16” is a necessary and sufficient number for associating the CMY color space and the Lab color space with an extremely high accuracy for an arbitrary coordinate on each ridge line.
[0053]
Thus, since the reference correspondence is highly accurate, the trial correspondence is defined by the ten colorimetric coordinates, and the Lab coordinates for any CMY coordinates calculated based on the trial correspondence. Is sufficiently close to the Lab coordinates calculated based on the reference correspondence, it can be said that the trial correspondence is sufficiently accurate. Therefore, the trial unit 32 and the colorimetric coordinate specifying unit 36 in FIG. 2 move the ten colorimetric coordinates on each ridge line by recursive calculation, and refer to the respective coordinate positions after the movement to determine the spline interpolation function. To determine the correspondence for trial. If Lab coordinates corresponding to all coordinates on the CMY color space ridge line are obtained for each state by the trial correspondence, the deviation from the Lab coordinates calculated by the reference correspondence can be calculated. The coordinate position for colorimetry in a state where is minimized can be specified as the optimum coordinate position.
[0054]
That is, the trial unit 32 includes a colorimetric coordinate provisional determination unit 33, a trial correspondence definition unit 34, and a deviation calculation unit 35. In the color measurement coordinate provisional determination unit 33, a predetermined initial coordinate 37 (this In the embodiment, the coordinates are uniformly arranged on the ridgeline), and 10 colorimetric coordinates are moved on each ridgeline in each trial process. The trial correspondence defining unit 34 determines a spline interpolation function for each colorimetric coordinate and defines the trial correspondence. The deviation calculation unit 35 obtains Lab coordinates for each gradation value on each ridge line by using the reference correspondence relation and the trial correspondence relation, and calculates the deviation between the two coordinates. When the trial unit 32 repeats these steps, the colorimetric coordinate position in which the deviation is minimized can be grasped, and the colorimetric coordinate specifying unit 36 determines the optimum measurement by discriminating this deviation. The color coordinates can be grasped.
[0055]
In the color measurement coordinate provisional determination unit 33, the distance between adjacent coordinates is limited when the color measurement coordinates are moved on each ridge line. This distance is defined by the distance between coordinates in the Lab space as will be described later, and a flag indicating whether or not another coordinate can exist next to a certain coordinate before the trial by the trial unit 32 is an element. The adjacency matrix 38 is calculated in advance. In the adjacency matrix 38, whether or not a coordinate point can exist next to all gradation values on the edge of the CMY color space is specified, so there is no need to perform sequential conditional branching when the trial unit 32 tries. It is only necessary to refer to the adjacency matrix 38.
[0056]
If the coordinate position for color measurement on the ridge line is specified, the intersection point when these coordinate points are connected by a line is taken as the coordinate position for color measurement on the outer periphery of the CMY color space, and this grid point is connected by a line As the coordinate position for colorimetry inside the CMY color space, the intersection of³The coordinate position for color measurement is determined. When the coordinate positions for color measurement over the entire CMY color space are determined, a patch is printed from the printer 12 using these coordinates, and the color meter 20 measures the color of the patch.
[0057]
In order to create an LUT that defines the correspondence between the coordinates in the CMYK color space and the coordinates in the sRGB color space, the CMY coordinates are converted into Lab coordinates, the sRGB coordinates are converted into Lab coordinates, and both are converted in the Lab space. Correspondingly, the CMY coordinates are separated into CMYK coordinates. Here, since the coordinates for color measurement over the entire CMY color space and the colorimetric values (Lab coordinate values) can be associated with coordinates in any CMY color space and Lab coordinates with high accuracy, LUT creation is possible. The above conversion work can be performed with high accuracy, and a high-precision LUT can be created. Lab coordinates corresponding to arbitrary CMY coordinates can be calculated by spline interpolation. The correspondence relationship between the coordinates in the sRGB color space and the Lab coordinates can be obtained by calculating with a known formula or by measuring the color.
[0058]
In this embodiment, the color separation process is performed because the printer 12 is a four-color printer. However, in the case of a three-color printer, the color separation process can be omitted, and light cyan, light magenta, dark yellow, etc. LUTs for 6-color and 7-color printers may be created by performing color separation. In addition, it is not essential to correspond to the sRGB color space. The RGB color space unique to a device such as a scanner and the CMY color space may be associated with each other. The LUT may be associated with the CMY color space. Of course, the colorimetric coordinates may be used when creating the ICC profile instead of the LUT.
[0059]
The colorimetric coordinate determination method according to the present invention is realized by performing many calculations and repeating the calculations. Therefore, it is preferable to implement the method using a computer. In this case, the computer constitutes a colorimetric coordinate determination apparatus, and the executed program is a colorimetric coordinate determination program. FIG. 3 is a block diagram illustrating a configuration example of the colorimetric coordinate determination apparatus. In the figure, the computer 100 can be constituted by a general-purpose personal computer or the like, and the APL 110 that can be executed by the computer 100 corresponds to the colorimetric coordinate determination program.
[0060]
The computer 100 includes a user interface such as a display and a keyboard, a flexible disk drive (FDD) 130, a hard disk drive (HDD) 120, and the like, and can execute the APL 110 under operating system control (not shown). The APL 110 is composed of a plurality of programs obtained by modularizing each unit shown in FIG. 2. The trial module 32 a calculates the minimum value of the deviation as a result of repeated trials, and the colorimetric coordinate determination module 36 a minimizes the deviation. The coordinate position for colorimetry at the time point is set as the coordinate for colorimetry on each ridge line, and the coordinate position for colorimetry in the CMY color space is determined from these colorimetric coordinates.
