JP3620360B2

JP3620360B2 - Imaging device

Info

Publication number: JP3620360B2
Application number: JP22823399A
Authority: JP
Inventors: 勝野田; 明仁西沢
Original assignee: Hitachi Ltd
Current assignee: Hitachi Ltd
Priority date: 1999-08-12
Filing date: 1999-08-12
Publication date: 2005-02-16
Anticipated expiration: 2019-08-12
Also published as: JP2001054133A

Description

【０００１】
【発明の属する技術分野】
本発明は撮像装置及び撮像装置用信号処理回路に関し、例えば固体撮像素子を１個乃至２個用いた動画カメラ又は静止画カメラに適用して高画質の画像を得るのに好適なものである。
【０００２】
【従来の技術】
単一の撮像素子を用いたカラーカメラでは、撮像素子の感光画素は複数の種類の色フィルタで塗り分けたものを規則的に配列して成り、該色フィルタに対応して該撮像素子から得られる画素信号を元に輝度信号と色信号を得ているのが一般的である。そのようなカラーカメラの信号処理方法の一例は、特開平１−３９８９３号に記載されている。更に同公報では、上記のようにして生成した輝度信号ではその色信号成分比率が望ましい比率からはずれていることを問題として取り上げ、その改善策として、撮像素子から得られる画素信号から輝度信号とＲ，Ｇ，Ｂの色信号とを生成し、それぞれの信号をガンマ補正した後に、Ｒ，Ｇ，Ｂ信号から生成した色差信号の一部を輝度信号に混合することで輝度信号中の色信号成分比率を補正して画質を向上する方法を開示している。
【０００３】
又、特開昭５７−４１０９１号では、輝度信号と色信号に対してガンマ補正した時に特に彩度の高い色の再現性が悪化することを問題として取り上げ、その改善策として、ガンマ補正前に輝度信号の低域成分を緑色信号成分に置き換え、ガンマ補正後にＲ，Ｂ信号の一部を加算して輝度信号に戻すことで色再現を向上する方法を開示している。
【０００４】
又、テレビジョン学会誌第３６巻、１１号、１００３頁乃至１００９頁では、輝度信号と色信号に対してカメラ信号処理を施した場合の被写体照明光源の色温度変化に伴う色再現誤差を問題として取り上げ、その改善策として、ガンマ補正前に輝度信号の低域成分を緑色信号成分に置き換え、同じくガンマ補正前にＲ信号とＢ信号に対する白バランス利得調節をし、ガンマ補正後にＲ，Ｂ信号の一部を加算して輝度信号に戻すことで色温度が変化した時も安定して良好な色再現を得る方法を開示している。
【０００５】
上記した従来技術はいずれも、輝度信号に対する処理（Ｙ処理）を色信号に対する処理（ＲＧＢ処理又はＲＢ処理）と独立させて行ったときに輝度信号に発生する画質悪化要因を問題点としており、その改善方法を提案しているものである。その方法を一言でまとめれば、輝度信号とは別に処理した色信号を輝度信号に作用させていることである。
【０００６】
【発明が解決しようとする課題】
しかしながら、元々、輝度信号に対する処理（Ｙ処理）を色信号に対する処理（ＲＧＢ処理又はＲＢ処理）と独立させて行う狙いは、単一の撮像素子でカラーカメラを構成する場合、色信号に比べ輝度信号は良質の信号を得やすいから、これを単独で最も重要な輝度信号に利用することにある。この観点で上記の従来技術を見直すと、いずれの従来技術も比較的性質の良くない色信号を輝度信号に作用させるものであるから、色再現の改善などの画質向上の効果がある反面、目的外の画質項目で画質劣化を起こすという副作用を伴う恐れがある。色信号は複数の種類の色フィルタで塗り分けた画素から得られる信号であり、これは１種類毎の色フィルタの２次元配列によるサンプリングが粗いことに原因して生ずるいわゆる折り返し雑音を含んでいるので、特にこの折り返し雑音が輝度信号に混入する副作用が問題となる。
【０００７】
そこで本発明は、輝度信号に対する処理（Ｙ処理）を色信号に対する処理（ＲＧＢ処理）と独立させて行う撮像装置で、色再現及び輝度再現を向上でき、かつ、色信号が輝度信号に及ぼす副作用を低減できる方法を提供することを目的とする。
【０００８】
【課題を解決するための手段】
上記課題を解決するために、本発明は、導入された光信号のＲ，Ｇ，Ｂ３原色光の内の一つにそれぞれ感応する３種類の光電変換画素を配列してなる撮像素子と、該撮像素子から得られる画素信号から広帯域輝度信号（Ｙ_Ｗ）と低域赤色信号（Ｒ_Ｌ）と低域緑色信号（Ｇ_Ｌ）と低域青色信号（Ｂ_Ｌ）を生成する信号生成回路とを含む光電変換手段と、該光電変換手段から得られる広帯域輝度信号（Ｙ_Ｗ）に対して該低域赤色信号（Ｒ_Ｌ）と該低域青色信号（Ｂ_Ｌ）を作用させて、低域成分の一部又は全部が緑色信号成分に置換されたＧ置換輝度信号（Ｙ_Ｇ）を生成する第１の演算手段と、白バランスを取ることを目的として該光電変換手段から得られる低域赤色信号（Ｒ_Ｌ）と低域青色信号（Ｂ_Ｌ）のそれぞれに対して利得調節する白バランス利得調節手段と、該Ｇ置換輝度信号（Ｙ_Ｇ）と該白バランス利得調節手段を通過した低域赤色信号（Ｒ_Ｌ）と同じく該白バランス利得調節手段を通過した低域青色信号（Ｂ_Ｌ）と該光電変換手段から得られる低域緑色信号（Ｇ_Ｌ）の４種の信号のそれぞれに対してガンマ補正を行うガンマ補正手段と、該ガンマ補正手段を通過した前記４種の信号の内のＧ置換輝度信号（Ｙ_Ｇ）と低域赤色信号（Ｒ_Ｌ）と低域青色信号（Ｂ_Ｌ）から輝度信号を再生成する第２の演算手段と、該ガンマ補正手段を通過した該４種の信号の内の低域赤色信号（Ｒ_Ｌ）と低域緑色信号（Ｇ_Ｌ）と低域青色信号（Ｂ_Ｌ）を組み合わせて色差信号を生成する第３の演算手段とを具備し、該第１の演算手段において広帯域輝度信号（Ｙ_Ｗ）の低域成分の一部又は全部を緑色信号成分に置換する該置換率を０から１の間で選定し、該第２の演算手段は、前記第１の演算手段の逆演算式に基づいた演算によってＧ置換輝度信号（Ｙ_Ｇ）と低域赤色信号（Ｒ_Ｌ）と低域青色信号（Ｂ_Ｌ）から輝度信号を再生成する。ここに、第１演算式ｙ＝ｆ（ｘ）より導かれる第２演算式ｘ＝Ｇ（ｙ）を第１演算式に対する逆演算式として定義する。
【０００９】
上記の構成による信号処理を特に輝度信号処理に注目して見ると、
第１ステップとして、広帯域輝度信号（Ｙ_Ｗ）からＧ置換輝度信号（Ｙ_Ｇ）に変換し、
第２ステップとして、Ｇ置換輝度信号（Ｙ_Ｇ）に対してガンマ補正を施し、
第３ステップとして、ガンマ補正済みのＧ置換輝度信号（Ｙ_Ｇ）とガンマ補正済みの低域赤色信号（Ｒ_Ｌ）と低域青色信号（Ｂ_Ｌ）から輝度信号を再生成する。
【００１０】
このことは、低域成分については、Ｒ，Ｇ，Ｂの３原色信号に対してガンマ補正を施し、それらの信号によって輝度信号を再合成するというカラーテレビジョン方式の原則通りの処理が為されたことを意味している。そして色再現や輝度再現は低域成分によって支配されているので、これにより、色再現や輝度再現が改善されて画質が向上することが分かる。
【００１１】
また、色信号中の折り返し雑音が輝度信号に混入する副作用については、広帯域輝度信号（Ｙ_Ｗ）の低域成分を緑色信号成分に置換してＧ置換輝度信号（Ｙ_Ｇ）得るに当たって、その置換率を必ずしも１００％とせずに中位の値に選定することで、色再現や輝度再現の改善効果と副作用とのバランスを取って問題を解決することが出来る。折り返し雑音は、撮像素子の１種類毎の色フィルタの２次元配列によって決まる２次元サンプリング周波数の近傍に位置する被写体の２次元周波数成分がサンプリング周波数の１／２の周波数で折り返したようにして直流に近い低周波数に変換されて生ずるものであって、被写体のサンプリング週数近傍の高周波成分とサンプリング周波数とのゼロビートであるとしても解釈できる。被写体がゆっくりと動いた場合は相対的なサンプリング位相が変化するので、ゼロビートの位相が変化して、再生画像では特に目障りなものである。これに対して、１枚の静止画として撮影したときは、ゼロビートの位相が固定されて、動画の場合に比べて格段に目障りの程度が軽いものである。撮像装置の動作モードを切り換えて動画撮影と静止画撮影を選択できるものにおいては、この性質を利用すると、動画撮影モードでは該置換率を比較的低めに設定し、静止画撮影モードでは該置換率を比較的高めに設定することで、それぞれに最適な画質が得られる。なお、第２の演算手段において輝度信号を再合成する演算式は該置換率に関連づけされており、第１の演算手段での演算式の逆演算式に基づいた演算式と成っているので、該置換率が２種類以上に切り換えられた場合でも、それぞれの置換率にふさわしい輝度信号が得られる。
【００１２】
【発明の実施の形態】
以下、実施の形態により本発明を詳細に説明する。
【００１３】
図１は本発明の第１の実施形態を示す図である。同図において、１は光電変換手段、１０１と１０２は該光電変換手段１の構成要素である撮像素子と信号生成手段である。２は第１の演算手段、３は白バランス利得調節手段、４はガンマ補正手段、５は第２の演算手段、６は第３の演算手段である。なお、２乃至６の各手段はアナログ回路、ディジタル回路、ＣＰＵを使ったソフトウエア処理のいずれでも構築可能であることを最初に言及しておく。
【００１４】
被写体像は不図示のレンズによって撮像素子の受光面に結像される。撮像素子の一例は図２に示される色フィルタ配列を有するＣＣＤである。図示の色フィルタ配列はＲ，Ｇ，Ｂ３原色フィルタによるもので、網掛け表示した４画素を単位として繰り返し配列されている。撮像素子からの信号読み出しは、水平１行毎に１画素ずつ順次読み出されるものである。信号生成手段１０２は撮像素子から順次出力される画素信号から、アナログ回路処理又はディジタル回路処理又はＣＰＵを使ったソフトウエア処理によって、広帯域輝度信号（Ｙ_Ｗ）と低域色信号（Ｒ_Ｌ）（Ｇ_Ｌ）（Ｂ_Ｌ）を生成する。又信号生成のアルゴリズムには種々の方法があるが、本願発明の主題からはずれるので一例を述べるに留める。先ず、撮像素子から順次出力される画素信号を１行分の遅延手段に通した信号と通さない信号により画素配列に対応する２行の上下２画素の信号を同時化する。次に、上下２画素の信号を加算し、水平方向にローパスフィルタをかけて広帯域輝度信号（Ｙ_Ｗ）を得る。色信号は、上下２行のそれぞれの画素をサンプリングしてＲ，Ｇ，Ｂそれぞれの水平方向の時系列信号としたものの低域成分として得る。これが信号生成手段１０２の信号生成方法の一例である。
【００１５】
第１の演算手段２は、光電変換手段１から出力された広帯域輝度信号（Ｙ_Ｗ）と低域色信号（Ｒ_Ｌ）、（Ｂ_Ｌ）を入力信号とし、次式の演算によってＧ置換輝度信号（Ｙ_Ｇ）を生成する。
【００１６】
Ｙ_Ｇ＝（１−ｐ）Ｙ_Ｗ＋ｐ（Ｙ_Ｗ−０．３０Ｒ_Ｌ−０．１１Ｂ_Ｌ）／０．５９−−−（１）
ここで、広帯域輝度信号を低域と高域に分け、Ｙ_Ｗ＝Ｙ_Ｌ＋Ｙ_Ｈで表すと、上式は、

