JP3895878B2 - Image data processing device - Google Patents

Description

となり、Ｒ、Ｇ及びＢの各成分が１：２：１の割合で合成された輝度データＹを得ることができる。本来、輝度信号は、ＮＴＳＣ方式の規格によれば、Ｒ、Ｇ及びＢの各成分を３：６：１の割合で合成して生成されるものであるが、これに近い割合で合成して生成したものであれば、実用上問題はない。
【００１５】
ガンマ補正回路１７は、再生画面上の発光輝度と視覚的感度との線形性のずれを補正するため、非線形変換を行う。アパーチャ回路１８は、輝度データＹに含まれる特定の周波数成分を強調してアパーチャデータを生成し、このアパーチャデータを輝度データＹに加算する。即ち、被写体画像の輪郭を強調するため、画像信号Ｉ１から画像データＤ１を得る際のサンプリング周波数の１／４の周波数成分を強調するように画像データＤに対してフィルタリング処理を施し、アパーチャデータを生成するように構成される。このようにして生成される輝度データＹは、色差データＵ、Ｖと共に外部の表示機器あるいは記録機器へ供給される。
【００１６】
【発明が解決しようとする課題】
ホワイトバランス制御回路１３において各原色データＰ[R]、Ｐ[G]、Ｐ[B]毎に設定されるゲインは、原色データＰ[R]、Ｐ[G]、Ｐ[B]を所定の期間にそれぞれ積分し、その積分値をそろえるようにして設定される。例えば、原色データＰ[R]、Ｐ[G]、Ｐ[B]をそれぞれ１画面単位で積分し、原色データＰ[G]の積分値Ｉ[G]に対して原色データＰ[R]、Ｐ[B]の積分値Ｉ[R]、Ｉ[B]が一致するように、原色データＰ[R]に対するゲインをＩ[G]／Ｉ[R]とし、原色データＰ[B]に対するゲインをＩ[G]／Ｉ[R]として設定するようにしている。また、モアレバランス制御回路１５において画像データＤ１の各成分毎に設定されるゲインについても、ホワイトバランス制御回路１３において設定されるゲインと同様に、画像データＤ１の各成分毎の所定期間の成分値を算出し、その積分値に基づいて設定するように構成される。これらの演算処理では、同等の演算処理でありながらも、扱うデータが異なっているため共通化が困難であり、回路規模縮小の障害となっている。
【００１７】
そこで本発明は、ホワイトバランス制御とモアレバランス制御とを共通化できるようにして、回路規模の縮小を可能にすることを目的とする。
【００１８】
【課題を解決するための手段】
本発明は、上述の課題を解決するために成されたもので、その特徴とするところは、光の三原色に対する補色を表す第１乃至第３の補色成分が所定の規則で繰り返される画像データを取り込み、輝度データ及び第１及び第２の色差データを生成する画像データ処理装置であって、上記画像データに対して、各補色成分毎に固有のゲインを与えて、各補色成分の所定期間内の平均レベルを整える色バランス制御回路と、バランス調整された上記画像データを成分別に分離して第１乃至第３の補色成分データを生成する分離回路と、上記第１乃至第３の色成分データの内の２つを互いに乗算して第１乃至第３の積を生成する乗算回路と、上記第１乃至第３の積に基づいて上記第１及び第２の色差データを生成する色差データ算出回路と、バランス調整された上記画像データを合成して輝度データを生成する輝度データ生成回路と、を備えたことにある。
【００１９】
本発明によれば、色バランスを調整した後、それぞれの色成分を互いに加算または減算することなく原色データが生成されるため、原色成分についても色バランスは保たれ、ホワイトバランスの調整が不要になる。
【００２０】
【発明の実施の形態】