[0061]
When the coordinate position for color measurement is determined, printing is performed using the coordinate value, and the print result is measured. The coordinate colorimetric values 39c for colorimetry are converted into data, read out via the FDD 130, and supplied to the LUT creation module 39a. The LUT creation module 39a executes the color separation processing as described above based on these colorimetric coordinate colorimetric values and the colorimetric coordinate values, creates the LUT 39b and stores it in the HDD 120. Of course, the printer 12 may be connected to the computer 100, and printing using the colorimetric coordinates may be performed under the control of the computer 100.
[0062]
Specifically, the trial module 32a includes a spline interpolation calculation module 31a, a color measurement coordinate provisional determination module 33a, and a deviation calculation module 35a. The reference coordinate colorimetric value 30 is converted into computer-readable data by the colorimeter 20, read from the FDD 130 via a flexible disk (not shown), and supplied to the spline interpolation calculation module 31a. The spline interpolation calculation module 31a is a module that calculates a spline interpolation function, which will be described later, based on given data, and can calculate both the reference correspondence and the trial correspondence.
[0063]
The colorimetric coordinate provisional determination module 33a calculates the adjacency matrix 38 from the reference correspondence calculated by the spline interpolation calculation module 31a (a value obtained by substituting all the tone values on the CMY color space ridge line into the spline interpolation function), and stores it in the HDD 120. Stored as adjacency matrix data 38a. In addition, when the trial is executed, the adjacent matrix data 38a is referred to, and the colorimetric coordinates are provisionally determined while moving the colorimetric coordinates within the distance limit indicated by the matrix element. The spline interpolation calculation module 31a calculates the trial correspondence from the provisionally determined colorimetric coordinates. The deviation calculation module 35a substitutes all gradation values on the ridge line of the CMY color space for the reference correspondence relation and the trial correspondence relation calculated by the spline interpolation calculation module 31a, and calculates a deviation from both substitution results. The colorimetric coordinate values at the time when the deviation is minimized are output to the colorimetric coordinate determination module 36a.
[0064]
(2) Standard coordinate acquisition and standard correspondence relationship:
Hereinafter, each process of the colorimetric coordinate determination method will be described in detail. FIG. 4 is an explanatory diagram illustrating an example of the patch P printed by the printer 12 using the colorimetric coordinates. As described above, the patch P is printed for each coordinate considering 16 coordinates for each edge in the CMY color space. Further, since the CMY gradation values are 0 to 255 and can take independent values, the CMY color space is a cube as shown in FIG. 5 when orthogonal coordinates with CMY as axes are considered. . Since the reference coordinates exist on each ridge line of the cube and there are 12 ridge lines, the total number of reference coordinates is 192 and the number of patches P is 192. In the present embodiment, the reference coordinates are selected so as to be evenly arranged from end to end on the ridgeline. In FIG. 5, the ridge line WC (W is white) in the CMY color space is shown enlarged, and as shown in the enlarged view, the intervals between adjacent reference coordinates are all equal.
[0065]
Further, on the printing paper of the printer 12, as shown in FIG. 4, patches of coordinates existing on the same ridge line are printed in a horizontal direction. In the drawing, characters CMYRGBWK are attached to both sides of the printing paper, and cyan, magenta, yellow, red, green, blue, white, and black are respectively shown. For example, the patches arranged at the top of the printing paper are printed based on the coordinates on the edge line WC. If the color of the patch P printed in this way is measured by the colorimeter 20, the Lab coordinate value of each patch can be obtained, and the Lab coordinate corresponding to the coordinates on each ridge line in the CMY color space can be obtained. it can. That is, the reference coordinates can be acquired. Note that the configuration for selecting the reference coordinates on the ridge line is an example, and the configuration for selecting the reference coordinates on the gray axis (straight line WK), which is required to define the color correspondence with high accuracy, is adopted. You may do it.
[0066]
FIG. 6 shows a solid in the Lab color space corresponding to the cube of the CMY color space shown in FIG. When the color of the patch P is measured, Lab coordinates corresponding to the reference coordinates on each ridge line of the cube in the CMY color space are obtained. If the Lab coordinates are plotted in the Lab color space, the CMY is obtained as shown in FIG. A solid corresponding to the cube of the color space can be roughly obtained. Here, since the color measurement is performed on 192 patches as described above, only 192 coordinates are obtained even in a solid in the Lab color space, which is a part of the curve around the solid.
[0067]
The solid in the Lab color space has a distorted shape as shown in FIG. That is, even if the CMY color space is a cube, the color corresponding to the outer periphery of the CMY color space in the Lab color space does not necessarily form the outer periphery of the cube in the Lab space. For example, a ridge line WC that forms a straight line in the CMY color space becomes a curved line WC shown in FIG. 6 in the Lab color space, and is bent while connecting the WC in the Lab color space. Also, as shown in the enlarged view of FIG. 6, the Lab coordinates on the curve WC are not evenly spaced, but if there are 16 coordinates on the curve WC, these coordinates and the reference coordinates in the CMY color space When the coordinates in the Lab color space are calculated by performing spline interpolation based on the set of the two, the two can be associated with sufficient accuracy. That is, it is sufficient as a criterion to be converged by recursive calculation, and sufficient accuracy is obtained when an LUT is created based on this criterion.
[0068]
7 to 9 are explanatory diagrams for explaining the definition of the reference correspondence, that is, the determination of the spline interpolation function. In this embodiment, when determining the spline interpolation function for the reference coordinates on the edge of the CMY color space, a spline interpolation function is determined in which any one of CMY is a variable and any one of Lab is a variable. That is, since one of CMY fluctuates and the other two are fixed on the edge line in the CMY color space, the fluctuating value is used as a variable. Further, each value of Lab is considered independently, and only one of them is changed to determine the spline interpolation function. FIGS. 7 to 9 show examples showing calculation when determining the spline interpolation function with respect to the reference coordinates on the ridge line WC, and are graphs showing the relationship between L-C, a-C, and b-C, respectively. .