となる。式（２）の第１項と第２項は低域成分を、第３項は高域成分を表す。第２項はＹ_ＬとＲ_ＬとＢ_Ｌから合成された緑色信号成分を意味し、第１項と第２項によって、広帯域輝度信号Ｙ_Ｗの低域成分（Ｙ_Ｌ）が置換率ｐでもって緑色信号成分に置換された信号を表す。Ｇ置換輝度信号（Ｙ_Ｇ）はこれに第３項の高域輝度信号が加わったものである。
【００１７】
式（１）の演算を行う演算手段の具体例を図３（イ）に示す。本図は演算ブロック図で表されていてアナログ又はディジタルのハードウエア回路で実現したものに相当するが、これと等価な機能はＣＰＵを使ったソフトウエア処理によっても実現できることは周知である。同図において、２０１は緑色信号成分生成部、２０２、２０３は置換率の係数部である。
【００１８】
図１の説明に戻って、白バランス利得調節手段３は、Ｒ_Ｌ，Ｂ_Ｌ信号に対して利得調節をして、被写体照明光源の色温度の違いに対して信号レベルのバランスを取る。ガンマ補正手段４は、Ｇ置換輝度信号（Ｙ_Ｇ）と白バランス利得調節手段を通過した低域赤色信号（Ｒ_Ｌ）と同じく白バランス利得調節手段を通過した低域青色信号（Ｂ_Ｌ）と光電変換手段１から得られる低域緑色信号（Ｇ_Ｌ）の４種の信号のそれぞれに対してガンマ補正を行う。第３の演算手段６は、ガンマ補正された低域赤色信号（Ｒ_Ｌ）と低域緑色信号（Ｇ_Ｌ）と低域青色信号（Ｂ_Ｌ）を入力信号とし、周知のマトリクス演算により赤色色差信号（Ｃｒ）と青色色差信号（Ｃｂ）を生成する。
【００１９】
第２の演算手段５は、ガンマ補正手段を通過したＧ置換輝度信号（Ｙ_Ｇ）と同じくガンマ補正手段を通過した低域赤色信号（Ｒ_Ｌ）と低域青色信号（Ｂ_Ｌ）を入力信号とし、次式の演算によって輝度信号（Ｙ）を再合成する。
【００２０】
Ｙ＝（１−ｑ）Ｙ_Ｇ＋ｑ（０．５９Ｙ_Ｇ＋０．３０Ｒ_Ｌ＋０．１１Ｂ_Ｌ）−−−−−（３）
前記式（２）と同様に、Ｇ置換輝度信号を低域と高域に分け、Ｙ_Ｇ＝Ｙ_ＧＬ＋Ｙ_ＧＨで表すと、上式は、

となる。式（４）の第１項と第２項は低域成分を、第３項は高域成分を表す。第２項はＹ_ＧＬとＲ_ＬとＢ_Ｌから再合成された低域輝度信号成分を意味し、第１項と第２項によって、Ｇ置換輝度信号（Ｙ_Ｇ）の低域成分（Ｙ_ＧＬ）が置換率ｑでもって再合成された低域輝度信号成分に置換された信号を表す。再合成輝度信号（Ｙ）はこれに第３項の高域信号が加わったものである。
【００２１】
ここに、置換率ｑは前記第１の演算手段の置換率ｐと関連づけされており、次式で定義される。
【００２２】
ｑ＝ｐ／（０．５９＋０．４１ｐ）−−−−−−−−−−−−−−−−−（５）
又は、
ｐ＝０．５９ｑ／（１−０．４１ｑ）−−−−−−−−−−−−−−−（５’）
図５に式（５）で定義されるｐとｑの関係を示す。
【００２３】
この時、式（１）を方程式と見なしてこの方程式をＹ_Ｗについて解き、Ｙ_ＷをＹに置き替えると式（３）に一致し、また逆に、式（３）を方程式と見なしてこの方程式をＹ_Ｇについて解き、ＹをＹ_Ｗに置き替えると式（１）に一致することが分かる。即ち、式（１）と式（３）は互いに逆演算式の関係にある。言い換えれば、本発明における第２の演算手段は第１の演算手段の逆演算式に基づいた演算を行うと言うことである。
【００２４】
式（３）の演算を行う演算手段の具体例を図３（ロ）に示す。同図において、５０１は輝度信号成分生成部、５０２、５０３は置換率の係数部である。
【００２５】
ところで、式（１）及び式（３）はそれぞれ、数学的に等価な次式に変形することが出来る。
【００２６】
Ｙ_Ｇ＝（Ｙ_Ｗ−ｑ（０．３０Ｒ_Ｌ＋０．１１Ｂ_Ｌ））／（１−０．４１ｑ）−−−（１’）
Ｙ＝（１−０．４１ｑ）Ｙ_Ｇ＋ｑ（０．３０Ｒ_Ｌ＋０．１１Ｂ_Ｌ）−−−−−−（３’）
式（１’）と式（３’）の演算を行う演算手段の具体例をそれぞれ図４（ハ）と（ニ）に示す。（イ）と（ハ）及び（ロ）と（ニ）を見比べると、数学的には等価な処理が異なった処理ブロックで構成されている。これに相当するハードウエア回路又はソフトウエアプログラムを作る上で都合の良い方を選択することが出来る。又特に、（ハ）の構成では、置換率ｐが用いられず、代わって第２の演算手段と共通の置換率ｑが用いられている。置換率ｐの値を直接のパラメータにしなくても良いケースでは、ｑを直接のパラメータとすることで、ｐをｑに変換する式（５）又は式（５’）の演算を省略できる利点が得られる。
【００２７】
しかし、いずれにしろ、第１の演算手段あるいは第２の演算手段の機能としては全く等価なものであるので、それぞれの演算手段の機能を表現する上では置換率ｐと置換率ｑを用いることにする。
【００２８】
次に本発明の第２の実施形態を、図６を用いて説明する。同図において、７は動画又は静止画を記録する記録手段、８は撮像装置の動作モードを制御する動作モード制御手段、８１は該動作モード制御手段が光電変換手段１の動作を制御する制御線、８２は該動作モード制御手段が該第１の演算手段２の該置換率ｐを制御する制御線、８５は該動作モード制御手段が該第２の演算手段５の該置換率ｑを制御する制御線、８７は該動作モード制御手段が該記録手段７の動作モードを制御する制御線である。その他の記号は前出の第１の実施形態と同じである。
【００２９】
本実施形態は、撮像装置の動作モードと連動して該第１の演算手段の該置換率ｐと該第２の演算手段の該置換率ｑを選択することが特徴であり、これによって撮像装置の動作モードに適した画質を得ることを狙うものである。撮像装置の動作モードと該置換率ｐの選定の組み合わせ例を表１に示す。
【００３０】
【表１】