図１は、本発明の画像データ処理装置の構成を示すブロック図であり、図２は、その動作を説明するタイミング図である。ここで、装置に入力される画像データＤは、例えば、光の三原色の補色（Ｙｅ：イエロー、Ｍｇ：マゼンタ、Ｃｙ：シアン）からなるストライプフィルタが装着されたイメージセンサから得られるものであり、３種類の補色成分が繰り返されるものとする。
【００２１】
色バランス制御回路２０は、ゲイン算出回路２１から供給されるゲインデータＧに基づいて、画像データＤ１の各色成分毎に固有のゲインを与え、所定期間内の各色成分の平均レベルを所定の範囲に収束させる。ゲイン算出回路２１は、画像データＤ１に対して色成分毎に積分処理を施し、その積分値がそれぞれ所定の適正範囲に納まるように各色成分毎のゲインを算出する。
【００２２】
分離回路２２は、色バランス制御回路２０から１画素単位で入力される画像データＤ２を色成分毎に振り分け、Ｙｅ、Ｍｇ、Ｃｙに対応する補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]を生成する。即ち、画像データＤ２では、各補色成分が、例えば、Ｙｅ、Ｍｇ、Ｃｙの順に１画素単位で繰り返されるため、一定の順序で３つに振り分けることで、各補色成分を表す補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]が生成される。第１〜第３のラッチ回路２３ａ〜２３ｃは、色分離回路２２に対して並列に接続され、色分離回路２２から入力される補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]をそれぞれラッチし、色分離回路２１から次のデータが出力されるまでラッチしたデータを繰り返し出力する。画像データＤ２を３つに分離して得られる補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]の場合、データ量が画像データＤの１／３であり、それぞれ３画素毎に１回更新されるため、連続する３画素は同じデータで表されることになる。
【００２３】
第１の乗算回路２４ａは、第１及び第２のラッチ回路２３ａ、２３ｂに接続され、各ラッチ回路２３ａ、２３ｂから得られる補色データＣ[Ye]、Ｃ[Mg]を乗算して、第１の積Ｐ[r]を生成する。そして、第２の乗算回路２４ｂは、第１及び第３のラッチ回路２３ａ、２３ｃに接続され、各ラッチ回路２３ａ、２３ｃから得られる補色データＣ[Y e]、Ｃ[Cy]を乗算して、第２の積Ｐ[g]を生成する。さらに、第３の乗算回路２４ｃは、第２及び第３のラッチ回路２３ｂ、２３ｃに接続され、各ラッチ回路２３ｂ、２３ｃから得られる補色データＣ[Mg]、Ｃ[Cy]を乗算して、第３の積Ｐ[b]を生成する。第１〜第３の開平回路２５ａ〜２５ｃは、それぞれ第１〜第３の乗算回路２４ａ〜２４ｃに接続され、各乗算回路２４ａ〜２４ｃから入力される第１〜第３の積Ｐ[r]、Ｐ[g]、Ｐ[b]の平方根を算出し、第１〜第３の根Ｐ[R]、Ｐ[G]、Ｐ[B]を出力する。これら第１〜第３の根Ｐ[R]、Ｐ[G]、Ｐ[B]は、そのまま、光の三原色に対応する３種類の原色データとして色差マトリクス回路２６へ供給される。
【００２４】
色差マトリクス回路２６は、ホワイトバランス制御回路２５から入力される原色データＰ[R]、Ｐ[B]、Ｐ[G]に対して色差データＵ、Ｖを生成する。ここで、色差マトリクス回路２６に入力される原色データＰ[R]、Ｐ[B]、Ｐ[G]は、色バランス制御回路２０において所定期間の平均値がそろえられた補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]に対して乗算及び開平処理を施したものであるため、ゲインのバランスは維持されている。即ち、補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]から原色データＰ[R]、Ｐ[B]、Ｐ[G]を生成する過程において加算処理あるいは減算処理を用いていないため、補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]においてそろえられたゲインのバランスは、原色データＰ[R]、Ｐ[B]、Ｐ[G]においても維持されている。従って、図１３に示すようなホワイトバランス制御回路１３は必要ない。