[0069]
In the figure, the horizontal axis is all C gradation values, the vertical axis is L, a, and b, respectively, and ● on the graph is the value at the reference coordinates. Since there are 16 reference coordinates, 16 points are determined on each graph. Since this example relates to the coordinates on the edge line WC, C changes in the CMY color space, and the other M and Y are fixed to “0”. Since the values of L, a, and b are determined for each C by the above colorimetry, plotting them on graphs for each C results in FIGS. When the spline interpolation function is obtained for this graph, the spline interpolation function Lp = f is calculated from each graph.₁(C), ap = f₂(C), bp = f₃(C) is calculated. In addition, p attached | subjected to L, a, and b of the left side has shown that it is a reference | standard correspondence.
[0070]
Here, since each spline interpolation function has one variable on both the left side and the right side, the coefficient of each variable can be easily determined, and each spline interpolation function can be determined. When the spline interpolation function is determined in this way, L, a, and b can be uniquely determined by substituting arbitrary C for each function, and Lab coordinates for arbitrary coordinates on the edge line WC are calculated. be able to. That is, the curve WC in the Lab color space corresponding to the ridge line WC can be uniquely determined, and the correspondence between both spaces is defined. If the same processing is performed for other ridge lines other than the ridge line WC, Lab coordinates corresponding to the respective ridge lines in the CMY color space are determined.
[0071]
(3) Trial process:
(3-1) Provisional determination of colorimetric coordinates and trial correspondence relationship:
Next, a trial process for minimizing the deviation between the coordinates based on the reference correspondence and the coordinates based on the trial correspondence will be described. In the present invention, the colorimetric coordinates are first determined on the edge of the CMY color space, and then the colorimetric coordinates are determined over the entire CMY color space, and the colorimetric coordinates on the edge are determined. In doing so, an attempt is made to move the temporarily determined colorimetric coordinates little by little, the coordinate position is optimized, and the colorimetric coordinates on the ridgeline are determined. For this optimization, the process of provisionally determining the coordinates for colorimetry and the process of defining the correspondence for trial are repeated, and the deviation of coordinates between the trial correspondence calculated at each trial stage and the reference correspondence is calculated. To do. The coordinate position where the deviation is minimized is the optimum coordinate position.
[0072]
In the trial process, the trial correspondence is determined using the initial value of the predetermined colorimetric coordinate in the first trial, and the initial value is a coordinate evenly arranged on each ridgeline. The number of coordinate for colorimetry is ten. FIG. 5 also shows an enlarged colorimetric coordinate (◯) on the ridge line WC of the CMY color space. Each colorimetric coordinate is a coordinate in the CMY color space, and the corresponding coordinate in the Lab color space can be obtained by substituting any one of the CMY values into the reference correspondence relationship, that is, the spline interpolation function. it can. FIG. 6 shows the colorimetric coordinates on the ridge line WC in the CMY color space as shown in FIG. 5 converted into the coordinates in the Lab color space according to the above-mentioned reference correspondence relationship defined for the ridge line WC.
[0073]
In other words, the coordinates corresponding to the coordinates evenly arranged on the edge of the CMY color space are not necessarily evenly arranged on the curve in the Lab color space. As described above, the arrangement of the corresponding coordinates in the CMY color space in the Lab color space cannot be immediately determined just by determining the coordinates in the CMY color space. In order to realize the conversion, it is important to optimize the coordinate position as in the present invention. If the corresponding coordinates in the Lab color space of the coordinates in the CMY color space are biased to a part of the curve WC or are too far apart, the CMY color space and the Lab color space are accurately determined by the trial correspondence. This is because they cannot be matched.
[0074]
The trial correspondence is defined by calculating a spline interpolation function with reference to 10 CMY coordinates on the CMY color space ridge line temporarily determined as colorimetric coordinates and Lab coordinates corresponding to the CMY coordinates. The Here, the Lab coordinate value is a value obtained by substituting the CMY coordinate value into the spline interpolation function of the reference correspondence. According to the above reference correspondence relationship, coordinates in the Lab color space can be calculated for all the coordinates of each ridge line (only one of CMY is variable from 0 to 255, and the other is fixed). Therefore, Lab coordinates corresponding to ten colorimetric coordinates existing on each ridge line can be calculated with high accuracy from the reference correspondence relationship.
[0075]
If the spline interpolation function is calculated from the colorimetric coordinates in the CMY color space and the coordinates in the Lab color space, the trial correspondence can be defined. For example, the circles shown in FIGS. 7 to 9 are examples of colorimetric coordinates on the edge line WC, and the spline interpolation function Lq = f is obtained from each graph.₄(C), aq = f₅(C), bq = f₆(C) can be calculated. Here, q attached to L, a, and b on the left side indicates a trial correspondence.
[0076]
However, since the number of colorimetric coordinates is 10, even if a spline interpolation function is calculated from these 10 colorimetric coordinates and a new trial correspondence is defined, any value on the ridgeline can be converted. When the accuracy is considered, the color conversion accuracy is inferior to that of the reference correspondence relationship normally defined from the 16 reference coordinates. That is, as shown in FIGS. 7 to 9, since the colorimetric coordinates are distributed more coarsely than the reference coordinates on each ridgeline, in general, the trial correspondence spline interpolation function obtained by ○ is used. Lq, aq, and bq have poorer fitting accuracy with respect to the curves in FIGS. 7 to 9 than the spline interpolation functions Lp, ap, and bp having the above-mentioned reference correspondence relationship.