【００３１】
なお、ｑは前記式（５）によってｐに従属して決まる。
【００３２】
同表において、区分Ａは、標準的な動画撮影モードで色再現向上効果を控え目に狙って置換率ｐを０．３とやや低い値に選定したケースである。区分Ｂは、動画撮影の途中で指定の画面を静止画記録するモードで、比較的良好な色再現を狙って置換率ｐを０．５と中位の値に選定したケースである。区分Ｃは、静止画撮影モードで出来るだけ色の美しい静止画が得られるように置換率ｐを１又はそれに近い高い値に選定したケースである。区分Ｄは、文書やパネルの文字を明瞭に撮影するための文字モードとも言うべき静止画モードで、色再現向上を狙わず置換率ｐを最小値の０に選定したケースである。
【００３３】
このように本実施形態によれば、撮像装置の動作モードに応じて置換率ｐの値が適値に選定されるので、色再現や輝度再現向上の効果と折り返し雑音の混入と言う副作用とのバランスが取れた総合的に良好な画像が得られる。
【００３４】
また、該置換率だけを変えた幾通りかの組み合わせを設けておくことにより、撮影者の意図によって置換率を選定することも可能となる。
【００３５】
【発明の効果】
以上説明したように、本発明は、光電変換手段から得られる広帯域輝度信号（Ｙ_Ｗ）に対して低域赤色信号（Ｒ_Ｌ）と低域青色信号（Ｂ_Ｌ）を作用させて、低域成分の一部又は全部が緑色信号成分に置換されたＧ置換輝度信号（Ｙ_Ｇ）を生成する第１の演算手段と、ガンマ補正手段を通過したＧ置換輝度信号（Ｙ_Ｇ）と低域赤色信号（Ｒ_Ｌ）と低域青色信号（Ｂ_Ｌ）から輝度信号を再合成する第２の演算手段と、ガンマ補正手段を通過した低域赤色信号（Ｒ_Ｌ）と低域緑色信号（Ｇ_Ｌ）と低域青色信号（Ｂ_Ｌ）を組み合わせて色差信号を生成する第３の演算手段とを具備し、該第１の演算手段において広帯域輝度信号（Ｙ_Ｗ）の低域成分の一部又は全部を緑色信号成分に置換する該置換率を０から１の間で選定し、該第２の演算手段は、前記第１の演算手段の逆演算式に基づいた演算によってＧ置換輝度信号（Ｙ_Ｇ）と低域赤色信号（Ｒ_Ｌ）と低域青色信号（Ｂ_Ｌ）から輝度信号を再合成するので、低域成分については、Ｒ，Ｇ，Ｂの３原色信号に対してガンマ補正を施し、それらの信号によって輝度信号を再合成するというカラーテレビジョン方式の原則通りの処理が為されたと等価な作用が得られ、これにより、色再現や輝度再現が改善されて画質が向上する効果が得られる。一方、色信号中の折り返し雑音が輝度信号に混入する副作用については、広帯域輝度信号（Ｙ_Ｗ）の低域成分を緑色信号成分に置換してＧ置換輝度信号（Ｙ_Ｇ）得るに当たって、その置換率を必ずしも１００％とせずに中位の値に選定することで、色再現や輝度再現の改善効果と副作用とのバランスを取って問題を解決することが出来る。更に、撮像装置の動作モードに応じて例えば動画撮影モードでは該置換率を比較的低めに設定し、静止画撮影モードでは該置換率を比較的高めに設定することで、それぞれに最適な画質が得られる。
【図面の簡単な説明】
【図１】本発明の第１の実施形態を示す図。
【図２】撮像素子の色フィルタ配列の一例を示す図。
【図３】第１の演算手段及び第２の演算手段の具体例を示す図。
【図４】第１の演算手段及び第２の演算手段の具体例を示す図。
【図５】ｐとｑの関係を示す図。
【図６】本発明の第２の実施形態を示す図。
【符号の説明】
１… 光電変換手段
１０１… 撮像素子
１０２… 信号生成手段
２… 第１の演算手段
２０１… 緑色信号成分生成部
２０２… 置換率の係数部
２０３… 置換率の係数部
３… 白バランス利得調節手段
４… ガンマ補正手段
５… 第２の演算手段
５０１… 輝度信号成分生成部
５０２… 置換率の係数部
５０３… 置換率の係数部
６… 第３の演算手段
７… 記録手段
８… 動作モード制御手段
８１… 制御線
８２… 制御線
８５… 制御線
８７… 制御線[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image pickup apparatus and a signal processing circuit for the image pickup apparatus, and is suitable for obtaining a high-quality image by applying it to, for example, a moving image camera or a still image camera using one or two solid-state image sensors.
[0002]
[Prior art]
In a color camera using a single image sensor, the photosensitive pixels of the image sensor are formed by regularly arranging a plurality of types of color filters, and are obtained from the image sensor corresponding to the color filters. In general, a luminance signal and a color signal are obtained based on a pixel signal obtained. An example of such a signal processing method for a color camera is described in JP-A-1-39893. Furthermore, in this publication, the luminance signal generated as described above is taken up as a problem that the color signal component ratio deviates from a desirable ratio, and as a measure for improvement, the luminance signal and the R signal are obtained from the pixel signal obtained from the image sensor. , G, B color signals are generated, and each signal is gamma-corrected, and then a part of the color difference signal generated from the R, G, B signals is mixed with the luminance signal, whereby the color signal component in the luminance signal is obtained. A method for improving the image quality by correcting the ratio is disclosed.
[0003]
Japanese Patent Application Laid-Open No. 57-41091 takes up the problem that the reproducibility of a highly saturated color deteriorates when the luminance signal and the color signal are gamma-corrected. There is disclosed a method for improving color reproduction by replacing a low-frequency component of a luminance signal with a green signal component, adding a part of R and B signals after gamma correction, and returning to a luminance signal.
[0004]
Also, in the Journal of the Institute of Television Engineers of Japan, Vol. 36, No. 11, pages 1003 to 1009, there is a problem with color reproduction errors caused by changes in the color temperature of the subject illumination light source when camera signal processing is performed on luminance signals and color signals. As a measure for improvement, the low-frequency component of the luminance signal is replaced with a green signal component before gamma correction, the white balance gain is adjusted for the R signal and B signal before gamma correction, and the R and B signals are corrected after gamma correction. A method is disclosed in which a good color reproduction is stably obtained even when the color temperature is changed by adding a part of the color signal back to the luminance signal.
[0005]
Each of the above conventional techniques has a problem of image quality deterioration that occurs in the luminance signal when the processing on the luminance signal (Y processing) is performed independently of the processing on the color signal (RGB processing or RB processing). The improvement method is proposed. To sum up the method, a color signal processed separately from the luminance signal is applied to the luminance signal.
[0006]
[Problems to be solved by the invention]
However, originally, the purpose of performing the processing on the luminance signal (Y processing) separately from the processing on the color signal (RGB processing or RB processing) is that when a color camera is configured with a single image sensor, the luminance is higher than that of the color signal. Since it is easy to obtain a high-quality signal, the signal is used alone for the most important luminance signal. Reviewing the above prior art from this point of view, any of the prior arts causes a relatively poor color signal to act on the luminance signal, so there is an effect of improving the image quality such as improved color reproduction. There may be a side effect of causing image quality degradation in other image quality items. The color signal is a signal obtained from a pixel that is painted with a plurality of types of color filters, and this includes so-called aliasing noise caused by rough sampling by a two-dimensional array of color filters for each type. Therefore, a side effect that the aliasing noise is mixed into the luminance signal becomes a problem.
[0007]
Therefore, the present invention is an imaging device that performs processing on a luminance signal (Y processing) independently of processing on a color signal (RGB processing), and can improve color reproduction and luminance reproduction, and has side effects that the color signal has on the luminance signal. It is an object of the present invention to provide a method capable of reducing the above.
[0008]
[Means for Solving the Problems]
In order to solve the above-described problems, the present invention provides an imaging device in which three types of photoelectric conversion pixels each sensitive to one of R, G, and B3 primary color lights of an introduced optical signal are arranged; A signal generation circuit that generates a broadband luminance signal (Y _W ), a low-frequency red signal (R _L ), a low-frequency green signal (G _L ), and a low-frequency blue signal (B _L ) from the pixel signal obtained from the image sensor. And a low-frequency component by causing the low-frequency red signal (R _L ) and the low-frequency blue signal (B _L ) to act on the broadband luminance signal (Y _W ) obtained from the photoelectric conversion device. First arithmetic means for generating a G-replaced luminance signal (Y _G ) in which part or all of the signal is replaced with a green signal component, and a low-frequency red signal obtained from the photoelectric conversion means for the purpose of white balance _{(R L)} and the gain adjustment for each of the low blue signal _{(B L)} And white balance gain control means for, the G substituted luminance signal (Y _G) and the low-band red signal has passed through the white balance gain control means (R _L) and also the low-frequency blue signal which has passed through the said white balance gain control means (B _L ) and a gamma correction unit that performs gamma correction on each of the four types of signals of the low-frequency green signal (G _L ) obtained from the photoelectric conversion unit, and the four types of the four types of signals that have passed through the gamma correction unit A second arithmetic unit that regenerates a luminance signal from the G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ), and the low-frequency blue signal (B _L ), and the gamma correction unit. A third calculation means for generating a color difference signal by combining the low-frequency red signal (R _L ), the low-frequency green signal (G _L ), and the low-frequency blue signal (B _L ) of the four types of signals. comprising, wideband luminance signal in said first operation means (Y _W) The replacement rate for replacing part or all of the low-frequency component with the green signal component is selected from 0 to 1, and the second calculation means calculates based on the inverse calculation formula of the first calculation means To regenerate the luminance signal from the G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ), and the low-frequency blue signal (B _L ). Here, the second arithmetic expression x = G (y) derived from the first arithmetic expression y = f (x) is defined as an inverse arithmetic expression with respect to the first arithmetic expression.
[0009]
Looking at the signal processing with the above configuration, particularly focusing on luminance signal processing,
As a first step, a broadband luminance signal (Y _W ) is converted to a G replacement luminance signal (Y _G ),
As the second step, gamma correction is performed on the G replacement luminance signal (Y _G ),
As a third step, a luminance signal is regenerated from the g-corrected G replacement luminance signal (Y _G ), the gamma corrected low-frequency red signal (R _L ), and the low-frequency blue signal (B _L ).
[0010]
This means that the low-frequency component is processed according to the principle of the color television system in which gamma correction is performed on the three primary color signals of R, G, and B, and the luminance signal is recombined with these signals. It means that. Since color reproduction and luminance reproduction are dominated by low-frequency components, it can be seen that this improves color reproduction and luminance reproduction and improves image quality.
[0011]
As for the side effect of aliasing noise in the color signal mixed into the luminance signal, the low-frequency component of the wide-band luminance signal (Y _W ) is replaced with the green signal component to obtain the G-substituted luminance signal (Y _G ). By selecting a medium value without necessarily setting the rate to 100%, the problem can be solved by balancing the improvement effect of color reproduction and luminance reproduction with side effects. The aliasing noise is generated by directing the two-dimensional frequency component of the subject located in the vicinity of the two-dimensional sampling frequency determined by the two-dimensional arrangement of the color filters for each type of the image sensor at a frequency half that of the sampling frequency. It can be interpreted as a zero beat between a sampling frequency and a high-frequency component in the vicinity of the sampling week of the subject. Since the relative sampling phase changes when the subject moves slowly, the zero beat phase changes, which is particularly disturbing in the reproduced image. On the other hand, when shooting as a single still image, the phase of zero beat is fixed, and the degree of obstruction is much lighter than that of moving images. For those that can select between moving image shooting and still image shooting by switching the operation mode of the imaging device, using this property, the replacement rate is set relatively low in the moving image shooting mode and the replacement rate in the still image shooting mode. By setting relatively high, the optimum image quality can be obtained for each. The arithmetic expression for recombining the luminance signal in the second arithmetic means is associated with the replacement rate, and is an arithmetic expression based on the inverse arithmetic expression of the arithmetic expression in the first arithmetic means. Even when the replacement rate is switched to two or more types, a luminance signal suitable for each replacement rate can be obtained.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, the present invention will be described in detail by embodiments.
[0013]
FIG. 1 is a diagram showing a first embodiment of the present invention. In the figure, reference numeral 1 denotes a photoelectric conversion means, and 101 and 102 denote an image sensor and a signal generation means which are components of the photoelectric conversion means 1. 2 is a first computing means, 3 is a white balance gain adjusting means, 4 is a gamma correction means, 5 is a second computing means, and 6 is a third computing means. It should be first mentioned that each of the means 2 to 6 can be constructed by any of an analog circuit, a digital circuit, and a software process using a CPU.
[0014]
The subject image is formed on the light receiving surface of the image sensor by a lens (not shown). An example of the image sensor is a CCD having the color filter array shown in FIG. The illustrated color filter array is based on R, G, and B3 primary color filters, and is repeatedly arrayed in units of four pixels that are shaded. Signal readout from the image sensor is sequentially read out pixel by pixel for each horizontal row. The signal generation unit 102 generates a wide-band luminance signal (Y _W ) and a low-frequency color signal (R _L ) (from a pixel signal sequentially output from the image sensor by analog circuit processing, digital circuit processing, or software processing using a CPU. G _L ) (B _L ) is generated. There are various methods for generating signals, but these are not the subject of the present invention, so only an example will be described. First, the signals of the upper and lower two pixels in two rows corresponding to the pixel array are synchronized by a signal that does not pass the pixel signals sequentially output from the image sensor through the signal that has passed through the delay means for one row. Next, the signals of the upper and lower two pixels are added and a low-pass filter is applied in the horizontal direction to obtain a broadband luminance signal (Y _W ). The color signal is obtained as a low-frequency component obtained by sampling the pixels in the upper and lower two rows to obtain the R, G, B horizontal time-series signals. This is an example of a signal generation method of the signal generation means 102.
[0015]
The first calculation means 2 uses the broadband luminance signal (Y _W ) and the low-frequency color signals (R _L ) and (B _L ) output from the photoelectric conversion means 1 as input signals, and performs G-substitution luminance by calculating the following equation: A signal (Y _G ) is generated.
[0016]
_{_{_{_{Y G = (1-p)}}}} Y W + p (Y W -0.30R L -0.11B L) /0.59 --- (1)
Here, when the broadband luminance signal is divided into a low frequency region and a high frequency region and expressed as Y _W = Y _L + Y _H , the above equation is

