JP4434375B2 - Ranging device - Google Patents

Description

本実施形態では、以上の空間フィルタを、状況に応じて適宜選択して使用する。
【００６５】
なお、全画像について差分処理をする場合は、比較的演算が簡単で高速な１次微分オペレータ、ラプラシアンオペレータを使用する。一方、撮影画面内の一部の画像に関して差分処理を行う場合は、演算がやや複雑で演算時間が大きいが効果は大きいソーベルオペレータを選択して使用する。
【００６６】
また、低輝度でＡＦエリアセンサ１２の積分時間が長い場合は、１次微分オペレータまたはラプラシアンオペレータを使用し、一方高輝度で積分時間が小さい場合は、ソーベルオペレータを使用することによりＡＦタイムラグとしてのバランスをとってもよい。
【００６７】
（３）ステップＳ３０３：２値化処理（閾値処理）
この２値化処理を図１１に示すフローチャートを参照して説明する。
当該２値化処理においてマイクロコンピュータ１１はまず、画像内の各輝度を示す画素値の出現頻度を表わすヒストグラムを作成し（ステップＳ４０１）、次に閾値設定処理を行う（ステップＳ４０２）。ここで、ヒストグラムに基づいて閾値を決定する手法は種々知られているが、たとえばモード法では、上記のうちで頻度が最小の輝度値を閾値（スレッシュレベル）として、２値化処理を行なう（図１２参照）。
【００６８】
上記ステップＳ４０２において閾値が設定された後、マイクロコンピュータ１１は２値化を行う（ステップＳ４０３）。
【００６９】
なお、閾値設定の他の手法としては、取り出す図形の面積がある程度わかっている場合に有効なｐ−タイル法、図形の境界部分に閾値が設定されるように定める微分ヒストグラム法、濃度値の集合を２つのクラスに分けたときのクラス間の分離が最もよくなるようにパラメータｔを求める判別分析法、画像位置に応じて閾値を変化させる可変閾値法等の手法が知られている。
【００７０】
本実施形態では、これらの手法を状況に応じて適宜選択して使用する。たとえばヒストグラムの形状を判別して明確な最小値が存在するか否かを判定し、明確な場合はモード法を採用する。一方、不明確な場合は判別分析法を採用する。
【００７１】
このようにヒストグラムの形状判別を行い、その結果に応じて閾値設定方法を変更する。ヒストグラムの形状判別方法については、図１３に示すようにたとえば（谷）極値でありかつ頻度最小値ａ、２番目に小さい値ｂを求め、その差ｂ−ａを判別値ｄｔｈと比較して、所定値ｄｔｈより大きい場合、最小値ａの輝度値を閾値として採用する。一方、所定値以下の場合は、画像位置に応じて閾値を変化させる可変閾値法を採用する。
【００７２】
ここで、上記閾値設定処理を図１３及び図１４に示すフローチャートを参照して詳しく説明する。
マイクロコンピュータ１１は、図１３に示す如き最小値ａと２番目に小さい頻度ｂを求める（ステップＳ５０１）。次に、この差（ｂ−ａ）と所定の判定値ｄｔｈとを比較する（ステップＳ５０２）。そして、差（ｂ−ａ）が判定値ｄｔｈより大きい場合は、最小値ａに対応する輝度値Ｂａを閾値として採用する（ステップＳ５０４）。一方、差（ｂ−ａ）が判定値ｄｔｈ以下の場合は可変閾値法を採用する（ステップＳ５０５）。
【００７３】
撮影画面全体に対応する画像での２値化の場合は、最初にモード法により閾値を設定して２値化処理を行う。そして、２値化画像を評価した結果が良好ではない場合は画像を複数のブロックに分割して、分割ブロック毎にヒストグラムを作成し、改めて分割ブロック毎に閾値を設定するようにしてもよい。
【００７４】
（４）ステップＳ３０４：ラベリング・図形融合処理
マイクロコンピュータ１１は、画像中で同じ輝度値の画素が互いに連結している連結部分の魂に対してラベリングを行う。つまり異なる連結部分に対して異なるラベルを貼り付けて区別して領域（連結領域）を分離する（図１７ラベリング１〜９参照）。
【００７５】
また図形融合処理では、画像に含まれている穴のような面積の小さい図形や点状の図形は、本質的に有効でないばかりか、ノイズとして後の処理に悪影響を及ぼす可能性があるので、除去する必要がある。そのためマイクロコンピュータ１１は、元の図形を膨らませたり縮めたりしてノイズ成分を除去する。
【００７６】
（５）ステップＳ３０５：細線化処理
この処理は、得られた２値画像を対象としてその中に含まれる各々の連結領域に対して連結性を損なうことなく線幅１の線図形まで細める処理である。すなわち、任意の太さの線状の図形において、その幅方向の画素を順次取り除くことにより線図形の中心線を求める。
【００７７】
（６）ステップＳ３０６：形状判断処理
ここで連結領域の面積はその連結領域に属する画素の個数である。周囲長は連結領域のまわりに境界に位置する画素の個数である。ただし、斜め方向は水平、垂直方向に対して√２倍に補正する。
【００７８】
画像の形状を判定するために、以下の係数ｅが使用される。
【００７９】
ｅ＝（周囲長）＾２／（面積）
ｅは、形状が円形の時に最小値を示し、形状が複雑になるほど大きい値を示す。
【００８０】
人物の顔はほぼ円形に近いと考えられるので、上記ｅと所定値とを比較して対称画像が人物の顔か否かを判定する。
【００８１】
また上記連結領域面積も所定値と比較して、対称画像人物の顔か否かを判定する。また、形状判定に先立ち、面積を所定範囲の値と比較して所定範囲以外の場合は人物ではない画像と判別して、形状判定処理を行わないようにしてもよい。このようにして演算量を減少させてＡＦタイムラグを縮小させることができる。
【００８２】
図１５は、上記形状判定処理ルーチンを示したフローチャートである。
図に示すように、マイクロコンピュータ１１は、抽出領域があるか否かを判別し（ステップＳ６０１）、抽出領域がない場合はリターンする。ここで抽出領域がある場合は抽出領域の面積Ｓを求め、所定範囲内であるか否かを判別する（ステップＳ６０２）。そして抽出領域面積Ｓが所定範囲の場合はステップＳ６０３に移行し、所定範囲内ではない場合はステップＳ６０７に移行する
上記ステップＳ６０２において抽出領域面積Ｓが所定範囲の場合は、マイクロコンピュータ１１は形状判定値ｅを算出し、所定範囲内か否かを判別する（ステップＳ６０３）。そして、所定範囲内の場合は人物であると判定する（ステップＳ６０４）。
【００８３】
この後マイクロコンピュータ１１は、全抽出領域について形状判定したか否かを判別し（ステップＳ６０５）、終了している場合はリターンする。一方、終了していない場合は、次の抽出領域を設定して（ステップＳ６０６）、上記ステップＳ６０２に移行して上記処理を繰り返し実行する。
【００８４】
一方、マイクロコンピュータ１１は、上記ステップＳ６０２において抽出領域面積Ｓが所定範囲にない場合、及び上記ステップＳ６０３において形状判定値ｅが所定範囲内にない場合は、共に人物以外の被写体と判定する（ステップＳ６０７）。この後、上記同様に全抽出領域について形状判定したか否かを判別する（ステップＳ６０５）。
【００８５】
ここで、図１６乃至図１８を参照して本実施形態における人物判定画像について説明する。
【００８６】
図１６は、本実施形態における人物判定画像の一例を示した図であり、撮影画面の対応するＡＦエリアセンサ１２の画像領域当該画像である。なお、この画像を原画像とする。
【００８７】
図１７は、図１６に示す原画像を用いて差分処理、２値化処理を施した後の画像を示す図である。
図に示すようにエッジ部分（輪郭）のみ抽出された画像となっている。また、抽出エリアにラベリング処理を施している（ラベリング１〜９）。
【００８８】
図１８は、本実施形態における人物判定領域および人物判定領域内の設定した複数の測距エリアを示す説明図である。
図に示すように、本実施形態では、人物の顔と判定された領域３０１が抽出され、人物判定領域３０１内の複数の測距エリアを設定し測距するようになっている。
【００８９】
なお、図７のステップＳ２０９において、マイクロコンピュータ１１は人物判定領域３０１内の画素に対応する領域にモニタ範囲を設定するようＡＦエリアセンサ１２に対して司令を出力する。
【００９０】
この領域のモニタ信号に基づいて再度積分動作を行うことにより、逆光時の高輝度背景の影響を受けて人物判定領域のセンサデータがつぶれてしまうのを防止でき、人物判定領域３０１に対して最適なセンサデータが得られ、高精度な測距演算を行うことが可能となる。
【００９１】
次に形状判定の別の手法として、予め主要被写体のパターンを記憶しておき基準画像とし、この基準画像とパターンマッチング処理を行うことによって抽出する手法を、図１９乃至図２１を参照して説明する。
【００９２】
図１９は、形状判定の別の手法において使用する人物判定画像の例を示した図であり、撮影画面の対応するＡＦエリアセンサ１２の画像領域当該画像である。なお、この画像を原画像として以下、形状判定の別の手法を説明する。
【００９３】
図２０は、図１９に示す原画像を用いて差分処理、２値化処理を施した後の画像を示す図である。
【００９４】
図に示すようにエッジ部分（輪郭）のみ抽出された画像となっている。また、抽出エリアにラベリング処理を施している（ラベリング１〜９）。当該別手法においては、予めＥＥＰＲＯＭ１１ｅに記憶されている主要被写体のパターン３００を基準画像とし（図２２参照）、この基準画像３００と、上記２値化処理後の画像との間でパターンマッチング処理（相関演算）を行うことによって、人物像を抽出する。
【００９５】
上記基準画像３００は、図２２に示すように、被写体距離変化に対応して複数の相似パターンＡ、Ｂ…が準備されており、撮影レンズの焦点距離（ズーム駆動部２２からの情報）等の条件に応じて選択される。
【００９６】