[0077]
FIG. 10 is an explanatory diagram for explaining this situation, in which the reference coordinates are indicated by ●, certain colorimetric coordinates are indicated by ◯, and the graph is enlarged than FIG. The reference correspondence defined by the reference coordinates is indicated by a solid line, and the trial correspondence defined by the colorimetric coordinates is indicated by a broken line. In the example shown in FIG. 10, since a cubic spline interpolation function is used, if the number of coordinates for determining the spline interpolation function is small, an inflection point is inherently present as in the trial correspondence indicated by the broken line. Where the function should not be, it may be defined as a function having an inflection point.
[0078]
FIG. 10 shows such a state. When a spline interpolation function is calculated using only ◯, it should be like a solid line that connects ● to each other, but it does not have an inflection point. In this graph, since the gradation value of C is 0 to 255, Lpi and Lqi (i is 0 to 255) can be calculated by the spline interpolation function when C is 0 to 255. By comparing the two, the deviation between the reference correspondence and the trial correspondence can be determined. Since the Lab color space is a three-dimensional space, in this embodiment, a shift is calculated as a distance in the Lab color space based on each of the L component, the a component, and the b component. Since the correspondence for trial changes sequentially when the coordinate position for colorimetry is changed, how to move the coordinate for colorimetry, calculate the correspondence for trial, and calculate the deviation as one trial will show how. If the coordinate position for colorimetry is correct, it can be grasped whether the shift is minimized. By minimizing the deviation in this way, it is possible to determine the trial correspondence and the colorimetric coordinate position that can convert the color with high accuracy. In the present embodiment, the deviations in all coordinates (gradation values 0 to 255) on each ridge line are averaged and compared.
[0079]
If the number of colorimetric coordinates is 10 as in this embodiment, the colorimetric coordinates on the ridge line are used to determine the colorimetric coordinates over the entire CMY color space, and these are color-converted. By using it as a colorimetric point when creating a table, when color conversion is performed using the color conversion table, the color difference between the true value and the interpolation value is 2.0 or less can be suppressed to 10% or less. it can. That is, if there are nine colorimetric coordinates, the color with a color difference of 2.0 or more exceeds 10% of the total. In addition, since the colorimetric coordinates are distributed in a three-dimensional space (CMY color space), the number of colors that actually require colorimetry increases by the third power by increasing the colorimetric coordinates on the ridge line. A color conversion table capable of color conversion with high accuracy while keeping the number of color coordinates small can be created.
(3-2) Distance limitation:
[0080]
As described above, the trial is performed so as to minimize the deviation of the Lab coordinate calculated in each of the reference correspondence and the trial correspondence. However, the colorimetric coordinates are assumed to be freely movable. If the coordinates for use are tentatively determined, the coordinates in the Lab color space may become inconvenient as coordinates for color measurement because the coordinates are spatially dense. Therefore, in the present invention, there is a restriction on the movement of the colorimetric coordinates when the colorimetric coordinates are provisionally determined. That is, the distance between adjacent colorimetric coordinates is limited.
[0081]
FIG. 11 is an explanatory diagram for explaining the distance limitation according to the present embodiment. In the present embodiment, the maximum value of the distance between adjacent coordinates is calculated from the length of the curve in the Lab color space corresponding to the ridge line of the CMY color space. In the following description, the curve WC in the Lab color space is taken as an example. FIG. 11 is a partially enlarged view of the curve WC in the Lab color space. In the figure, on the curve WC, the coordinates in the Lab color space corresponding to the reference coordinates are indicated by ●. When the reference correspondence relationship is defined, coordinates pi (i is an integer from 0 to 255) in the Lab color space corresponding to all gradation values (any of CMY is 0 to 255) on the color space ridge line are calculated. be able to.
[0082]
As for the length of the curve WC, an approximate value l can be obtained by integrating the distance between adjacent coordinates as in the following equation (1).
[Expression 1]

Note that the coordinates pi are discrete coordinates, so even if the distances between the coordinates are integrated, the length of the curve is not strictly speaking. However, since the number of gradations is 256 and there are many, the length of the curve is almost the same. Can be considered. Hereinafter, the approximate value l will be described as being the same as the length of the curve. Moreover, the right side in Formula (1) has shown the distance between each coordinate.
[0083]
The maximum value Lmax of the distance between adjacent coordinates is a constant multiple (2.5 in this embodiment) of the distance when the distance between the coordinates is arranged on the curve WC to be equal. It can be calculated by equation (2).
[Expression 2]

In Equation (2), n is the number of coordinates for colorimetry, and in this embodiment, n = 10.
[0084]
By limiting Lmax in this way, it is possible to prevent adjacent coordinates from being separated by a constant multiple from the constant fold when the coordinates are evenly arranged on the curve WC. Therefore, it is possible to prevent the coordinates from becoming dense in a part on the curve WC. Here, 2.5 in the formula (2) can be calculated as appropriate depending on conditions such as the printing medium and experimental results, and can be changed as appropriate, but is preferably in the range of approximately 2.0 to 3.0. Since Lmax depends on the length of the curve in the Lab color space, it becomes a different value for each ridgeline in the CMY color space and is calculated respectively.
[0085]
On the other hand, in the present embodiment, the minimum value Lmin of the distance between adjacent coordinates employs a constant value of 0.7 for each ridge line. That is, since the Lab color space is a uniform color space and the distance is a color difference, the minimum value of the color difference at the coordinates can be defined by defining the minimum distance between the coordinates. Since the color recognition limit by the human eye is approximately 1.0, by setting Lmin to such a value, it is possible to prevent the coordinates from being too close, and at least a significant color change is felt by the human eye. The distance between the colorimetric coordinates can be secured. Of course, the visual recognition limit is relative, and the constant value 0.7 is an example of the value, but is preferably about 0.5 to 1.0.