輝度演算回路２７は、色バランス制御回路２０から入力される画像データＤ２に含まれる４つの色成分を合成することにより、輝度データＹを生成する。ガンマ補正回路２８は、輝度データＹを非線形変換することにより、再生画面上の発光輝度と視覚的感度との線形性を保つようにしている。そして、アパーチャ回路２９は、輝度データＹに含まれる特定の周波数成分を強調することで、被写体画像の輪郭を強調する。これらの輝度演算回路２７からアパーチャ回路２９までの構成は、図１１に示す輝度演算回路１６からアパーチャ回路１８までの構成と一致する。
【００２５】
ここで、画像データの処理方法について説明する。
【００２６】
図３は、光の三原色とそれらの補色とに対応付けられたカラーフィルタの分光特性を示す図である。
【００２７】
一般に、光の三原色（Ｒ：レッド、Ｇ：グリーン、Ｂ：ブルー）に対応するカラーフィルタは、図３の波線に示すような分光特性を示す。Ｒフィルタは、波長の長い赤色の光に対応する部分で透過率が極大値を示し、赤色よりも波長の短い光に対しては透過率が低下する。逆に、Ｂフィルタは、波長の短い青色の光に対応する部分で透過率が極大値を示し、青色よりも波長の長い光に対しては透過率が低下する。そして、Ｇフィルタは、ＲフィルタとＧフィルタとの中間の特性を示す。
【００２８】
これに対して、Ｒ、Ｇ、Ｂの補色（Ｙｅ、Ｍｇ、Ｃｙ）に対応するフィルタは、図３の実線に示すような分光特性を示す。即ち、Ｒの補色となるＣｙフィルタは、Ｒフィルタと相反する特性を示し、赤色の光に対応する部分を除いた短波長側で透過率が高くなる。また、Ｂの補色となるＹｅフィルタは、Ｂフィルタと相反する特性を示し、青色の光に対応する部分を除いた長波長側で透過率が高くなる。そして、Ｇの補色となるＭｇフィルタは、緑色の光に対応する部分を除いて透過率が高くなる。
【００２９】
ここで、Ｙｅフィルタを透過した光に応じて得られたＹｅ成分と、Ｍｇフィルタを透過した光に応じて得られたＭｇ成分とを乗算すると、両成分に含まれているＲ成分が強調され、一方の成分にしか含まれないＢ成分及びＧ成分は減衰される。同様に、Ｍｇフィルタを透過した光に応じて得られたＭｇ成分と、Ｃｙフィルタを透過した光に応じて得られたＣｙ成分とを乗算することにより、両成分に含まれているＢ成分が強調され、一方の成分にしか含まれないＧ成分及びＲ成分は減衰される。また、Ｃｙフィルタを透過した光に応じて得られたＣｙ成分と、Ｙｅフィルタを透過した光に応じて得られたＹｅ成分とを乗算することにより、両成分に含まれているＧ成分が強調され、一方の成分にしか含まれないＲ成分及びＢ成分は減衰される。このようにして得られる各補色成分の積は、特定の原色成分に対応付けられるが、補色成分の乗算により、実際の原色成分に対しては二乗倍の値として表される。そこで、各積の平方根をとることにより、各原色成分を近似的に表すことができる。
【００３０】
図４は、ゲイン算出回路２１の一例を示すブロックずである。この図においては、Ｙｅ、Ｍｇ、Ｃｙの各成分が一定の規則で繰り返される画像信号Ｄ１に対応するものであり、Ｙｅ成分を基準にしてＭｇ成分及びＣｙ成分のゲインを決定する場合を示している。
【００３１】