It becomes. In the expression (2), the first and second terms represent a low frequency component, and the third term represents a high frequency component. The second term denotes a green signal component synthesized from Y _L and R _L and B _L, the first and second terms, the low-frequency component (Y _L) of the wideband luminance signal Y _W with the substitution rate p Thus, the signal replaced with the green signal component is represented. The G replacement luminance signal (Y _G ) is obtained by adding the high-frequency luminance signal of the third term to this.
[0017]
A specific example of the calculation means for performing the calculation of Expression (1) is shown in FIG. This figure is represented by an arithmetic block diagram and corresponds to that realized by an analog or digital hardware circuit, but it is well known that a function equivalent to this can also be realized by software processing using a CPU. In the figure, 201 is a green signal component generation unit, and 202 and 203 are coefficient units of replacement rates.
[0018]
Returning to the description of FIG. 1, the white balance gain adjusting means 3 adjusts the gain for the R _L and B _L signals to balance the signal level against the difference in color temperature of the subject illumination light source. The gamma correction means 4 includes a G replacement luminance signal (Y _G ), a low-frequency red signal (R _L ) that has passed through the white balance gain adjustment means, and a low-frequency blue signal (B _L ) that has passed through the white balance gain adjustment means. Gamma correction is performed on each of the four types of signals of the low-frequency green signal (G _L ) obtained from the photoelectric conversion means 1. The third calculation means 6 receives the low-frequency red signal (R _L ), the low-frequency green signal (G _L ), and the low-frequency blue signal (B _L ) subjected to gamma correction as input signals, and performs a red color difference by a known matrix calculation. A signal (Cr) and a blue color difference signal (Cb) are generated.
[0019]
The second calculation means 5 receives the low-frequency red signal (R _L ) and low-frequency blue signal (B _L ) that have passed through the gamma correction means in the same manner as the G replacement luminance signal (Y _G ) that has passed through the gamma correction means. Then, the luminance signal (Y) is re-synthesized by the calculation of the following equation.
[0020]
_{_{Y = (1-q) Y}} G + q (0.59Y G + 0.30R L + 0.11B L) ----- (3)
Similarly to the equation (2), when the G replacement luminance signal is divided into a low frequency region and a high frequency region and expressed as Y _G = Y _GL + Y _GH ,