また、カメラの姿勢に応じて複数のパターンが準備されており、カメラ姿勢検出部１２０の出力に基づいて姿勢を判別し、パターンを選択することができる。
【００９７】
さらに人物パターンに限らず、さまざまな物体のパターンが準備されており、人物パターンが検出できない場合に、予め決められた優先順位に従って選択されたパターンマッチング処理がなされる。
【００９８】
図２１は、当該別手法における人物判定領域および人物判定領域内の設定した複数の測距エリアを示す説明図である。
【００９９】
図に示すように、当該別手法では、人物と判定された領域が抽出され人物判定領域３０２内の複数の測距エリアを設定して測距を行う。これら複数の測距エリアの測距結果は、平均処理や最至近選択等の処理により一個の測距データにまとめられレンズ駆動が行われる。
【０１００】
なお、図７のステップＳ２０９においては、マイクロコンピュータ１１は人物判定領域３０２内の画素に対応する領域にモニタ範囲を設定するようＡＦエリアセンサ１２に対して司令を出力する。そして、この領域のモニタ信号に基づいて再度ＡＦエリアセンサ１２の積分動作を行うことにより、人物判定領域３０２に対して最適なセンサデータが得られ、高精度な測距演算が可能である。
【０１０１】
以上説明したように、本第１の実施形態の測距装置によると、主要被写体の輪郭を検出し、広範囲な測距エリアのうちで輪郭内について重点的に測距を行うので、背景に影響されることなく主要被写体に正確にピントをあわせることが可能となり、かつ、タイムラグを縮小した測距装置を提供することができる。
【０１０２】
次に、本発明の第２の実施形態について説明する。
この第２の実施形態は、その構成は図１に示す限りにおいて上記第１の実施形態と同様である。したがってここでは差異のみの言及に留め、作用において異なる部分のみを説明する。
【０１０３】
本第２の実施形態の測距装置は、上記第１の実施形態に比して主要被写体検出ルーチンのみを異にし、その他の構成、作用は第１の実施形態と同様であるので、ここでの詳しい説明は省略する。
【０１０４】
図２４は、本第２の実施形態の測距装置における主要被写体検出ルーチンを示したフローチャートである。
【０１０５】
図に示すように、まずマイクロコンピュータ１１は、プリ主要被写体検出を行い（ステップＳ７０１）、撮影画面全体に対応するＡＦエリアセンサ１２のセンサデータよりおおまかに主要被写体の位置を求める。次に、マイクロコンピュータ１１は、本主要被写体検出を行い（ステップＳ７０２）、プリ主要被写体検出の結果に基づいて検出ブロックを新たに設定し、このブロック内で詳細に主要被写体本検出を行う。
【０１０６】
ここで、上記プリ主要被写体検出（プリ検出）ルーチンについて図２５を参照して詳しく説明する。
【０１０７】
このプリ主要被写体検出では、マイクロコンピュータ１１は、まず全領域内のセンサデータについて差分処理を行う（ステップＳ８０１）。その後、全領域内のセンサデータに関するヒストグラムを作成し、閾値を設定して２値化処理を行う（ステップＳ８０２）。さらに、２値化処理結果についてラベリング処理を行い（ステップＳ８０３）、ラベリングされた連結領域につい形状判定処理を行う（ステップＳ８０４）。
【０１０８】
この後、マイクロコンピュータ１１は上記ステップＳ８０４において求めた形状判定処理結果より抽出エリアを仮決定する（ステップＳ８０５）。なお、本第２の実施形態においては、処理時間短縮のために当該プリ検出を行う際、センサデータを所定画素毎に間引いたり、加算平均をとって処理画素数を減らすようにしている。
【０１０９】
次に、上記主要被写体本検出（本検出）ルーチンについて図２６を参照して詳しく説明する。
【０１１０】
この主要被写体本検出では、マイクロコンピュータ１１は、まず上記プリ主要被写体検出処理の結果に基づいて、抽出エリアを含む検出ブロック３０５を設定する（ステップＳ９０１、図２７参照）。次に、マイクロコンピュータ１１は上記検出ブロック内について差分処理を行い（ステップＳ９０２）、該ブロック内のセンサデータに基づいてヒストグラムを作成してブロック毎に閾値を決定する（ステップＳ９０３）。
【０１１１】
この後マイクロコンピュータ１１はブロック毎に２値化処理（ステップＳ９０４）、ラベリング処理（ステップＳ９０５）、形状判定処理（ステップＳ９０６）を行い、抽出エリアを決定する（ステップＳ９０７）。
【０１１２】
このように撮影画面に対応するＡＦエリアセンサ１２の画素部分のセンサデータ全体について、エッジ検出して輪郭より主要被写体をおおまかに検出する。
【０１１３】
そして、その結果に基づいて検出領域付近に新たな検出ブロックを設定し、詳細に主要被写体検出を行うので、閾値を最適に設定することができ、良好な２値化画像が得られ、主要被写体検出精度を大幅に向上させることができる。
【０１１４】
また、処理時間がやや長くなるのを許容するならば、最初から撮影画面に対応するＡＦエリアセンサ１２の画素領域を複数のブロックに分割して、ブロック毎に本検出（図２６参照）を行ってもよい。すなわち、ブロック毎にヒストグラム作成、閾値設定、２値化処理、形状判定処理を行い、主要被写体エリアを抽出する。このとき分割ブロックの大きさを、撮影レンズの焦点距離（ズーム駆動部２２からの情報）等の条件に応じて可変（広角側で大、望遠側で小）としてもよい。
【０１１５】
また、ストロボ２０ａ、ストロボ回路部２０による被写体への投光動作と同時に、ＡＦエリアセンサ１２による積分動作を行い、かつ定常光除去回路１２ｃを動作させて得られるセンサデータについて主要被写体検出を行ってもよい。
【０１１６】
主要被写体と背景の輝度が非常に近く、エッジ（輪郭）を検出しにくい場合であっても、上記投光により距離の差による反射光量の差が発生するので、精度よくエッジ（輪郭）検出を行うことができる。
【０１１７】
このように、本第２の実施形態の測距装置によると、主要被写体の輪郭を検出して広範囲な測距エリアのうちで輪郭内について重点的に測距を行うので、背景に影響されることなく主要被写体に正確にピントをあわせることが可能となり、かつ、タイムラグを縮小した測距装置を提供することができる。
【０１１８】
次に、本発明の第３の実施形態の測距装置について説明する。
図２８は、本第３の実施形態である測距装置を備えるカメラの主要構成を示したブロック図である。
当該カメラは、いわゆるスーパーコンビネーションＡＦと称されるオートフォーカス方式を採用したカメラであって、カメラ全体の制御を司るＣＰＵ５０１と、被写体像を受像して光電変換した後に上記ＣＰＵ５０１に測距情報を提供する測距部５０２と、ＣＰＵ５０１の制御下に被写体に向けてストロボ光を発光する発光源５０５ａと、この発光源５０５ａの発光を制御する回路５０５と、測距種別情報あるいは主要被写体の判別の可否情報を音声として発する音声信号発生回路５０７と、で主要部が構成される。
【０１１９】
上記測距部５０２は、２つの受光レンズから被写体光を入射して光電変換するエリアセンサ５０２ａ、５０２ｂと、このエリアセンサ５０２ａ、５０２ｂの出力をＡ／Ｄ変換して各画素のデジタル値をカメラ制御用のマイクロコンピュータである上記ＣＰＵ５０１に対して出力するＡ／Ｄ変換回路５０２ｃと、ＣＰＵ５０１の制御下にエリアセンサ５０２ａ、５０２ｂの出力より定常光を除去する定常光除去回路５１１と、を備えている。
【０１２０】
上記定常光除去回路５１１は被写体から定常的に入射する直流的な光の信号を除去する回路であり、エリアセンサ５０２ａ、５０２ｂで受光した発光源５０５ａからのパルス光のみを出力信号とする作用をする回路である。
【０１２１】
ＣＰＵ５０１の制御の下、定常光除去回路５１１を作動させた状態で発光回路５０５により発光源５０５ａが発光すると、被写体５２０からの反射信号光を受けたエリアセンサ５０２ａ上には、図２９に示す如き像が結ばれる。なお、この図２９において、黒で示す部分が光を入射したところを示している。
【０１２２】
また、ＣＰＵ５０１には、上述の如きエリアセンサ上の像のパターンを分析するソフトウェア（パターン判別プログラム５０１ａ）が組み込まれている。このパターン判別プログラム５０１ａにより、被写体像が人間の形であると判定されれば、これを主要被写体と考えることができることになる。なお、パターン判別プログラムは、上記第１実施形態における図１０に示す「主要被写体検出」と同様の処理を行う。
【０１２３】
次に、本実施形態の測距装置による測距動作を図３０に示すフローチャートを参照して説明する。
【０１２４】
図３０は、本第３の実施形態の測距装置における測距ルーチンを示したフローチャートである。
【０１２５】
ＣＰＵ５０１は、まず測距に先立って、いわゆるプリ発光のために発光源５０５ａを作動させて補助光照射を行うとともに、定常光除去回路５１１を作動させてエリアセンサ５０２ａ、５０２ｂからの出力より定常光を除去する（ステップＳ１１０１）。このプリ発光、定常光除去動作により被写体５２０の反射信号光のパターンのみが図２９に示す如く抽出される。ＣＰＵ５０１はその内部プログラムであるパターン判別プログラム５０１ａに基づいて当該反射信号光より被写体像が人間の形であるか否かを判別する（ステップＳ１１０２）。
【０１２６】
次にＣＰＵ５０１は、上記パターン判別プログラム５０１ａに基づく処理により主要被写体の位置を検出する（ステップＳ１１０３）。ここでＣＰＵ５０１は、被写体像が人間の形であると判別できる場合には、当該被写体は主要被写体であるとの判定を行いステップＳ１１０４に移行する。このときＣＰＵ５０１は、上記パターンを形成する光信号が弱いか否かあるいは十分なコントラストがあるかに基づいてかかる判別を行う。
【０１２７】