ゲイン算出回路２１は、分離回路３１、第１〜第３の積分回路３２ａ〜３２ｃ及び第１、第２の除算回路３３ａ、３３ｂにより構成される。分離回路３１は、入力される画像データＤ１を色成分毎に分離し、第１〜第３の分離データＤ[Ye]、Ｄ[Mg]、Ｄ[Cy]を生成する。第１〜第３の積分回路３２ａ〜３２ｃは、それぞれ分離回路３１に接続され、第１〜第３の分離データＤ[Ye]、Ｄ[Mg]、Ｄ[Cy]を所定の期間、例えば、１垂直走査期間毎に積分し、第１〜第３の積分データＩ[Ye]、Ｉ[Mg]、Ｉ[Cy]を生成する。第１の除算回路３３ａは、第１の積分データＩ[Ye]を第２の積分データＩ[Mg]で除算し、Ｍｇ成分に対応する第１のゲインデータＧ[Mg]を生成する。そして、第２の除算回路３３ｂは、第１の積分データＩ[Ye]を第３の積分データＩ[Cy]で除算し、Ｃｙ成分に対応する第２のゲインデータＧ[Cy]を生成する。このようにして生成される第１、第２のゲインデータＧ[Mg]、Ｇ[Cy]は、画像データＤ１に含まれるＭｇ成分及びＣｙ成分の平均レベルをＹｅ成分にそろえる。
【００３２】
図５は、本発明の処理方法を採用するのに適したモザイク型のカラーフィルタの構成を示す平面図であり、図６は、そのカラーフィルタを装着したイメージセンサから得られた画像信号を各補色成分に分離する分離部の構成を示すブロック図である。この分離部は、図１に示す処理装置において、色分離回路２２及び第１〜第３のラッチ回路２３ａ〜２３ｃに置き換えて用いるものである。
【００３３】
カラーフィルタは、各セグメントがＹｅ、Ｍｇ、Ｃｙ及びＧの各成分に対応付けられている。このうち、Ｍｇ成分及びＧ成分が奇数行に交互に配置され、Ｙｅ成分及びＣｙ成分が偶数行に交互に配置される。ここで、Ｙｅ、Ｍｇ、Ｃｙの各成分は、色情報の生成に加えて、輝度情報の生成に用いられ、Ｇ成分は、輝度情報の生成のみに用いられる。このようなカラーフィルタを装着したイメージセンサでは、奇数行から読み出された画像信号においてＭｇ成分とＧ成分とが交互に連続し、偶数行から読み出された画像信号においてＹｅ成分とＣｙ成分とが交互に連続することになる。
【００３４】
分離部は、第１〜第３のラインメモリ４１ａ〜４１ｃ、フィルタ演算回路４２及びセレクタ回路４３より構成される。第１〜第３のラインメモリ４１ａ〜４１ｃは、直列に接続され、画像データＤを３行分記憶すると共に、各行の画像データＤ[L1]、Ｄ[L2]、Ｄ[L3]を同時に読み出してフィルタ演算回路４２へ供給できるできるように構成される。
【００３５】
フィルタ演算回路４２は、第１〜第３のラインメモリ４１ａ〜４１ｃから入力される画像データＤ[L1]、Ｄ[L2]、Ｄ[L3]を３列分保持し、３行×３列単位のフィルタ演算によって、カラーフィルタの各セグメントに対応する色成分データＣ[1]、Ｃ[2]、Ｃ[3]、Ｃ[4]を生成する。例えば、図５において、波線に示す９画素に対しては、中心の画素を目標画素とし、目標画素のＭｇ成分は、そのまま色成分データＣ[1]として出力される。そして、目標画素の左右に位置するＧ成分の２画素から、それらの平均値が色成分データＣ[2]として出力され、目標画素の上下に位置するＹｅ成分の２画素から、それらの平均値が色成分データＣ[3]として出力される。さらに、４角に位置するＣｙ成分の４画素は、それらの平均値が色成分データＣ[4]として出力される。このような色成分データＣ[1]、Ｃ[2]、Ｃ[3]、Ｃ[4]と各色成分Ｙｅ、Ｍｇ、Ｃｙ、Ｇとの対応関係は、目標画素がずれる毎に切り換えられる。
【００３６】
セレクタ回路４３は、色成分データＣ[1]、Ｃ[2]、Ｃ[3]、Ｃ[4]の内、常にＹｅ、Ｍｇ及びＣｙの各成分を表すものを選択し、補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]として出力する。この選択動作は、目標画素の色成分に応じて、４つのパターンが切り換えられる。従って、図１に示す第１〜第３のラッチ回路２３ａ〜２３ｃと同様に、全ての期間で欠落することのない３種類の補色データＣ[Ye]、Ｃ[Mg]、Ｃ[Cy]が出力され、図１に示す処理装置において第１〜第３の乗算回路２４ａ〜２４ｃにそれぞれ入力される。このような分離部によれば、モザイク型のカラーフィルタにおいても、本願発明の処理装置を適用することが可能になる。