[0086]
In order to determine which coordinate position can be adjacent to a certain coordinate on the curve in the Lab color space corresponding to the edge line in the CMY color space, the above equation (1) is used. Although it can be discriminated by calculating (2), it is necessary to actually calculate the distance between each coordinate and the other coordinates for the specific coordinate position. Although the distance limitation is taken into account in the trial in the present invention, since it is preferable to automatically repeat the trial process by computer calculation, it is not preferable in terms of calculation speed to execute the distance limitation condition determination one by one in the trial process.
[0087]
Therefore, in this embodiment, the calculation speed of the trial process is improved by using an adjacency matrix in which distance restriction is flagged in advance. FIG. 12 is an explanatory diagram of the adjacency matrix. The element of the adjacency matrix is aij, i and j each take an integer value of 0 to 255, and the actual value of the element aij is 0 or 1. The value of aij is defined by the following conditional expression (3).
[Equation 3]

[0088]
That is, regarding the coordinate pi in the Lab color space, the element aij is "1" when there is a coordinate pj that satisfies the distance limitation when the coordinate pj is assumed as the adjacent coordinate.
[0089]
If such an adjacency matrix is defined, a coordinate sequence X for colorimetry that is provisionally determined in the CMY color space is xk (k is 0 to 9, xk is 0 to 255) in order from the end. It is possible to determine that the k-th coordinate is within the above distance limit when the elements of the adjacency matrix satisfy the following conditional expression (4).
[Expression 4]

That is, in the adjacency matrix, since the coordinate pj (condition i <j according to the equation (3)) having a larger number than the coordinate pi of interest is used as a comparison target, the condition for adjacent coordinates on one side of the coordinates is flagged. However, considering the number k-1 and the number k + 1, it is possible to take into account the distance restriction on both sides of the coordinate of the number k.
[0090]
FIG. 13 is an explanatory diagram for explaining how distance limitation is determined by the adjacency matrix and the conditional expression (4). In the figure, for simplicity, a part of the adjacency matrix is shown, and the element values are also determined as an example for explanation as shown in the figure. In the trial process, one of the colorimetric coordinate sequences X is moved, and the above deviation is sequentially calculated. In the example shown in the figure, the first coordinate is moved as x₁It shows as'. X₀Value is 0, x₂It is assumed that the value of is 8.
[0091]
x₁When the distance limitation of ′ is considered by the above conditional expression (4), the x₁In order to consider the movement of only ', xk-1 and xk + 1 are fixed and are "0" and "8" in this example, respectively. So first a_0xk= 1. At this time, it is only necessary to consider whether or not the first row of the adjacency matrix shown in FIG.₀₂, A₀₃, A₀₄, A₀₅, A₀₆It turns out that corresponds. On the other hand, a_xk8= 1, it is only necessary to consider whether the eighth column of the adjacency matrix shown in FIG. 13 is “1”.₁₈, A₂₈, A₃₈It turns out that corresponds.
[0092]
Therefore, xk satisfying conditional expression (4) is 2 and 3, and x₁The coordinate value that can be taken by ′ is 2 or 3. So x₁For, substituting 2 and 3 defines the trial correspondence, and calculates the deviation. If the condition determination based on the adjacency matrix, the definition of the correspondence for trial, and the calculation of the deviation are performed for all the numbers k of the color measurement coordinate sequence X, the coordinates for color measurement in which the coordinates are moved within the distance limit Therefore, the deviation can be calculated for all the combinations, and therefore the minimum value of the deviation can be grasped. By performing this operation for each ridge line in the CMY color space, the coordinate for color measurement can be specified on each ridge line.
[0093]
(4) Extension of colorimetric coordinates in the first color space:
After specifying the colorimetric coordinates on each ridge line as described above, it is necessary to extend the colorimetric coordinates over the entire area in the CMY color space. FIG. 14 is an explanatory diagram for explaining how the colorimetric coordinates in the CMY color space are determined from the colorimetric coordinates on the ridge line. The figure shows a cube formed in the CMY color space, and the coordinates for colorimetry specified on the ridge line are indicated by ◯. On the outer peripheral surface of the CMY color space, an intersection when the colorimetric coordinates on the parallel ridge lines in the CMY color space are connected by a straight line is defined as a colorimetric coordinate position.
[0094]
For example, an intersection can be determined by assigning a number starting from the smallest coordinate value in each colorimetric coordinate and connecting the colorimetric coordinates having the same number with a straight line. In FIG. 14, numbers 0 to 9 are assigned to the coordinates for color measurement on the ridge line WC and the coordinates for color measurement on the ridge line YG as shown in FIG. 14, and the numbers 4 are connected by a straight line. When the colorimetric coordinates are connected by straight lines on the ridgeline WY and the ridgeline CG, they intersect each other, and the intersection is set as the colorimetric coordinate xn. If the same operation is performed for all the colorimetric coordinates for all the ridgelines, the colorimetric coordinates can be determined on the outer peripheral surface of the CMY color space.
[0095]
Further, for the colorimetric coordinates inside the CMY color space, consider the intersection of the surfaces. That is, when a straight line is drawn on the outer peripheral surface of the CMY color space as described above, a surface surrounded by the straight line is formed. In the example shown in FIG. 14, the surfaces S1, S2, and S3 surrounded by a straight line connecting the colorimetric coordinates with a certain number are formed. The surface S1 is sandwiched between the surfaces WMRY and CBKG, the surface S2 is sandwiched between the surfaces WCBM and YGKR, and the surface S3 is coupled to the surface WCGI and the surface, no matter how the position of the colorimetric coordinates changes on each ridgeline. Since these are sandwiched between MBKR, these surfaces S1, S2, and S3 intersect in the CMY color space, and a common intersection is one point. Therefore, this intersection is set as a color measurement coordinate xm in the CMY color space.