ステップＳ１１０４においてＣＰＵ５０１は、測距方式がいわゆるアクティブ方式（信号光をカメラ側から投射して、その反射信号光を用いて測距するタイプ）と、パッシブタイプ（被写体の像信号をもとに測距するタイプ）のいずれの方式を用いるかの選択を行なう。ここで像信号のコントラストが弱いときには再度測距用光（発光源５０５ａによる補助光）を照射して（ステップＳ１１１０）、その反射信号光によるアクティブオートフォーカスを、先に求められた主要被写体位置、具体的には主要被写体の輪郭領域内部に複数の測距エリアを設定して重点的に行う（ステップＳ１１１１）。
【０１２８】
この後、ＣＰＵ５０１は音声信号発生回路５０７を制御して、アクティブオートフォーカスによる測距であること、および主要被写体の判別の可否情報を音声として発して（ステップＳ１１１２）、リターンする。なお、かかる情報に対応した音声パターンを（音声パターン３）とする。
【０１２９】
一方、上記ステップＳ１１０４において、被写体からの反射信号光が弱いと判断された場合、ＣＰＵ５０１は、パッシブオートフォーカスの方が適しているとしてステップＳ１１０５に分岐する。このステップでは、すでに求められた主要被写体位置の像信号の輪郭領域内に複数の測距エリアを設定し重点的に用いたパッシブ方式による測距を行う。
【０１３０】
この後、ＣＰＵ５０１は音声信号発生回路５０７を制御して、パッシブオートフォーカスによる測距であること、および主要被写体の判別の可否情報を音声として発して（ステップＳ１１０６）、リターンする。なお、かかる情報に対応した音声パターンを（音声パターン２）とする。
【０１３１】
また、上記ステップＳ１１０３において主要被写体を検出することができなかった場合は、輝度情報等を加味してアクティブ方式またはパッシブ方式を選択した後、被写体の存在確率の高い画面中央部を重点的に測距する（ステップＳ１１２０）。
【０１３２】
この後、ＣＰＵ５０１は音声信号発生回路５０７を制御して、中央測距による測距であることおよび主要被写体の判別の可否情報を音声として発して（ステップＳ１１２１）、リターンする。なお、かかる情報に対応した音声パターンを（音声パターン１）とする。
【０１３３】
このように、各測距方式、すなわちステップＳ１１０５におけるパッシブオートフォーカス、ステップＳ１１１１におけるアクティブオートフォーカス、ステップＳ１１２０における中央測距の各測距方式毎に、測距種別の情報および主要被写体の判別の可否情報が音声信号発生回路５０７から発せされる。
【０１３４】
また、本実施形態は、単にアクティブ方式とパッシブ方式とをハイブリッド的に組み合わせたのではなく、２つの方式を用いて、主要被写体検知まで行っていることを特徴とするものである。
【０１３５】
以上説明したように、本第３の実施形態によると、ユーザーにわかりやすく、安心感のある測距を行い得るカメラを提供することができる。
【０１３６】
以上述べたように、上記各実施形態の測距装置によると、主要被写体の輪郭を検出し、広範囲な測距エリアのうちで輪郭内について重点的に測距を行うので、背景に影響されることなく主要被写体に正確にピントをあわせることが可能であり、かつタイムラグを縮小可能な測距装置を提供することができる。
【０１３７】
なお、本実施形態の測距装置は外光式測距装置としているが、これに限らず、ＴＴＬ方式の測距装置にも適用することができ、同様の効果を奏する。
【０１３８】
[付記]
以上詳述した如き本発明の実施形態によれば、以下の如き構成を得ることができる。即ち、
（１）
二次元領域内を撮像する積分型エリアセンサと、
この積分型エリアセンサの積分動作を制御する積分制御手段と、
上記積分型エリアセンサの出力に基づいて、上記二次元領域内に存在する主要被写体を含む領域を抽出する領域抽出手段と、
この領域抽出手段により抽出された領域に対して測距演算を行う測距演算手段と、
この測距演算手段により得られた測距結果を所定の判定条件と比較する比較手段と、
この比較手段による比較結果が所定条件に合致しない場合、上記積分制御手段による制御パラメータを変更してから上記積分型エリアセンサの積分動作を再度実行させる制御手段と、
を具備したことを特徴とする測距装置。
【０１３９】
（２）
上記領域抽出手段は、上記積分型エリアセンサから出力される画像データに所定の演算処理を行い、画像データに含まれる物体像同士をそれぞれ分離し、分離された各データと特定のパターンとを照合して物体像が人物像であるか否かを判定すると共に、人物像と判定した領域の情報を出力することを特徴とする（１）に記載の測距装置。
【０１４０】
（３）
上記所定の演算処理は、エッジ抽出処理および二値化処理を含むことを特徴とする（２）に記載の測距装置。
【０１４１】
【発明の効果】
以上説明したように本発明によれば、広範囲な測距領域を有する測距装置であっても、主要被写体の領域内に対応する画素のモニタ信号に応じた積分制御を行い、タイムラグを増大させることなく、かつ背景に影響を受けずに主要被写体に対して正確なオートフォーカス処理を行い得る測距装置を提供することができる。
【図面の簡単な説明】
【図１】本発明の第１の実施形態である測距装置の構成を示したブロック図である。
【図２】上記第１の実施形態の測距装置において、マイクロコンピュータのメインルーチンを示すフローチャートである。
【図３】上記第１の実施形態の測距装置における測距光学系を示した説明図である。
【図４】三角測距の原理により被写体距離を求める方法を説明するための図である。
【図５】上記第１の実施形態の測距装置において、ＡＦエリアセンサの構成を示したブロック図である。
【図６】上記第１の実施形態の測距装置における撮影画面（ワイドとテレ）と測距領域との関係について説明した図である。
【図７】上記第１の実施形態の測距装置におけるオートフォーカス（ＡＦ）ルーチンを示したフローチャートである。
【図８】上記第１の実施形態の測距装置における主要被写体検出動作を示したフローチャートである。
【図９】上記第１の実施形態の測距装置において、差分処理によるエッジ検出処理の前後画像を示した線図である。
【図１０】上記第１の実施形態の測距装置において、差分処理における空間フィルタテーブルの例を示した説明図である。
【図１１】上記第１の実施形態の測距装置において、２値化処理を示したフローチャートである。
【図１２】上記第１の実施形態の測距装置において、２値化処理の過程でモード法により閾値を決定する手法を説明するための線図である。
【図１３】上記第１の実施形態の測距装置において、閾値設定処理を説明するための線図である。
【図１４】上記第１の実施形態の測距装置において、閾値設定処理を示したフローチャートである。
【図１５】上記第１の実施形態の測距装置において、形状判定処理ルーチンを示したフローチャートである。
【図１６】上記第１の実施形態の測距装置において、人物判定画像の一例を示した図である。
【図１７】上記第１の実施形態の測距装置において、図１６に示す原画像を用いて差分処理、２値化処理を施した後の画像を示す図である。
【図１８】上記第１の実施形態の測距装置において、人物判定領域および人物判定領域内の設定した複数の測距エリアを示す説明図である。
【図１９】上記第１の実施形態の測距装置において、形状判定の別の手法において使用する人物判定画像の例を示した図である。
【図２０】上記第１の実施形態の測距装置において、図１９に示す原画像を用いて差分処理、２値化処理を施した後の画像を示す図である。
【図２１】上記第１の実施形態の測距装置において、形状判定の別の手法における人物判定領域および人物判定領域内の設定した複数の測距エリアを示す説明図である。
【図２２】上記第１の実施形態の測距装置におけるＥＥＰＲＯＭに記憶されている、形状判定の別の手法における主要被写体のパターンの一例を示した説明図である。
【図２３】上記第１の実施形態の測距装置において、差分処理によるエッジ検出処理の際に用いるセンサデータｓ（ｉ，ｊ）について示した説明図である。
【図２４】本発明の第２の実施形態の測距装置における主要被写体検出ルーチンを示したフローチャートである。
【図２５】上記第２の実施形態の測距装置におけるプリ主要被写体検出（プリ検出）ルーチンを示したフローチャートである。
【図２６】上記第２の実施形態の測距装置における主要被写体本検出（本検出）ルーチンを示したフローチャートである。
【図２７】上記第２の実施形態の測距装置における主要被写体本検出において、新たな検出ブロックの一例を示した説明図である。
【図２８】本発明の第３の実施形態の測距装置を備えるカメラの主要構成を示したブロック図である。
【図２９】上記第３の実施形態の測距装置におけるたエリアセンサ上に結像される像の一例を示した図である。
【図３０】上記第３の実施形態の測距装置における測距ルーチンを示したフローチャートである。
【符号の説明】
１１…マイクロコンピュータ
１２…ＡＦエリアセンサ
１２ａ…画像領域
１２ｂ…処理回路
１２ｃ…定常光除去回路
１３…フォーカスレンズ駆動部
１４…フォーカスレンズ
１５…フォーカスレンズエンコーダ
１６…シャッタ駆動部
１７…１ＲＳＷ
１８…２ＲＳＷ
１９…表示部
２０…ストロボ回路部
２１…フィルム駆動部
２２…ズームレンズ駆動部
２３…測光部
１２０…カメラ姿勢検出部[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a distance measuring device, and more particularly to a distance measuring device that can be used for a camera or the like.