It becomes. In the expression (4), the first and second terms represent a low frequency component, and the third term represents a high frequency component. The second term refers to a low-frequency luminance signal component is re-synthesized from _{Y GL} and _{R L} and _{B L,} the first and second terms, the low-frequency component _{(Y GL} of G substitution luminance signal _{(Y G)} ) Represents the signal replaced with the low-frequency luminance signal component recombined with the replacement rate q. The recombined luminance signal (Y) is obtained by adding the high-frequency signal of the third term to this.
[0021]
Here, the replacement rate q is related to the replacement rate p of the first computing means, and is defined by the following equation.
[0022]
q = p / (0.59 + 0.41p) ---------------- (5)
Or
p = 0.59q / (1-0.41q) -------------- (5 ')
FIG. 5 shows the relationship between p and q defined by equation (5).
[0023]
At this time, solve for the equation Y _W considers Equation (1) and equation, the Y _W match changing the formula (3) placed in Y, and conversely, regards equation (3) and equation this It can be seen that when the equation is solved for Y _G and Y is replaced with Y _W , the equation (1) is satisfied. In other words, the expressions (1) and (3) are in inverse relation to each other. In other words, the second calculation means in the present invention performs a calculation based on the inverse calculation formula of the first calculation means.
[0024]
A specific example of the calculation means for performing the calculation of Expression (3) is shown in FIG. In the figure, reference numeral 501 denotes a luminance signal component generation unit, and