[0096]
If the same operation is performed for all the colorimetric coordinates for all the ridgelines, the colorimetric coordinates can be determined within the CMY color space. As a result, 10 in the CMY color space including the ridge line and the outer periphery.³Individual colorimetric coordinates can be determined. These 10³If you measure the color of a patch printed with individual colorimetric coordinates, 10³The CMY value and the Lab value can be accurately associated with each coordinate. If spline interpolation is performed for an arbitrary CMY value based on these coordinates, a Lab value (interpolation value) for an arbitrary CMY value can be obtained. As described above, this interpolation value and a true value (for example, actual value) The coordinates at which the color difference with respect to 2) is 2.0% or less are 10% or less of the total. Accordingly, the CMY color space and the Lab color space can be accurately associated with each other by this colorimetric value, and an accurate LUT can be created while referring to this correspondence.
[0097]
Further, according to an experiment conducted by the applicant using dye ink on plain paper, the average value of the color difference is about 1.0, and the histogram of the color difference with respect to many CMY values (the vertical axis indicates the number of distributions). If the horizontal axis is the color difference), the standard deviation is about 0.6. When the color measurement coordinates are uniformly arranged in the CMY color space while using dye ink on plain paper, the coordinates at which the color difference is 2.0 or more are 20% of the total, and the average value of the color difference is The standard deviation is about 1.4. Accordingly, it can be said that, particularly for plain paper, a very significant accuracy improvement effect can be achieved by determining the colorimetric coordinates according to the present invention.
[0098]
(5) Color coordinate determination work:
Hereinafter, an operation of determining the colorimetric coordinates on the ridge line in the CMY color space by performing the above steps as a series of operations will be described with reference to a flowchart. FIG. 15 is a general flow showing the entire color measurement coordinate determination operation. In step S100, the printer 10 is controlled by the computer 10 in step S100, and CMY values corresponding to 16 reference coordinates arranged uniformly on each ridge line in the CMY color space as shown in FIG. Execute printing. As a result, 192 patches P as shown in FIG. 4 are printed on the printing paper.
[0099]
In step S105, the color of each printed patch P is measured by the colorimeter 20, and coordinate values in each Lab color space are acquired. That is, steps S100 and S105 correspond to the reference coordinate acquisition process. In step S110, the spline interpolation function is determined with reference to the colorimetric value and the coordinate value in the CMY color space. The determined spline interpolation function is the reference correspondence. In step S115, in order to give a distance restriction between the colorimetric coordinates in the trial process, a process for determining the distance restriction for any one of the ridge lines in the CMY color space is performed.
[0100]
In step S120, ten colorimetric coordinates are evenly arranged on the edge being processed as shown in FIG. 5 as initial coordinates, and an initial colorimetric coordinate array X is obtained. In step S125, the number k is initialized to “1”, and a deviation calculation process is performed in step S130. In the deviation calculation process, as will be described later, a colorimetric coordinate for moving the kth colorimetric coordinate within the above distance limit is provisionally determined, and calculated based on the Lab coordinate calculated based on the reference correspondence and the trial correspondence. The deviation from the Lab coordinates is calculated sequentially. In step S135, it is determined whether or not the number k has reached n-2 (n is the number of coordinates for color measurement on one ridge line; in this embodiment, n = 10), and k has reached n-2. If not discriminated, k is incremented in step S137 and the processes in and after step S130 are repeated.
[0101]
That is, in the deviation calculation process, the coordinates of the number k are moved within the movable range, the deviation is calculated sequentially, and the same process is repeated for the coordinates of the numbers 1 to 8 (10-2) for color measurement on the ridgeline. The shift is calculated by moving the coordinates within the entire movable range. Note that the number k is a number assigned sequentially from the end with respect to the colorimetric coordinates, and is a value of 0 to 9. In the present embodiment, both end points on the ridge line are fixed colorimetric coordinates, and in steps S125 to S137, the numbers 1 to 8 are moved within the movable range. Yes.
[0102]
In the deviation calculation process in step S130, the colorimetric coordinate sequence X having the smallest deviation at that time is recorded as Xbest in a predetermined storage area. Accordingly, Xbest at the time when the loop of steps S125 to S137 is completed is a coordinate sequence for colorimetry when the shift is minimized. Therefore, in step S140, the Xbest is specified as a colorimetric coordinate for the edge line being processed. In step S145, it is determined whether or not the above colorimetric coordinates have been specified for all ridge lines in the CMY color space.
[0103]
If it is not determined in step S145 that the colorimetric coordinates have been specified for all ridge lines, the ridge line to be processed is changed in step S150, and the processes in steps S115 to S145 are repeated. If it is determined in step S145 that the specification of the colorimetric coordinates has been completed for all the ridgelines, the colorimetric coordinates on the specified ridgeline are determined as the colorimetric coordinates on the ridgeline (step S155). ) And finish the work. That is, the processes in steps S115 to S150 correspond to the trial process.
[0104]
Next, the distance restriction determination process in step S115 will be described in detail. FIG. 16 is a flowchart of the distance limit determination process. In step S200, all gradation values on the edge line being processed, that is, gradation values 0 to 255 are substituted into the spline interpolation function having the above reference correspondence to calculate respective interpolation Lab values pi (i is 0 to 255). . That is, in this embodiment, the coordinates of the gradation values 0 to 255 correspond to the specific coordinates described in the above claims. In step S210, equation (1) is calculated using the calculated Lab value pi, and the length l of the curve in the Lab color space corresponding to the ridge line in the CMY color space is calculated. In step S220, equation (2) is calculated using the length l of the curve in the Lab color space, and Lmax is obtained. In this embodiment, n-1 = 9.