[0002]
[Prior art]
Conventionally, various cameras having an autofocus function for measuring a distance at a plurality of distance measuring points have been proposed. For example, Japanese Patent Laid-Open No. 10-142490 discloses a camera having a distance measuring device that has a number of distance measuring points over a wide range in a shooting screen and measures all the distance measuring points for each shooting. Yes.
[0003]
In recent years, cameras that have been devised so as not to be affected by the background have been proposed. Japanese Patent Laid-Open No. 5-7326 proposes a distance measuring device that detects an edge portion in a shooting screen, generates a gate signal to set a distance measuring frame, and is devised so as not to be affected by the background. Yes.
[0004]
[Problems to be solved by the invention]
However, the distance measuring device disclosed in the above-mentioned Japanese Patent Application Laid-Open No. 10-142490 has a problem that the time lag becomes very large and the usability is inconvenient because distance measurement is performed at every distance measuring point for each photographing.
[0005]
In addition, if a very high-speed processing circuit or microcomputer is used in order to shorten the time lag, there is a problem that the cost is greatly increased and it cannot be adopted for a general-purpose camera.
[0006]
Further, the distance measuring device disclosed in the above Japanese Patent Laid-Open No. 5-7326 is limited to a hill-climbing autofocus device using a contrast method for detecting a high frequency component from a video signal. This contrast method detects the position of the focus lens at which the highest frequency component is greatest while scanning the focus lens, so there is a problem that the time lag is very large and a photo opportunity is missed.
[0007]
The present invention has been made in view of such problems, and even a distance measuring device having a wide range of distance measurement does not increase the time lag and is not affected by the background. An object of the present invention is to provide a distance measuring device capable of performing an automatic focusing process.
[0008]
[Means for Solving the Problems]
  In order to achieve the above object, a first distance measuring device of the present invention mainly includes an image sensor that has a plurality of images, picks up a subject image, and outputs an image pickup signal, based on an output of the image sensor. Contour extraction means for detecting a subject and extracting the contour of the main subject, monitor selection means for setting a monitor signal extraction range indicating the integration amount of pixels of the image sensor, and a range set by the monitor selection means Based on the monitor signal, the integration control means for controlling the integration operation of the imaging signal, and the integration control means executes the integration control of the imaging element to divide the subject image obtained from the imaging element into two images A distance measuring unit that performs a distance measuring operation by performing a correlation operation on the captured image signal, and the monitor selecting unit obtains an appropriate image signal for the main subject region extracted by the contour extracting unit. As the extraction range of the monitor signal of the main subject areaOf the extracted contourInsideIn the corresponding pixel area of the image sensorIt is characterized by setting.
[0009]
  In order to achieve the above object, the second distance measuring device of the present invention provides:An auxiliary light unit that irradiates the subject with auxiliary light; and a stationary light removal unit that removes a stationary light component from the output of the image sensor, wherein the contour extraction unit irradiates the subject with the auxiliary light unit, The contour of the subject is extracted based on the output of the image sensor from which the steady light component has been removed by the steady light removing means.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings.
[0013]
FIG. 1 is a block diagram showing the configuration of the distance measuring apparatus according to the first embodiment of the present invention.
[0014]
As shown in the figure, the distance measuring device of the present embodiment captures a subject image formed by a microcomputer 11 that controls each circuit in the entire camera including the distance measuring device and a distance measuring optical system 100 described later. Then, the AF area sensor 12 that converts the sensor data that is an electrical signal, the focus lens drive unit 13 that drives the focussing lens 14, and the focus lens encoder that generates a pulse signal corresponding to the amount of movement of the focussing lens 14 15 and a photometric unit 23 for processing a photocurrent signal generated by a photometric light-receiving element 23a divided into a plurality of parts and generating a photometric output, and a shutter for driving the shutter to expose the film. Drive unit 16, strobe circuit unit 20 that emits strobe light 20a as an auxiliary light source during photographing, and camera Display unit 19 that displays information on the display unit using a display element such as an LCD, 1RSW (first release switch) 17 and 2RSW (second release switch) 18 that are linked to a release button, auto load, one frame winding, winding A film drive unit 21 that performs a return film drive operation, a zoom lens drive unit 22 that performs a zoom operation of the photographing lens, and a camera posture detection that detects the posture (vertical and horizontal) of the camera and outputs the detected result to the microcomputer 11 The main part is composed of the part 120.
[0015]
The microcomputer 11 includes a CPU (central processing unit) 11a, a ROM 11b, a RAM 11c, and an A / D converter ADC 11d therein. Among these, the CPU 11a performs a series of operations according to a sequence program stored in the ROM 11b.
[0016]
The microcomputer 11 further has an EEPROM 11e, and stores correction data relating to autofocus (AF), photometry / exposure calculation, and the like for each camera. The EEPROM 11e stores various parameters for detecting a main subject in a shooting screen described later.
[0017]
The AF area sensor 12 includes a light receiving element group that is two-dimensionally arranged in the horizontal direction and the vertical direction, which are the imaging region 12a, and a processing circuit 12b. Then, the electric charge generated by the incident light to the light receiving element (photodiode) is converted into a voltage by a pixel amplification circuit for each pixel, amplified and output. The microcomputer 11 controls the integration operation of the AF area sensor 12 and reads out the sensor data, processes the sensor data output from the AF area sensor 12 and performs a distance measurement calculation.
[0018]
The AF area sensor 12 has a steady light removal circuit 12c. The stationary light removal circuit 12c has a function of switching whether or not to remove stationary light under the control of the microcomputer 11.
[0019]
The focus lens driving unit 13 drives a focusing lens 14 which is a part of the photographing lens 130, and the focusing lens encoder 15 generates a pulse signal corresponding to the amount of movement of the focusing lens 14. The microcomputer 11 controls the position of the focus lens 14 by outputting a drive signal to the focus lens drive unit 13 based on the distance measurement calculation result and monitoring the output of the focus encoder 15.
[0020]
The photometry unit 23 corresponds to the photographing screen and processes a photocurrent signal generated by the photometric light receiving element 23a divided into a plurality of parts to generate a photometric output. The microcomputer 11 AD converts this photometric output by the AD converter ADC 11d to perform photometric / exposure calculation.
[0021]
The shutter driving unit 16 drives the shutter under the control of the microcomputer 11 to expose the film.
[0022]
The strobe circuit unit 20 has a function of causing the strobe 20a to emit light as an auxiliary light source at the time of photographing, and charging and emission control for light emission of the strobe 20a are performed under the control of the microcomputer 11. The strobe circuit unit 20 performs light emission control under the control of the microcomputer 11 when the strobe 20a is used as autofocus auxiliary light during the distance measuring operation.
[0023]
The display unit 19 displays information inside the camera on a display element such as an LCD under the control of the microcomputer 11.
[0024]
1RSW (first release switch) 17 and 2RSW (second release switch) 18 are interlocked with the release button, and 1RSW 17 is turned on when the release button is depressed in the first stage, and then 2RSW 18 is activated when the release button is depressed in the second stage. Turn on. The microcomputer 11 performs AF and photometry operations when the 1RSW 17 is turned on, and performs an exposure operation and a film winding operation when the 2RSW 18 is turned on.
[0025]
The film drive unit 21 performs auto-load, 1-frame roll-up and rewind film drive operations under the control of the microcomputer 11, and the zoom lens drive unit 22 also performs the zoom operation of the photographing lens under the control of the microcomputer 11. Do. Further, the focal length information of the photographing lens is output to the microcomputer 11.
[0026]
The camera posture detection unit 120 detects the posture (vertical, horizontal) of the camera and outputs it to the microcomputer 11.
[0027]
Next, the operation of the distance measuring apparatus of the present embodiment having such a configuration will be described.
FIG. 2 is a flowchart showing a main routine of the microcomputer 11 in the distance measuring device of the present embodiment.