reference numerals

502 and 503 denote replacement rate coefficient units.
[0025]
By the way, the equations (1) and (3) can be transformed into the following mathematically equivalent equations.
[0026]
_{_{_{_{Y G = (Y W -q (}}}} 0.30R L + 0.11B L)) / (1-0.41q) --- (1 ')
Y = (1−0.41q) Y _G + q (0.30R _L + 0.11B _L ) −−−−−− (3 ′)
Specific examples of the calculation means for performing the calculations of the equations (1 ′) and (3 ′) are shown in FIGS. 4 (c) and (d), respectively. Comparing (a) and (c) and (b) and (d), mathematically equivalent processes are composed of different processing blocks. It is possible to select a hardware circuit or a software program corresponding to this which is convenient for creating. In particular, in the configuration of (c), the replacement rate p is not used, and instead, the replacement rate q common to the second calculation means is used. In the case where the value of the replacement rate p does not need to be a direct parameter, by using q as a direct parameter, there is an advantage that the calculation of Expression (5) or Expression (5 ′) for converting p to q can be omitted. can get.
[0027]
However, in any case, since the functions of the first calculation means or the second calculation means are completely equivalent, the replacement rate p and the replacement rate q are used to express the functions of the respective calculation means. To.
[0028]
Next, a second embodiment of the present invention will be described with reference to FIG. In the figure, 7 is a recording means for recording a moving image or a still image, 8 is an operation mode control means for controlling the operation mode of the imaging apparatus, and 81 is a control line for controlling the operation of the photoelectric conversion means 1 by the operation mode control means. , 82 is a control line for the operation mode control means to control the replacement rate p of the first calculation means 2, and 85 is a control line for the operation mode control means to control the replacement rate q of the second calculation means 5. A control line 87 is a control line for controlling the operation mode of the recording means 7 by the operation mode control means. Other symbols are the same as those in the first embodiment.
[0029]
The present embodiment is characterized in that the replacement rate p of the first calculation means and the replacement rate q of the second calculation means are selected in conjunction with the operation mode of the image pickup device. The aim is to obtain an image quality suitable for the operation mode. Table 1 shows a combination example of the selection of the operation mode of the imaging apparatus and the replacement rate p.
[0030]
[Table 1]

[0031]
Note that q is determined depending on p by the above equation (5).
[0032]
In the table, category A is a case where the replacement rate p is selected to be a slightly low value of 0.3, aiming conservatively at the color reproduction improvement effect in the standard moving image shooting mode. Category B is a mode in which a designated screen is recorded as a still image in the middle of moving image shooting, and the replacement rate p is selected to a medium value of 0.5 with the aim of relatively good color reproduction. Category C is a case where the replacement rate p is selected to be 1 or a high value close to it so that a still image having the color as beautiful as possible can be obtained in the still image shooting mode. Category D is a still image mode that should be referred to as a character mode for clearly photographing characters on a document or a panel, and the replacement rate p is selected to be the minimum value 0 without aiming at improving color reproduction.
[0033]
As described above, according to the present embodiment, since the value of the replacement rate p is selected to an appropriate value according to the operation mode of the imaging apparatus, the effect of improving color reproduction and luminance reproduction and the side effect of mixing aliasing noise are present. A balanced and comprehensively good image can be obtained.
[0034]
Also, by providing several combinations in which only the replacement rate is changed, it is possible to select the replacement rate according to the photographer's intention.
[0035]
【The invention's effect】
As described above, the present invention causes the low-frequency red signal (R _L ) and the low-frequency blue signal (B _L ) to act on the broadband luminance signal (Y _W ) obtained from the photoelectric conversion means, thereby reducing the low frequency. first calculation means some components or the whole to produce a substituted G substituted luminance signal (Y _G) green signal components, G substitution luminance signal passed through the gamma correction means (Y _G) and low red signal (R _L) and second calculating means for recombining the luminance signal from the low blue signal (B _L), the low-frequency red signal (R _L) and the low-frequency green signal that has passed through the gamma correction means (G _L ) And a low-frequency blue signal (B _L ) to generate a color-difference signal, and a part of the low-frequency component of the broadband luminance signal (Y _W ) or The replacement rate for replacing all the green signal components is selected from 0 to 1, and the second performance is selected. The arithmetic means regenerates the luminance signal from the G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ), and the low-frequency blue signal (B _L ) by an operation based on the inverse arithmetic expression of the first arithmetic means. As a result of the synthesis, the low-frequency component is processed in accordance with the principle of the color television system in which gamma correction is performed on the three primary color signals of R, G, and B, and the luminance signal is re-synthesized with these signals. Thus, an effect equivalent to the above is obtained, whereby the color reproduction and luminance reproduction are improved and the image quality is improved. On the other hand, with respect to the side effect of aliasing noise in the color signal being mixed into the luminance signal, the low-frequency component of the broadband luminance signal (Y _W ) is replaced with the green signal component to obtain the G-substituted luminance signal (Y _G ). By selecting a medium value without necessarily setting the rate to 100%, the problem can be solved by balancing the improvement effect of color reproduction and luminance reproduction with side effects. Furthermore, according to the operation mode of the imaging device, for example, the replacement rate is set to be relatively low in the moving image shooting mode, and the replacement rate is set to be relatively high in the still image shooting mode. can get.
[Brief description of the drawings]
FIG. 1 is a diagram showing a first embodiment of the present invention.
FIG. 2 is a diagram illustrating an example of a color filter array of an image sensor.
FIG. 3 is a diagram illustrating a specific example of a first calculation unit and a second calculation unit.
FIG. 4 is a diagram illustrating a specific example of a first calculation unit and a second calculation unit.
FIG. 5 is a diagram showing the relationship between p and q.
FIG. 6 is a diagram showing a second embodiment of the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Photoelectric conversion means 101 ... Image pick-up element 102 ... Signal generation means 2 ... First calculating means 201 ... Green signal component generation part 202 ... Substitution rate coefficient part 203 ... Substitution rate coefficient part 3 ... White balance gain adjustment means 4 Gamma correction unit 5 Second calculation unit 501 Luminance signal component generation unit 502 Replacement factor unit 503 Replacement factor unit 6 Third calculation unit 7 Recording unit 8 Operation mode control unit 81 ... Control line 82 ... Control line 85 ... Control line 87 ... Control line