[0105]
As a result, the maximum value of the distance is calculated in the Lab color space of the adjacent colorimetric coordinates for the ridge line being processed. On the other hand, in the present embodiment, the minimum distance value Lmin in the Lab color space of adjacent colorimetric coordinates is set to 0.7 by default. Thus, if Lmax and Lmin are determined, the adjacency matrix can be calculated. That is, in step S230, an element aij is determined based on the conditional expression (3), and an adjacency matrix is generated. As described above, since the movable range of each colorimetric coordinate can be determined from this adjacency matrix, it can be said that the distance limitation is determined by determining the adjacency matrix.
[0106]
Next, the deviation calculation process in step S130 will be described in detail. FIG. 17 is a flowchart of the deviation calculation process. In step S300, each coordinate value of the color measurement coordinate sequence X is substituted into the spline interpolation function having the above reference correspondence relationship, and Lab values corresponding to these color measurement coordinate sequences X are calculated. As a result, the colorimetric coordinate sequence X in the CMY color space and its Lab value are calculated, and thus, a spline interpolation function is determined using them in step S310. The determined spline interpolation function is the trial correspondence.
[0107]
In step S320, all gradation values on the edge line being processed, that is, gradation values 0 to 255, are substituted into the trial correspondence spline interpolation function, and each interpolation Lab value qi (i is 0 to 255) is calculated. To do. In step S340, in order to evaluate the shift between the coordinates qi and the coordinates pi, the following equation (5) is calculated to calculate the average value E of the shift between the coordinates.
[0108]
[Equation 5]

Of course, it is not essential to use the average value E here, and an integrated value may be used.
[0109]
As for this average value, the minimum value at that time is recorded in advance, and the minimum value is set as Ebest. Therefore, in step S350, it is determined whether or not the average value E calculated in step S340 is smaller than Ebest. When the deviation calculation process is executed for a certain ridgeline for the first time, the value calculated by the above equation (5) is substituted into Ebest. If it is determined in step S350 that the average value E is smaller than Ebest, the colorimetric coordinate sequence X is recorded as Xbest and the average value E is recorded as Ebest in step S360.
[0110]
If it is not determined in step S350 that the average value E is smaller than Ebest, step S360 is skipped. In step S370, it is determined whether or not the processing in steps S300 to S360 has been completed for all xk satisfying the conditional expression (4), and the total of xk satisfying conditional expression (4) is determined in step S370. If it is not determined that the process has been completed, the coordinate position of xk is changed in step S380, and the processes in and after step S300 are repeated. If it is determined in step S370 that the processing has been completed for all xk satisfying conditional expression (4), the processing returns to the processing shown in FIG.
[0111]
That is, as a result of the determination in step S370, when the coordinate position of xk is changed, the kth coordinate position of the colorimetric coordinate sequence X is changed, and the kth coordinate is moved within the movable range to calculate the sequential deviation. Is done. Further, since the deviation calculation process shown in FIG. 17 is a part of the loop of steps S125 to S137 shown in FIG. 15, this process is executed for all of the first to n-2th colorimetric coordinates. As described above, steps S120 and S380 correspond to the colorimetric coordinate provisional determination step, steps S300 and 310 correspond to the trial correspondence definition step, and steps S350, S360, and S140 correspond to the colorimetric coordinates. It corresponds to a specific process.
[0112]
(6) Color conversion processing based on the LUT created using the colorimetric coordinates:
When the coordinates for color measurement on the ridge line are specified as described above and the coordinates for color measurement over the entire area of the CMY color space are determined, the CMY coordinates and the Lab coordinates are increased when the LUT is created as described above. Color conversion with accuracy is possible. Therefore, the color conversion accuracy of the created LUT becomes very high, and printing can be executed while performing color conversion with very high accuracy. The created LUT can be used in a printing process that is performed on a general-purpose computer. FIG. 18 is a block diagram illustrating a computer configuration example that uses the LUT during printing.
[0113]
In the figure, a computer 140 is a general-purpose personal computer, and a printer driver (PRTDRV) 210, an input device driver (DRV) 220, and a display driver (DRV) 230 are incorporated in the OS 200. The display DRV 230 is a driver that controls display of image data and the like on the display 180, and the input device DRV 220 receives code signals from the keyboard 310 and the mouse 320 input via the serial communication I / O 190a and receives a predetermined signal. A driver that accepts input operations.
[0114]
The APL 250 is an application program that can execute retouching of a color image, and the user can operate the input device for operation under the execution of the APL 250 and cause the printing apparatus 400 to print the color image. The LUT created by the present invention is referred to when such a color image is printed. The color image data 150 a of the color image created by the APL 250 is dot matrix data representing gradations of RGB color components, is data compliant with the sRGB standard, and is stored in the HDD 150.
[0115]
The PRTDRV 210 includes an image data acquisition module 210a, a color conversion module 210b, a halftone processing module 210c, and a print data generation module 210d in order to execute printing. The LUT 150b created using the colorimetric coordinates determined according to the present invention is stored in the HDD 150. When the user issues a print execution instruction when executing the APL 250, the image data 150a to be printed is acquired by the image data acquisition module 210a, and the image data acquisition module 210a activates the color conversion module 210b. The color conversion module 210b is a module for converting sRGB gradation values into CMYK gradation values.³64 from the reference points³In addition to generating reference points, any sRGB dot data is converted into CMYK dot data by performing an interpolation operation using these reference points.