[0028]
First, when a power switch (not shown) is turned on or a battery is inserted, the microcomputer 11 starts to operate and executes a sequence program stored in the ROM 11b. Then, the microcomputer 11 expands the adjustment / correction data such as initialization of each block in the camera, AF in the EEPROM 11e, and photometry in the RAM 11c (step S101).
[0029]
Next, the microcomputer 11 detects the state of the 1RSW 17, and waits for the ON operation of the 1RSW 17 (step S102). When the 1RSW 17 is turned on, the microcomputer 11 controls the corresponding circuit to perform an autofocus (AF) operation (step S103). Subsequently, photometric / exposure calculation processing (step S104) is performed to detect the state of the 2RSW 18 (step S105).
[0030]
When the 2RSW 18 is turned on in step S105, the microcomputer 11 instructs to perform a shutter operation to expose the film (step S106), and winds up the film by one frame (step S107).
[0031]
On the other hand, when the 1RSW 17 is not turned on in step S102, the microcomputer 11 detects an input of a switch other than 1RSW17 and 2RSW18 (step S108). Here, when another switch input is detected, each circuit is instructed to perform processing corresponding to the switch input, for example, zoom up / down processing for zoom switch up / down switch input (step S109).
[0032]
Next, the distance measuring optical system in the distance measuring apparatus of this embodiment will be described.
FIG. 3 is an explanatory view showing the distance measuring optical system 100 in the distance measuring apparatus of the present embodiment, and shows the arrangement of the optical system and the AF area sensor. FIG. 4 is a diagram for explaining a method for obtaining the subject distance based on the principle of triangulation.
[0033]
The distance measuring optical system 100 measures the distance to a subject by a so-called external light passive method, and as shown in FIG. 3, the light receiving lenses 101 and 102 are arranged with a base line length B apart from each other. The image 103 is divided into two images and formed on the light receiving region 12 a of the AF area sensor 12.
[0034]
As shown in FIG. 4, the relative position difference x between the two images is based on the principle of triangulation, and the subject distance L is obtained from the focal length f and the base length B of the light receiving lens according to the following equation.
[0035]
L = (B · f) / x
The distance measurement calculation described above is performed by the microcomputer 11. More specifically, a distance measurement block is set in the light receiving area 12a of the AF area sensor 12, and correlation calculation is performed using sensor data corresponding to the two images to detect a relative positional difference x between the two images. .
[0036]
Next, the configuration of the AF area sensor 12 will be described with reference to FIG.
As shown in FIG. 5, the AF area sensor 12 includes a plurality of pixels 53 corresponding to a shooting screen, a monitor selection circuit 57 for controlling an integration operation, a horizontal shift register 56, a vertical shift register 54, and a fixed pattern. And a noise removal circuit 55.
[0037]
As described above, the AF area sensor 12 is provided with a plurality of pixels 53 corresponding to the shooting screen. Among the plurality of pixels, one pixel 50 includes a photodiode 52 that is a light receiving element, and a photodiode 52. And an amplifier 51 (including a storage capacitor 58) for converting the signal charge output from the signal charge into a voltage signal. The amplifier 51 also includes a function for removing a steady light component.
[0038]
The monitor selection circuit 57 creates and outputs a monitor signal indicating the integration amount for the pixel range based on the command from the microcomputer 11.
[0039]
The horizontal shift register 56 and the vertical shift register 54 are controlled by a command from the microcomputer 11 and select and output the signal output of each pixel.
[0040]
The fixed pattern noise removal circuit 55 is a circuit for removing fixed pattern noise included in the signal output of each pixel.
[0041]
Next, with reference to FIG. 6, the relationship between the photographing screen (wide and tele) and the distance measurement area in this embodiment will be described.
As described above, since the distance measuring device of the present embodiment employs the outside light distance measuring method, there is a parallax between the shooting screen and the distance measuring area. For this reason, in this embodiment, the area | region used for ranging is limited according to the focal distance information (zoom information) of an imaging optical system. Distance measurement area position correction data corresponding to such a change in focal length is stored in advance in the EEPROM 11e, and is expanded in the RAM 11d along with the initialization of the microcomputer 11. Then, the distance measurement area to be used for the distance measurement operation in the light receiving area of the AF area sensor 12 is determined with reference to the correction data according to the zoom operation. Further, the distance measurement calculation is performed based on the sensor data within the distance measurement area range.
[0042]
Further, the microcomputer 11 outputs a control signal to the AF area sensor 12 so as to generate a peak monitor for integration control corresponding to this distance measurement area. The AF area sensor 12 outputs a peak signal within the range of the designated distance measuring area to the microcomputer 11. The microcomputer 11 refers to the monitor signal and controls the integration amount to be a predetermined level.
[0043]
In the present embodiment, it is possible to prevent the influence of the subject from being influenced outside the shooting screen by such a device. Also, when reading sensor data, unnecessary sensor data outside the shooting screen is read out and stored in the RAM 11c with reference to the distance measurement area correction data corresponding to the shooting screen. Alternatively, a readout range setting signal is output to the AF area sensor 12 so that only sensor data within the set range is output.
[0044]
Next, an autofocus (AF) routine in the distance measuring apparatus of the present embodiment will be described with reference to the flowchart of FIG.
[0045]
First, the microcomputer 11 outputs an integration control signal to the AF area sensor 12 to instruct to perform an integration operation (step S201). Next, a monitor signal corresponding to a peak (brightest pixel) output within a predetermined range is output from the AF area sensor 12. The microcomputer 11 adjusts the integration time so that the amount of light received by the light receiving portion of the AF area sensor 12 is appropriate while referring to the monitor signal.
[0046]
Thereafter, the microcomputer 11 outputs the read clock CLK to the AF area sensor 12 (step S202), causes the sensor data (pixel data) to be output to the AD converter ADC 11d, performs AD conversion, and stores it in the read RAM 11c.
[0047]
Further, the microcomputer 11 performs a process of extracting the main subject (step S203), and sets a plurality of distance measurement areas in the extracted main subject area (step S204). Further, ranging calculation is performed for the plurality of ranging areas (step S205), and it is determined whether or not the obtained ranging data satisfies a predetermined condition (reliability or the like) (step S206).
[0048]
If the distance measurement result satisfies the predetermined condition in step S206, the microcomputer 11 performs an averaging process on a plurality of distance measurement data satisfying the predetermined condition (step S207), and the measurement obtained as a result is performed. Adopt distance data. Then, the focusing lens 14 is driven based on the adopted distance measurement data (step S208), and the process returns.
[0049]
On the other hand, if the distance measurement result does not satisfy the predetermined condition in step S206, the microcomputer 11 performs the integration again by controlling the sensor data in the main subject area to be appropriate. That is, the monitor selection circuit 57 in the AF area sensor 12 is controlled so that monitor signals are acquired only for pixels corresponding to the main subject area, and for example, the maximum value is output to the microcomputer 11 as a monitor output. Setting is made (step S209).
[0050]
Thereafter, the process returns to step S201 again, and integration, reading, and distance measurement processing are performed again. In this case, since integration control is performed so that the sensor data in the area of the main subject is appropriate for performing the distance measurement calculation, good distance measurement data can be obtained for the main subject.
[0051]
Next, the main subject detection operation in the distance measuring apparatus of the present embodiment will be described with reference to the flowchart of FIG.
[0052]
In this main subject detection routine, a person is detected assuming a person as the main subject. Although two images are obtained by the AF area sensor 12, either one or both of the image data (sensor data) used for main subject detection may be used. The sensor data of the AF area sensor 12 is stored in the RAM 11c in the microcomputer 11, and the following processing is performed based on this sensor data.
[0053]
First, an overview of the process will be described.
First, the microcomputer 11 performs a smoothing process (step S301). This process is a process of removing random noise in the image, and the noise is removed by filter processing or Fourier transform. The random noise to be removed is random noise that the AF area sensor 12 itself has, or noise that is generated by external noise such as fluctuations in the power supply voltage of the AF area sensor 12.
[0054]
Next, the microcomputer 11 performs difference processing (step S302). In this process, the microcomputer 11 performs a difference process on the sensor data, and an edge candidate region and its strength are given in the process of performing edge detection.
[0055]
Thereafter, the microcomputer 11 performs binarization processing (step S303). In this processing, the microcomputer 11 obtains a binary image by extracting a portion below a certain value from the image by threshold processing.
[0056]
Further, the microcomputer 11 performs a connection / graphic fusion process (step S304), and then performs a thinning process (step S305). Since a graphic having a certain width corresponding to the edge is obtained by this processing, the line width is reduced to about 1 by applying a thinning algorithm.
[0057]
Thereafter, the microcomputer 11 performs shape determination processing for determining the shape of the image and extracting the main subject (step S306), and returns.
[0058]
Next, the process of each step will be described in more detail.
(1) Step S301: Smoothing process
This smoothing process is a process for removing random noise mixed in the image. Various methods are known for this process, but the median filter that obtains the median of the pixel values in the neighborhood area or the neighborhood area is divided into small areas, and the variance is obtained for each small area to minimize the variance. An edge preserving filter that obtains a small area and outputs the average value is effective.
[0059]
The median filter has a side effect that the edge of the image is blurred, but the edge preserving filter is more effective because the edge is not blurred. In addition, there is a method using Fourier transform.
[0060]
(2) Step S302: Edge detection processing by difference processing
In this step, the microcomputer 11 performs edge detection on the sensor data s (i, j) (see FIG. 23) by performing the following processing.