【００３７】
図７は、図５に示すモザイクフィルタに対応して得られる画像データＤ１に対応するように構成したゲイン算出回路２１の一例を示すブロックずである。この図においては、Ｇ成分を基準にしてＹｅ成分、Ｍｇ成分及びＣｙ成分のゲインを決定する場合を示している。
【００３８】
ゲイン算出回路２１は、分離回路５１、第１〜第４の積分回路５２ａ〜５２ｄ、乗算回路５３及び第１〜第３の除算回路５４ａ〜５４ｃにより構成される。分離回路５１は、入力される画像データＤ１を色成分毎に分離し、第１〜第４の分離データ、Ｄ[Ye]、Ｄ[Mg]、Ｄ[Cy]、Ｄ[G]を生成する。第１〜第４の積分回路５２ａ〜５２ｄは、それぞれ分離回路５１に接続され、第１〜第４の分離データＤ[G]、Ｄ[Ye]、Ｄ[Mg]、Ｄ[Cy]を所定の期間、例えば、１垂直走査期間毎に積分し、第１〜第４の積分データＩ[G]、Ｉ[Ye]、Ｉ[Mg]、Ｉ[Cy]を生成する。乗算回路５３は、積分データＩ[G]を２倍する。この乗算回路５３による乗算処理は、補色系のフィルタに対して光の透過率が半分程度になる原色系（Ｇ）のフィルタの感度不足を補うためのものである。第１の除算回路５４ａは、２倍された第１の積分データ２Ｉ[G]を第２の積分データＩ[Ye]で除算し、Ｙｅ成分に対応する第１のゲインデータＧ[Ye]を生成する。第２の除算回路５４ｂは、２倍された第１の積分データ２Ｉ[G]を第３の積分データＩ[Mg]で除算し、Ｍｇ成分に対応する第２のゲインデータＧ[Mg]を生成する。そして、第３の除算回路５４ｂは、２倍された第１の積分データ２Ｉ[Ye]を第４の積分データＩ[Cy]で除算し、Ｃｙ成分に対応する第３のゲインデータＧ[Cy]を生成する。このようにして生成される第１〜第３のゲインデータＧ[Ye]、Ｇ[Mg]、Ｇ[Cy]は、画像データＤ１に含まれるＹｅ成分、Ｍｇ成分及びＣｙ成分の平均レベルをＧ成分の２倍にそろえる。
【００３９】
図８は、乗算回路２４ａ〜２４ｃで生成される第１〜第３の積Ｐ[r]、Ｐ[g]、Ｐ[b]から色差データＵ、Ｖを生成するための別の回路構成を示すブロック図である。この図に示す色差データ生成回路は、図１において、第１〜第３の開平回路２５ａ〜２５ｃ及び色差マトリクス回路２６に置き換えられるものである。
【００４０】
色差データ生成回路は、第１、第２の減算回路６１ａ、６１ｂ、第１、第２の絶対値回路６２ａ、６２ｂ、第１、第２の開平回路６３ａ、６３ｂ、第１、第２の符号付加回路６４ａ、６４ｂ及び色差マトリクス回路６５により構成される。第１の減算回路６１ａは、第１の積Ｐ[r]から第２の積Ｐ[g]を減算して第１の差Ｓ[r]を生成する。第２の減算回路６１ｂは、第３の積Ｐ[b]から第２の積Ｐ[g]を減算して第２の差Ｓ[b]を生成する。
【００４１】
第１及び第２の絶対値回路６２ａ、６２ｂは、それぞれ第１及び第２の減算回路６１ａ、６１ｂに接続され、第１及び第２の絶対値Ａ[r]、Ａ[b]を生成する。同時に、第１及び第２の絶対値回路６２ａ、６２ｂは、第１及び第２の差Ｓ[r]、Ｓ[b]の正／負を示す第１及び第２の符号Ｑ[r]、Ｑ[b]をそれぞれ取り出し、後述する符号付加回路６４ａ、６４ｂへそれぞれ供給する。第１及び第２の開平回路６３ａ、６３ｂは、それぞれ第１及び第２の絶対値回路６２ａ、６２ｂに接続され、各絶対値回路６２ａ、６２ｂから入力される第１及び第２の絶対値Ａ[r]、Ａ[b]の平方根を算出し、第１、第２の根Ｒ１[r]、Ｒ１[b]を生成する。そして、第１及び第２の符号付加回路６４ａ、６４ｂは、第１及び第２の開平回路６３ａ、６３ｂから入力される第１、第２の根Ｒ１[r]、Ｒ１[b]に、第１及び第２の絶対値回路６２ａ、６２ｂから供給される第１及び第２の符号Ｑ[r]、Ｑ[b]を付加し、極性を有する第１、第２の根Ｒ２[r]、Ｒ２[b]を生成する。