[0116]
When the color conversion module 210b performs color conversion to generate CMYK gradation data, the CMYK gradation data is transferred to the halftone processing module 210c. The halftone processing module 210c is a module that performs a halftone process for converting the CMYK gradation value of each dot and expressing it with the recording density of the ink droplets, and is a head for attaching ink at the converted recording density. Generate drive data. The print data generation module 210d receives the head drive data and performs rasterization that rearranges them in the order used by the printing apparatus 400. After the rasterization, print data is generated by adding predetermined information such as the resolution of the image, and is output to the printing apparatus 400 via the parallel communication I / O 190b. The printing apparatus 400 prints the image displayed on the display 180 based on the print data.
[0117]
In this printing process, the color conversion is performed with reference to the LUT created using the colorimetric coordinates determined by the present invention, so that the display 180 and the printing apparatus 400 can be accurately processed over the entire color gamut. Color conversion can be performed, and high-quality printing without tone jump can be performed. The above description is a very general-purpose printing process by the PRTDRV 210. Therefore, if the colorimetric coordinates are determined by the colorimetric coordinate determination method according to the present invention and an LUT is created, many LUTs used in the conventional printing process can be replaced with the LUT according to the present invention. It is possible to execute printing with very high image quality without changing the hardware configuration at all in the printing apparatus.
[0118]
As described above, in the present invention, the reference correspondence relationship is defined by a larger number of reference coordinates than the color measurement coordinates in the first color space, the trial correspondence relationship is defined by the color measurement coordinates, and the reference coordinates are set for specific coordinates. Coordinates in the second color space are calculated from the correspondence relationship and the trial correspondence relationship. Then, the trial is repeated so as to minimize the shift of the coordinates in each second color space. Therefore, it is possible to define the color correspondence relationship with high accuracy over the entire color space using the necessary and sufficient number of colorimetric coordinates.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram schematically illustrating steps of a colorimetric coordinate determination method.
FIG. 2 is a block diagram illustrating a relationship between steps of a colorimetric coordinate determination method.
FIG. 3 is a block diagram illustrating a configuration example of a colorimetric coordinate determination apparatus.
FIG. 4 is an explanatory diagram illustrating an example of a patch P printed by colorimetric coordinates.
FIG. 5 is a diagram showing a cube formed in a CMY color space.
FIG. 6 is a diagram illustrating a solid formed in the Lab color space.
FIG. 7 is an explanatory diagram for explaining determination of a spline interpolation function.
FIG. 8 is an explanatory diagram for explaining determination of a spline interpolation function.
FIG. 9 is an explanatory diagram for explaining determination of a spline interpolation function.
FIG. 10 is an explanatory diagram for explaining color conversion accuracy by a spline interpolation function.
FIG. 11 is an explanatory diagram illustrating distance limitation.
FIG. 12 is an explanatory diagram of an adjacency matrix.
FIG. 13 is an explanatory diagram for explaining a state in which distance limitation determination is performed.
FIG. 14 is an explanatory diagram illustrating a state in which colorimetric coordinates in a CMY color space are determined from colorimetric coordinates on a ridge line.
FIG. 15 is a general flow showing the entire color measurement coordinate determination operation.
FIG. 16 is a flowchart of a distance restriction determination process.
FIG. 17 is a flowchart of a deviation calculation process.
FIG. 18 is a block diagram illustrating a computer configuration example using an LUT during printing.
[Explanation of symbols]
10 ... Computer
11 ... Display
12 ... Printer
20 ... Colorimeter
30: Standard coordinate colorimetric value
31 ... Standard correspondence definition part
31a ... Spline interpolation calculation module
32 ... Trial part
32a ... Trial module
33 ... Temporary color coordinate determination unit
33a ... Color provisional coordinate provisional determination module
34 ... Corresponding relationship defining part for trial
35. Deviation calculation unit
35a: deviation calculation module
36 ... Color measurement coordinate specifying part
36a ... Color measurement coordinate determination module
37 ... Initial coordinates
38 ... Adjacency matrix
38a ... Adjacency matrix data
39a ... LUT creation module
39c ... Coordinate colorimetric value for colorimetry

Claims (3)

By obtaining the coordinate of the second color space colorimetry output color at a plurality of coordinates of the first color space used for a given image device a correspondence relationship between the first color space and the second color space In prescribing, in the color measurement coordinate determination method for determining the plurality of coordinate positions as color measurement coordinates,
With the coordinates of the first color space to get the coordinates of the second color space, and the reference relationship defining step of defining a correspondence relationship between the first color space and the second color space,
Determining the coordinates at a coarse interval than the coordinate of the first color space in the standard correspondence defining process, based on the determined coordinates and the reference relationship of correspondence between the first color space and the second color space A process for defining the correspondence relationship for trial,
A coordinate adjacent among the determined coordinate given limit on the distance coordinate mutual of the second color space defined by the reference relationship, the reference relationship for the particular coordinates in the first color space a calculation step of calculating a deviation between the coordinates in the second color space defined by the coordinate and the trial correspondence between the second color space defined by,
A step in which the displacement is determined as the color for the coordinate measuring coordinates when minimized,
Comprising
In the calculation step, in the second color space defined by the reference correspondence relationship, all coordinates that can exist on a line connecting the reference coordinates defined by fixing the predetermined color while making the predetermined color variable. The maximum distance between the coordinates in the second color space corresponding to the adjacent coordinates by a value obtained by dividing the reference length obtained by integrating the distances between the coordinates by (number of coordinates for colorimetry-1) and a predetermined constant. And
A coordinate determination method for colorimetry characterized by the above . The first color space is a color space including predetermined three colors used in an imaging device as a component, and the second color space is a uniform color space.
The color measurement coordinate determination method according to claim 1. The predetermined constant is 2 or more and 3 or less,
  The colorimetric coordinate determination method according to claim 1 or 2.

Images