[0061]
In the method using the first-order differential operator, the x-direction differentiation and the y-direction differentiation are calculated by the following equations, respectively.
Δxs (i, j) = s (i, j) −s (i−1, j)
Δys (i, j) = s (i, j) −s (i, j−1)
As a result, data as shown in FIG. 9A is obtained.
[0062]
Further, in the method using the second-order differential operator, it is obtained by the following equation.
Δ ^ 2xs (i, j) = s (i−1, j) −2s (i, j) −s (i + 1, j)
Δ ^ 2ys (i, j) = s (i, j−1) −2s (i, j) −s (i, j + 1)
A Laplacian operator, which is a kind of second-order differential operator, emphasizes the shoulder portion of the edge, and therefore shifts from a positive region to a negative region. Then, an edge is obtained by obtaining a portion that becomes “0” (FIG. 9B).
[0063]
As a specific processing method, a product-sum operation with a spatial filter table (weight table) is performed.
FIG. 10 is an explanatory diagram showing an example of the spatial filter table. In FIG. 10,
FIG. 10A shows the first-order differential operator (lateral direction).
FIG. 10B shows a first-order differential operator (vertical direction).
FIG. 10C shows a Laplacian operator.
FIG. 10 (d) shows a Sobel operator (first derivative in X direction and Y direction, absolute value data conversion, addition).
Respectively.
[0064]
Further, the arithmetic expression of the processing is as shown below.
[Expression 1]

In the present embodiment, the above spatial filter is appropriately selected and used according to the situation.
[0065]
Note that, when the difference processing is performed on all images, a first-order differential operator and a Laplacian operator that are relatively simple and high-speed are used. On the other hand, when the difference process is performed on a part of the images on the shooting screen, a Sobel operator is selected and used because the calculation is slightly complicated and the calculation time is long but the effect is large.
[0066]
When the integration time of the AF area sensor 12 is low and the luminance is long, a first-order differential operator or a Laplacian operator is used. You may balance.
[0067]
(3) Step S303: Binarization processing (threshold processing)
This binarization processing will be described with reference to the flowchart shown in FIG.
In the binarization process, the microcomputer 11 first creates a histogram representing the appearance frequency of pixel values indicating each luminance in the image (step S401), and then performs a threshold setting process (step S402). Here, various methods for determining the threshold value based on the histogram are known. For example, in the mode method, the binarization process is performed using the luminance value with the lowest frequency among the above as the threshold value (threshold level) ( (See FIG. 12).
[0068]
After the threshold is set in step S402, the microcomputer 11 performs binarization (step S403).
[0069]
Other methods for setting the threshold include a p-tile method that is effective when the area of the figure to be extracted is known to some extent, a differential histogram method that determines that a threshold is set at the boundary of the figure, and a set of density values. There are known methods such as a discriminant analysis method for obtaining the parameter t so that the separation between the two classes is best, and a variable threshold method for changing the threshold according to the image position.
[0070]
In this embodiment, these methods are appropriately selected according to the situation and used. For example, the shape of the histogram is discriminated to determine whether or not there is a clear minimum value. If it is clear, the mode method is adopted. On the other hand, discriminant analysis is adopted when it is unclear.
[0071]
In this way, the shape of the histogram is determined, and the threshold setting method is changed according to the result. For example, as shown in FIG. 13, the histogram shape discrimination method is (valley) extreme value and frequency minimum value a, the second smallest value b is obtained, and the difference b−a is compared with the discrimination value dth. If the value is larger than the predetermined value dth, the luminance value of the minimum value a is adopted as the threshold value. On the other hand, when the value is equal to or smaller than the predetermined value, a variable threshold method is adopted in which the threshold is changed according to the image position.
[0072]
Here, the threshold value setting process will be described in detail with reference to the flowcharts shown in FIGS.
The microcomputer 11 obtains the minimum value a and the second smallest frequency b as shown in FIG. 13 (step S501). Next, the difference (b−a) is compared with a predetermined determination value dth (step S502). When the difference (b−a) is larger than the determination value dth, the luminance value Ba corresponding to the minimum value “a” is adopted as a threshold (step S504). On the other hand, when the difference (b−a) is equal to or smaller than the determination value dth, the variable threshold method is employed (step S505).
[0073]
In the case of binarization with an image corresponding to the entire shooting screen, first, a threshold value is set by the mode method and binarization processing is performed. If the result of evaluating the binarized image is not good, the image may be divided into a plurality of blocks, a histogram may be created for each divided block, and a threshold may be set for each divided block.
[0074]
(4) Step S304: Labeling and figure fusion processing
The microcomputer 11 performs labeling on the connected souls in which pixels having the same luminance value are connected to each other in the image. That is, different labels are attached to different connected portions to distinguish the regions (connected regions) (see labeling 1 to 9 in FIG. 17).
[0075]
In addition, in figure fusion processing, small-area figures such as holes and dot-like figures included in the image are not essentially effective, but they may adversely affect the subsequent processing as noise. Need to be removed. Therefore, the microcomputer 11 removes noise components by expanding or contracting the original figure.
[0076]
(5) Step S305: Thinning process
This process is a process of narrowing down the line figure having a line width of 1 with respect to each connected area included in the obtained binary image without losing the connectivity. That is, the center line of a line figure is obtained by sequentially removing pixels in the width direction of a line figure having an arbitrary thickness.
[0077]
(6) Step S306: Shape determination process
Here, the area of the connection region is the number of pixels belonging to the connection region. The perimeter is the number of pixels located at the boundary around the connected area. However, the diagonal direction is corrected to √2 times the horizontal and vertical directions.
[0078]
In order to determine the shape of the image, the following coefficient e is used.
[0079]
e = (perimeter length) ^ 2 / (area)
e indicates a minimum value when the shape is circular, and indicates a larger value as the shape becomes more complicated.
[0080]
Since the person's face is considered to be almost circular, the e is compared with a predetermined value to determine whether or not the symmetric image is a person's face.
[0081]
Also, the area of the connected region is compared with a predetermined value to determine whether or not the face is a symmetric image person. Prior to shape determination, the area may be compared with a value within a predetermined range, and if it is outside the predetermined range, it may be determined that the image is not a person and the shape determination process may not be performed. In this way, the amount of calculation can be reduced and the AF time lag can be reduced.
[0082]
FIG. 15 is a flowchart showing the shape determination processing routine.
As shown in the figure, the microcomputer 11 determines whether or not there is an extraction area (step S601), and returns if there is no extraction area. If there is an extraction region, the area S of the extraction region is obtained to determine whether it is within a predetermined range (step S602). If the extraction area S is within the predetermined range, the process proceeds to step S603, and if not within the predetermined range, the process proceeds to step S607.
If the extraction area S is within the predetermined range in step S602, the microcomputer 11 calculates the shape determination value e and determines whether it is within the predetermined range (step S603). If it is within the predetermined range, it is determined that the person is a person (step S604).
[0083]
Thereafter, the microcomputer 11 determines whether or not the shape has been determined for all the extraction regions (step S605), and returns if it has been completed. On the other hand, if not completed, the next extraction area is set (step S606), the process proceeds to step S602, and the above process is repeatedly executed.
[0084]
On the other hand, if the extraction area S is not in the predetermined range in step S602 and the shape determination value e is not in the predetermined range in step S603, the microcomputer 11 determines that the subject is not a person (step S603). S607). Thereafter, similarly to the above, it is determined whether or not the shape is determined for all the extraction regions (step S605).
[0085]
Here, the person determination image in the present embodiment will be described with reference to FIGS.
[0086]
FIG. 16 is a diagram showing an example of a person determination image in the present embodiment, which is the image area corresponding to the AF area sensor 12 corresponding to the shooting screen. This image is the original image.
[0087]
FIG. 17 is a diagram illustrating an image after performing difference processing and binarization processing using the original image illustrated in FIG. 16.
As shown in the figure, only an edge portion (contour) is extracted. Moreover, the labeling process is performed to the extraction area (labeling 1-9).
[0088]
FIG. 18 is an explanatory diagram showing a person determination area and a plurality of distance measurement areas set in the person determination area in the present embodiment.
As shown in the figure, in this embodiment, an area 301 determined as a person's face is extracted, and a plurality of distance measurement areas in the person determination area 301 are set and distance measurement is performed.
[0089]
In step S209 in FIG. 7, the microcomputer 11 outputs a command to the AF area sensor 12 to set a monitor range in an area corresponding to the pixel in the person determination area 301.
[0090]
By performing the integration operation again based on the monitor signal in this area, it is possible to prevent the sensor data in the person determination area from being crushed due to the influence of the high-luminance background during backlighting. Sensor data can be obtained, and highly accurate ranging calculation can be performed.
[0091]
Next, as another method of shape determination, a method of storing a pattern of a main subject in advance and using it as a reference image and performing pattern matching processing with the reference image will be described with reference to FIGS. 19 to 21. To do.
[0092]
FIG. 19 is a diagram illustrating an example of a person determination image used in another method of shape determination, and is an image area corresponding image of the AF area sensor 12 corresponding to the shooting screen. Hereinafter, another method for shape determination will be described using this image as an original image.
[0093]
FIG. 20 is a diagram illustrating an image after performing difference processing and binarization processing using the original image illustrated in FIG. 19.