【００４２】
色差マトリクス回路６５は、第１及び第２の符号付加回路６４ａ、６４ｂに接続され、極性が付加された第１、第２の根Ｒ２[r]、Ｒ２[b]を合成して色差データＵ、Ｖを生成する。この色差マトリクス回路６４においては、例えば、第２の根Ｒ２[b]に所定の乗数を乗算し、その積を第１の根Ｒ２[r]に加算して色差データＶを生成し、第１の根Ｒ２[r]に所定の乗数を乗算し、その積を第２の根Ｒ２[g]に加算して色差データＵを生成するように構成される。この結果、図１に示す回路構成の場合を同様に、色差データＵ、Ｖを得ることができる。
【００４３】
以上の実施形態においては、イメージセンサとしてＣＣＤを例示しているが、イメージセンサとしては、その他にＭＯＳセンサを用いることも可能である。また、カラーフィルタの構成としては、少なくとも３種類の補色（Ｙｅ、Ｍｇ、Ｃｙ）を備えていれば適用可能であり、その他の色成分と組み合わせることも可能である。
【００４４】
【発明の効果】
本発明によれば、画像データの入力段階で色バランスを調整し、その後、そのバランスを維持した状態で輝度データ及び色差データを生成することができる。従って、色差データの生成過程でのホワイトバランス制御と輝度データの生成過程でのモアレバランス制御とを共通化することができ、ホワイトバランス制御回路と、そのゲインを算出するための回路を省略することが可能になる。
【図面の簡単な説明】
【図１】本発明の画像データ処理装置の構成を示すブロック図である。
【図２】本発明の画像データ処理装置の動作を説明するタイミング図である。
【図３】カラーフィルタの分光特性を示す図である。
【図４】ゲイン算出回路の構成の第１の例を示すブロック図である。
【図５】モザイク型カラーフィルタの一例を示す構成図である。
【図６】モザイクフィルタに対応した色分離部の構成を示すブロック図である。
【図７】ゲイン算出回路の構成の第２の例を示すブロック図である。
【図８】色差データ生成回路の別の構成例を示すブロック図である。
【図９】従来の固体撮像素子の構成を示すブロック図である。
【図１０】モザイク型カラーフィルタの一例を示す構成図である。
【図１１】従来の画像データ処理装置の構成を示すブロック図である。
【図１２】従来の画像データ処理装置の動作を説明するタイミング図である。
【符号の説明】
１ ＣＣＤイメージセンサ
２ アナログ処理回路
３ Ａ／Ｄ変換回路
４ デジタル処理回路
５ 駆動回路
６ タイミング制御回路
１１ 色分離回路
１２ 色演算回路
１３ ホワイトバランス制御回路
１４ 色差マトリクス回路
１５ モアレバランス制御回路
１６、２７ 輝度演算回路
１７、２８ ガンマ補正回路
１８、２９ アパーチャ回路
２０ 色バランス制御回路
２１ ゲイン算出回路
２２ 分離回路
２３ａ〜２３ｃ ラッチ回路
２４ａ〜２４ｃ 乗算回路
２５ａ〜２５ｃ、６３ａ、６３ｂ 開平回路
２６、６５ 色差マトリクス回路
３１、５３ 分離回路
３２ａ〜３２ｃ、５２ａ〜５２ｄ 積分回路
３３ａ、３３ｂ、５４ａ〜５４ｃ 除算回路
４１ａ〜４１ｃ ラインメモリ
４２ フィルタ演算回路
４３ セレクタ回路[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image data processing apparatus that generates color difference data and luminance data from image data including complementary color components.