Claims

An imaging device in which at least three types of photoelectric conversion pixels sensitive to the introduced optical signal are arrayed, and a broadband luminance signal (Y _W ) and a low-frequency red signal (R _L ) from the pixel signal obtained from the imaging device Photoelectric conversion means including a low-frequency green signal (G _L ) and a signal generation circuit that generates a low-frequency blue signal (B _L );
A low-frequency red signal (R _L ) and a low-frequency blue signal (B _L ) obtained from the photoelectric conversion means are caused to act on the broadband luminance signal (Y _W ) obtained from the photoelectric conversion means, so that the low-frequency component First calculating means for generating a G-replaced luminance signal (Y _G ) partially or entirely replaced with a green signal component;
White balance gain adjusting means for correcting white balance by adjusting the gain for each of the low-frequency red signal (R _L ) and the low-frequency blue signal (B _L ) obtained from the photoelectric conversion means; and the G replacement luminance signal (Y _G ) and the low-frequency blue signal (B _L ) that has passed through the white balance gain adjusting means as well as the low-frequency red signal (R _L ) that has passed through the white balance gain adjusting means and the low-frequency signal obtained from the photoelectric conversion means Gamma correction means for performing gamma correction on each of the four signals of the gamut green signal (G _L );
Secondly, a luminance signal is recombined from the G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ), and the low-frequency blue signal (B _L ) among the four types of signals that have passed through the gamma correction means. And a computing means of
A color difference signal is generated by combining a low-frequency red signal (R _L ), a low-frequency green signal (G _L ), and a low-frequency blue signal (B _L ) among the four types of signals that have passed through the gamma correction means. 3 computing means;
In an imaging apparatus comprising:
The replacement rate for replacing part or all of the low-frequency component of the broadband luminance signal (Y _W ) with the green signal component in the first calculation means is a value selected from 0 to 1.
The second calculating means calculates the G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ), and the low-frequency blue signal (B _L ) by calculation based on the inverse calculation formula of the first calculating means. An imaging device characterized by recombining luminance signals.

An imaging device in which at least three types of photoelectric conversion pixels sensitive to the introduced optical signal are arrayed, and a broadband luminance signal (Y _W ) and a low-frequency red signal (R _L ) from the pixel signal obtained from the imaging device Photoelectric conversion means including a low-frequency green signal (G _L ) and a signal generation circuit that generates a low-frequency blue signal (B _L );
A low-frequency red signal (R _L ) and a low-frequency blue signal (B _L ) obtained from the photoelectric conversion means are caused to act on the broadband luminance signal (Y _W ) obtained from the photoelectric conversion means, so that the low-frequency component First calculating means for generating a G-replaced luminance signal (Y _G ) partially or entirely replaced with a green signal component;
White balance gain adjusting means for adjusting the gain for each of the low-frequency red signal (R _L ) and the low-frequency blue signal (B _L ) obtained from the photoelectric conversion means for the purpose of white balance;
The G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ) that has passed through the white balance gain adjusting means, and the low-frequency blue signal (B _L ) that has passed through the white balance gain adjusting means and the photoelectric conversion Gamma correction means for performing gamma correction on each of the four types of signals of the low-frequency green signal (G _L ) obtained from the means;
A low-frequency red signal (R _L ) and a low-frequency blue signal (B _L ) that have also passed through the gamma correction means are caused to act on the G-replacement luminance signal (Y _G ) that has passed through the gamma correction means, thereby reducing the low-frequency range. Second computing means for generating a luminance signal in which a part or all of the components are replaced with a recombined luminance signal component;
A color difference signal is generated by combining a low-frequency red signal (R _L ), a low-frequency green signal (G _L ), and a low-frequency blue signal (B _L ) among the four types of signals that have passed through the gamma correction means. 3 computing means;
In an imaging apparatus comprising:
The replacement rate for replacing part or all of the low-frequency component of the broadband luminance signal (Y _W ) with the green signal component in the first calculation means is a first replacement rate p,
In the second computing means, a second substitution rate q for substituting a part or all of the low frequency component of the G replacement luminance signal (Y _G ) with the recombined luminance signal component is defined as the first substitution rate p. In connection,
An imaging apparatus characterized in that a value determined by q = p / (0.59 + 0.41 × p) or a value in the vicinity thereof is used.

3. The imaging apparatus according to claim 1, wherein two or more types of values are prepared for the replacement rate of the first calculation means, and one of the values can be selected and executed. 4. apparatus.

An imaging device in which at least three types of photoelectric conversion pixels sensitive to the introduced optical signal are arrayed, and a broadband luminance signal (Y _W ) and a low-frequency red signal (R _L ) from the pixel signal obtained from the imaging device Photoelectric conversion means including a low-frequency green signal (G _L ) and a signal generation circuit that generates a low-frequency blue signal (B _L );
A low-frequency red signal (R _L ) and a low-frequency blue signal (B _L ) obtained from the photoelectric conversion means are caused to act on the broadband luminance signal (Y _W ) obtained from the photoelectric conversion means, so that the low-frequency component First calculating means for generating a G-replaced luminance signal (Y _G ) partially or entirely replaced with a green signal component;
White balance gain adjusting means for correcting white balance by adjusting the gain for each of the low-frequency red signal (R _L ) and the low-frequency blue signal (B _L ) obtained from the photoelectric conversion means;
The G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ) that has passed through the white balance gain adjusting means, and the low-frequency blue signal (B _L ) that has passed through the white balance gain adjusting means and the photoelectric conversion Gamma correction means for performing gamma correction on each of the four types of signals of the low-frequency green signal (G _L ) obtained from the means;
Secondly, a luminance signal is recombined from the G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ), and the low-frequency blue signal (B _L ) among the four types of signals that have passed through the gamma correction means. And a computing means of
A color difference signal is generated by combining a low-frequency red signal (R _L ), a low-frequency green signal (G _L ), and a low-frequency blue signal (B _L ) among the four types of signals that have passed through the gamma correction means. 3 computing means;
In an imaging apparatus comprising means for controlling an operation mode of the imaging apparatus,
The replacement rate for replacing part or all of the low-frequency component of the wide-band luminance signal (Y _W ) with the green signal component in the first calculation means is an operation mode set by the means for controlling the operation mode of the imaging device. It is a value selected between 0 and 1 in conjunction,
The second calculating means calculates the G replacement luminance signal (Y _G ), the low-frequency red signal (R _L ), and the low-frequency blue signal (B _L ) by calculation based on the inverse calculation formula of the first calculating means. An imaging device characterized by recombining luminance signals.

5. The imaging apparatus according to claim 4, wherein the first computing means replaces part or all of the low-frequency component of the wide-band luminance signal (Y _W ) with a green signal component according to the operation mode of the imaging device. An image pickup apparatus characterized in that a lower value is selected when the operation mode set by the control means is a moving image shooting mode, and a higher value is selected when it is a still image shooting mode.