[0094]
As shown in the figure, only an edge portion (contour) is extracted. Moreover, the labeling process is performed to the extraction area (labeling 1-9). In the other method, the main subject pattern 300 stored in the EEPROM 11e in advance is used as a reference image (see FIG. 22), and a pattern matching process between the reference image 300 and the binarized image (see FIG. 22). A person image is extracted by performing a correlation operation.
[0095]
As shown in FIG. 22, the reference image 300 includes a plurality of similar patterns A, B,... Corresponding to changes in the subject distance, such as the focal length of the photographing lens (information from the zoom drive unit 22). It is selected according to conditions.
[0096]
Also, a plurality of patterns are prepared according to the camera posture, and the posture can be determined based on the output of the camera posture detection unit 120 and the pattern can be selected.
[0097]
Furthermore, not only a person pattern but also various object patterns are prepared, and when a person pattern cannot be detected, a pattern matching process selected according to a predetermined priority order is performed.
[0098]
FIG. 21 is an explanatory diagram showing a person determination area and a plurality of distance measurement areas set in the person determination area in the different method.
[0099]
As shown in the figure, in this other method, an area determined to be a person is extracted, and distance measurement is performed by setting a plurality of distance measuring areas in the person determination area 302. The distance measurement results of the plurality of distance measurement areas are combined into one distance measurement data by an average process, a closest selection process, and the like, and the lens is driven.
[0100]
In step S209 of FIG. 7, the microcomputer 11 outputs a command to the AF area sensor 12 so as to set a monitor range in an area corresponding to the pixel in the person determination area 302. Then, by performing the integration operation of the AF area sensor 12 again based on the monitor signal in this area, optimal sensor data is obtained for the person determination area 302, and highly accurate distance measurement calculation is possible.
[0101]
As described above, according to the distance measuring apparatus of the first embodiment, the outline of the main subject is detected, and the distance is focused on the inside of the outline within a wide range of distance measurement areas. Accordingly, it is possible to focus on the main subject accurately, and to provide a distance measuring device with a reduced time lag.
[0102]
Next, a second embodiment of the present invention will be described.
The configuration of the second embodiment is the same as that of the first embodiment as far as the configuration is shown in FIG. Therefore, only the differences will be mentioned here, and only the portions that differ in operation will be described.
[0103]
The distance measuring apparatus according to the second embodiment differs from the first embodiment only in the main subject detection routine, and other configurations and operations are the same as those in the first embodiment. Detailed description of is omitted.
[0104]
FIG. 24 is a flowchart showing a main subject detection routine in the distance measuring apparatus according to the second embodiment.
[0105]
As shown in the figure, the microcomputer 11 first performs pre-main subject detection (step S701), and roughly determines the position of the main subject from the sensor data of the AF area sensor 12 corresponding to the entire shooting screen. Next, the microcomputer 11 performs main main subject detection (step S702), newly sets a detection block based on the result of pre-main subject detection, and performs main subject main detection in detail within this block.
[0106]
Now, the pre-main subject detection (pre-detection) routine will be described in detail with reference to FIG.
[0107]
In this pre-main subject detection, the microcomputer 11 first performs a difference process on the sensor data in the entire area (step S801). Thereafter, a histogram relating to sensor data in the entire region is created, a threshold value is set, and binarization processing is performed (step S802). Further, a labeling process is performed on the binarization result (step S803), and a shape determination process is performed on the labeled connected area (step S804).
[0108]
Thereafter, the microcomputer 11 provisionally determines an extraction area from the shape determination processing result obtained in step S804 (step S805). In the second embodiment, when pre-detection is performed in order to shorten the processing time, the sensor data is thinned out for each predetermined pixel, or the average is taken to reduce the number of processing pixels.
[0109]
Next, the main subject main detection (main detection) routine will be described in detail with reference to FIG.
[0110]
In this main subject main detection, the microcomputer 11 first sets a detection block 305 including an extraction area based on the result of the pre-main subject detection process (step S901, see FIG. 27). Next, the microcomputer 11 performs a difference process on the detection block (step S902), creates a histogram based on the sensor data in the block, and determines a threshold value for each block (step S903).
[0111]
Thereafter, the microcomputer 11 performs binarization processing (step S904), labeling processing (step S905), and shape determination processing (step S906) for each block, and determines an extraction area (step S907).
[0112]
As described above, edge detection is performed on the entire sensor data of the pixel portion of the AF area sensor 12 corresponding to the shooting screen, and the main subject is roughly detected from the contour.
[0113]
Based on the result, a new detection block is set in the vicinity of the detection area, and the main subject is detected in detail. Therefore, the threshold can be set optimally, and a good binarized image can be obtained. The detection accuracy can be greatly improved.
[0114]
If the processing time is allowed to be slightly longer, the pixel area of the AF area sensor 12 corresponding to the shooting screen is divided into a plurality of blocks from the beginning, and main detection is performed for each block (see FIG. 26). May be. That is, histogram creation, threshold setting, binarization processing, and shape determination processing are performed for each block to extract a main subject area. At this time, the size of the divided block may be variable (large on the wide angle side and small on the telephoto side) according to conditions such as the focal length of the photographing lens (information from the zoom driving unit 22).
[0115]
Also, simultaneously with the light projecting operation to the subject by the strobe 20a and the strobe circuit unit 20, the integration operation by the AF area sensor 12 is performed, and the main subject is detected for the sensor data obtained by operating the steady light removal circuit 12c. Also good.
[0116]
Even when the brightness of the main subject and the background are very close and it is difficult to detect the edge (contour), the difference in the amount of reflected light due to the difference in distance occurs due to the light projection, so the edge (contour) can be detected accurately. It can be carried out.
[0117]
As described above, according to the distance measuring apparatus of the second embodiment, the contour of the main subject is detected and the distance is focused on the contour within the wide distance measuring area, which is influenced by the background. Thus, it is possible to focus on the main subject accurately, and to provide a distance measuring device with a reduced time lag.
[0118]
Next, a distance measuring apparatus according to a third embodiment of the present invention will be described.
FIG. 28 is a block diagram illustrating a main configuration of a camera including the distance measuring apparatus according to the third embodiment.
The camera employs an auto-focus method called a so-called super combination AF, and provides distance measurement information to the CPU 501 that controls the entire camera and receives the subject image and photoelectrically converts the subject image. A distance measuring unit 502, a light source 505a that emits strobe light toward the subject under the control of the CPU 501, a circuit 505 that controls the light emission of the light source 505a, and whether or not distance type information or main subject can be discriminated An audio signal generation circuit 507 that emits information as audio constitutes a main part.
[0119]
The distance measuring unit 502 includes area sensors 502a and 502b that photoelectrically convert object light from two light receiving lenses, and A / D convert the outputs of the area sensors 502a and 502b to convert the digital value of each pixel into a camera. An A / D conversion circuit 502c that outputs to the CPU 501 that is a control microcomputer, and a stationary light removal circuit 511 that removes stationary light from the outputs of the area sensors 502a and 502b under the control of the CPU 501. Yes.
[0120]
The stationary light removal circuit 511 is a circuit that removes a direct-current light signal that is steadily incident from the subject, and has an effect of using only the pulsed light from the light source 505a received by the area sensors 502a and 502b as an output signal. It is a circuit to do.
[0121]
When the light emission source 505a emits light by the light emission circuit 505 with the steady light removal circuit 511 operated under the control of the CPU 501, the area sensor 502a receiving the reflected signal light from the subject 520 is as shown in FIG. The image is tied. In FIG. 29, a black portion indicates a position where light is incident.
[0122]
The CPU 501 incorporates software (pattern discrimination program 501a) for analyzing the image pattern on the area sensor as described above. If the pattern determination program 501a determines that the subject image has a human shape, it can be considered as the main subject. Note that the pattern determination program performs the same processing as the “main subject detection” shown in FIG. 10 in the first embodiment.
[0123]
Next, the distance measuring operation by the distance measuring apparatus of the present embodiment will be described with reference to the flowchart shown in FIG.
[0124]
FIG. 30 is a flowchart showing a distance measurement routine in the distance measuring apparatus according to the third embodiment.
[0125]
Prior to ranging, the CPU 501 first activates a light source 505a for so-called pre-emission to irradiate auxiliary light, and activates a steady light removal circuit 511 to output steady light from outputs from the area sensors 502a and 502b. Is removed (step S1101). Only the reflected signal light pattern of the subject 520 is extracted as shown in FIG. 29 by this pre-light emission and steady light removal operation. The CPU 501 determines whether or not the subject image has a human shape based on the reflected signal light based on the pattern determination program 501a that is the internal program (step S1102).
[0126]
Next, the CPU 501 detects the position of the main subject by processing based on the pattern discrimination program 501a (step S1103). If the CPU 501 can determine that the subject image is a human figure, the CPU 501 determines that the subject is a main subject, and proceeds to step S1104. At this time, the CPU 501 makes such a determination based on whether the optical signal forming the pattern is weak or has sufficient contrast.
[0127]
In step S1104, the CPU 501 uses a so-called active method (a type in which signal light is projected from the camera side and measured using the reflected signal light) and a passive type (based on the image signal of the subject). The type of distance) is selected. Here, when the contrast of the image signal is weak, the distance measuring light (auxiliary light from the light emission source 505a) is irradiated again (step S1110), and the active autofocus by the reflected signal light is determined as the main subject position previously obtained, Specifically, a plurality of distance measuring areas are set inside the contour area of the main subject and the focus is performed (step S1111).
[0128]
Thereafter, the CPU 501 controls the audio signal generation circuit 507 to emit information indicating that the distance measurement is based on active autofocus and whether or not the main subject can be determined as a sound (step S1112), and then returns. The voice pattern corresponding to such information is referred to as (voice pattern 3).
[0129]
On the other hand, if it is determined in step S1104 that the reflected signal light from the subject is weak, the CPU 501 branches to step S1105 because passive autofocus is more suitable. In this step, a plurality of distance measurement areas are set in the contour region of the image signal of the main subject position that has already been obtained, and distance measurement is performed by a passive method that is used with priority.
[0130]
Thereafter, the CPU 501 controls the audio signal generation circuit 507 to emit information indicating that the distance measurement is based on passive autofocus and whether or not the main subject can be identified (step S1106), and then returns. The voice pattern corresponding to such information is referred to as (voice pattern 2).
[0131]
If the main subject cannot be detected in step S1103, the active method or passive method is selected in consideration of luminance information and the like, and then the center of the screen with a high probability of existence of the subject is intensively measured. The distance is set (step S1120).
[0132]
After that, the CPU 501 controls the audio signal generation circuit 507 to emit the information indicating that the distance measurement is based on the center distance measurement and whether or not the main subject can be identified as a sound (step S1121), and then returns. The voice pattern corresponding to such information is referred to as (voice pattern 1).
[0133]
As described above, the distance type information and the main subject can be discriminated for each distance measurement method, that is, each of the distance measurement methods of passive autofocus in step S1105, active autofocus in step S1111, and center distance measurement in step S1120. Information is emitted from the audio signal generation circuit 507.
[0134]
In addition, the present embodiment is characterized in that the main subject detection is performed using two methods instead of simply combining the active method and the passive method in a hybrid manner.
[0135]
As described above, according to the third embodiment, it is possible to provide a camera that is easy to understand for the user and can perform distance measurement with a sense of security.
[0136]
As described above, according to the distance measuring apparatus of each of the embodiments described above, the outline of the main subject is detected, and the distance measurement is focused on the inside of the outline within a wide range of distance measurement areas. Thus, it is possible to provide a distance measuring device that can accurately focus on the main subject without reducing the time lag.
[0137]
Although the distance measuring device of the present embodiment is an outside light type distance measuring device, the present invention is not limited to this and can be applied to a TTL distance measuring device, and the same effect can be obtained.
[0138]
[Appendix]
According to the embodiment of the present invention as described in detail above, the following configuration can be obtained. That is,
(1)
An integral area sensor for imaging in a two-dimensional area;
Integration control means for controlling the integration operation of the integration type area sensor;
A region extracting means for extracting a region including a main subject existing in the two-dimensional region based on an output of the integral area sensor;
Ranging calculation means for performing a ranging calculation on the area extracted by the area extracting means;
A comparison means for comparing the distance measurement result obtained by the distance calculation means with a predetermined determination condition;
If the comparison result by the comparison means does not meet a predetermined condition, the control means for changing the control parameter by the integration control means and then executing the integration operation of the integration type area sensor again;
A distance measuring device comprising:
[0139]
(2)
The region extracting means performs predetermined arithmetic processing on the image data output from the integral area sensor, separates object images included in the image data, and collates each separated data with a specific pattern. The distance measuring device according to (1), wherein whether or not the object image is a person image is determined, and information on a region determined to be a person image is output.
[0140]
(3)
The distance measuring apparatus according to (2), wherein the predetermined calculation process includes an edge extraction process and a binarization process.
[0141]
【The invention's effect】
  As described above, according to the present invention, even a ranging device having a wide ranging area can be used., Integral control according to the monitor signal of the corresponding pixel in the area of the main subject,It is possible to provide a distance measuring device that can perform an accurate autofocus process on a main subject without increasing a time lag and without being affected by the background.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a distance measuring apparatus according to a first embodiment of the present invention.
FIG. 2 is a flowchart showing a main routine of the microcomputer in the distance measuring apparatus according to the first embodiment.
FIG. 3 is an explanatory diagram showing a distance measuring optical system in the distance measuring apparatus according to the first embodiment.
FIG. 4 is a diagram for explaining a method of obtaining a subject distance based on the principle of triangulation.
FIG. 5 is a block diagram showing a configuration of an AF area sensor in the distance measuring apparatus according to the first embodiment.
FIG. 6 is a diagram illustrating a relationship between a shooting screen (wide and tele) and a distance measurement area in the distance measuring device according to the first embodiment.
FIG. 7 is a flowchart showing an autofocus (AF) routine in the distance measuring apparatus according to the first embodiment.
FIG. 8 is a flowchart showing a main subject detection operation in the distance measuring apparatus according to the first embodiment.
FIG. 9 is a diagram showing images before and after edge detection processing by difference processing in the distance measuring apparatus according to the first embodiment.
FIG. 10 is an explanatory diagram showing an example of a spatial filter table in difference processing in the distance measuring apparatus according to the first embodiment.
FIG. 11 is a flowchart showing binarization processing in the distance measuring apparatus according to the first embodiment.
FIG. 12 is a diagram for explaining a method of determining a threshold value by a mode method in the course of binarization processing in the distance measuring apparatus according to the first embodiment.
FIG. 13 is a diagram for explaining threshold setting processing in the distance measuring apparatus according to the first embodiment.
FIG. 14 is a flowchart showing threshold setting processing in the distance measuring apparatus according to the first embodiment.
FIG. 15 is a flowchart showing a shape determination processing routine in the distance measuring apparatus according to the first embodiment.
FIG. 16 is a diagram showing an example of a person determination image in the distance measuring apparatus according to the first embodiment.
17 is a diagram showing an image after performing difference processing and binarization processing using the original image shown in FIG. 16 in the distance measuring apparatus according to the first embodiment. FIG.
FIG. 18 is an explanatory diagram showing a person determination area and a plurality of distance measurement areas set in the person determination area in the distance measuring apparatus according to the first embodiment.
FIG. 19 is a diagram illustrating an example of a person determination image used in another shape determination method in the distance measuring apparatus according to the first embodiment.
20 is a diagram showing an image after performing difference processing and binarization processing using the original image shown in FIG. 19 in the distance measuring apparatus according to the first embodiment. FIG.
FIG. 21 is an explanatory diagram showing a person determination area and a plurality of distance measurement areas set in the person determination area in another shape determination method in the distance measuring apparatus according to the first embodiment.
FIG. 22 is an explanatory diagram showing an example of a pattern of a main subject in another method for shape determination, which is stored in an EEPROM in the distance measuring apparatus according to the first embodiment.
FIG. 23 is an explanatory diagram showing sensor data s (i, j) used in edge detection processing by difference processing in the distance measuring apparatus according to the first embodiment.
FIG. 24 is a flowchart showing a main subject detection routine in the distance measuring apparatus according to the second embodiment of the present invention.
FIG. 25 is a flowchart showing a pre-main subject detection (pre-detection) routine in the distance measuring apparatus according to the second embodiment.
FIG. 26 is a flowchart showing a main subject main detection (main detection) routine in the distance measuring apparatus according to the second embodiment;
FIG. 27 is an explanatory diagram showing an example of a new detection block in main subject main detection in the distance measuring apparatus according to the second embodiment.
FIG. 28 is a block diagram illustrating a main configuration of a camera including a distance measuring device according to a third embodiment of the present invention.
FIG. 29 is a diagram showing an example of an image formed on an area sensor in the distance measuring apparatus according to the third embodiment.
FIG. 30 is a flowchart showing a distance measurement routine in the distance measuring apparatus according to the third embodiment.
[Explanation of symbols]
11 ... Microcomputer
12 ... AF area sensor
12a: Image area
12b ... Processing circuit
12c ... Stationary light removal circuit
13: Focus lens drive unit
14 ... Focus lens
15 ... Focus lens encoder
16 ... Shutter drive section
17 ... 1RSW
18 ... 2RSW
19 ... Display section
20: Strobe circuit section
21 ... Film drive unit
22 ... Zoom lens drive unit
23. Metering section
120 ... Camera posture detection unit

Claims (2)

An imaging device having a plurality of images, imaging a subject image and outputting an imaging signal;
Contour extraction means for detecting a main subject based on the output of the image sensor and extracting the contour of the main subject;
Monitor selection means for setting an extraction range of a monitor signal indicating an integration amount of pixels of the image sensor;
Integration control means for controlling the integration operation of the imaging signal based on a monitor signal in a range set by the monitor selection means;
A distance measuring means for performing a distance measuring operation by performing a correlation operation on an imaging signal obtained by dividing the subject image into two images obtained by the integration control means by executing integration control of the image sensor;
Comprising
The monitor selection means corresponds to the extraction range of the monitor signal corresponding to the inside of the extracted outline of the main subject area so that an appropriate imaging signal for the main subject area extracted by the outline extraction means can be obtained. A distance measuring device characterized by being set in a pixel region of an image sensor . An auxiliary light means for irradiating the subject with auxiliary light;
A stationary light removing means for removing a stationary light component from the output of the imaging device;
Further comprising
The contour extracting unit extracts the contour of the subject based on the output of the imaging device that irradiates the subject with the auxiliary light unit and removes the stationary light component by the stationary light removing unit. The described distance measuring device.

Images