[0002]
[Prior art]
FIG. 9 is a block diagram illustrating a configuration of an imaging apparatus using a CCD image sensor, and FIG. 10 is a plan view illustrating an example of a mosaic type color filter mounted on the CCD image sensor.
[0003]
The CCD image sensor 1 has a plurality of light receiving pixels, a plurality of vertical shift registers, and usually one horizontal shift register. The plurality of light receiving pixels are arranged in a matrix at regular intervals on the light receiving surface, and generate and store information charges corresponding to the received subject images. The plurality of vertical shift registers are arranged corresponding to the columns of the light receiving pixels, and sequentially transfer the information charges accumulated in the light receiving pixels in the vertical direction. The horizontal shift register is arranged on the output side of the vertical shift register, receives information charges transferred from a plurality of vertical shift registers, and transfers and outputs the information charges in units of one row. As a result, an image signal I0 that changes the voltage value according to the amount of information charge accumulated in each light receiving pixel is output.
[0004]
The analog processing circuit 2 performs processing such as sample hold and level clamp on the image signal I0 input from the CCD 1 to generate an image signal I1 according to a predetermined format. For example, in the sample and hold process, only the signal level is extracted from the image signal I 0 in which the reset level and the signal level are alternately repeated in synchronization with the output operation of the CCD 1. In the level clamping process, the black reference level set at the end of the horizontal scanning period of the image signal I0 is clamped to a predetermined level every horizontal scanning period. The A / D conversion circuit 3 quantizes the image signal I1 input from the analog processing circuit 2 at a timing according to the operation of the analog processing circuit 2, that is, the output operation of the CCD1, and digitally converts information corresponding to each light receiving pixel of the CCD1. Image data D1 represented by a value is generated.
[0005]
The digital processing circuit 4 performs processing such as color separation and matrix calculation on the image data D1 input from the A / D conversion circuit 3 to generate luminance data Y and color difference data U and V. For example, in the color separation process, the image data D1 is sorted according to the color arrangement of the color filter mounted on each light receiving surface of the CCD 1, and a plurality of color component data is generated. In the matrix calculation process, primary color data corresponding to the three primary colors of light is generated from each distributed color component data, and color difference data is generated by combining the primary color data at a predetermined ratio.
[0006]
The drive circuit 5 supplies multiphase drive clocks to the shift registers of the CCD 1 in response to various timing signals from the timing control circuit 6 described later. For example, a four-phase vertical transfer clock φv is supplied to the vertical shift register, and a two-phase horizontal transfer clock φh is supplied to the horizontal shift register. The timing control circuit 6 generates a vertical timing signal for determining the vertical scanning timing of the CCD 1 and a horizontal timing signal for determining the horizontal scanning timing in accordance with a reference clock having a fixed period, and supplies the generated timing to the driving circuit 5. At the same time, a timing clock CT for synchronizing the operation of each circuit with the output operation of the CCD 1 is supplied to the analog processing circuit 2, the A / D conversion circuit 3, and the digital processing circuit 4.
[0007]
By the way, when performing color imaging, a color filter for color separation is mounted on the light receiving surface in order to associate each light receiving pixel of the CCD 1 with a predetermined color component. Examples of the color filter include a stripe type in which a plurality of segments connected in the vertical direction are arranged, and a mosaic type in which a plurality of segments are arranged in association with each light receiving pixel. For example, as shown in FIG. 9, the mosaic color filter is divided into a plurality of segments corresponding to each pixel of the light receiving unit of the CCD 1, and each segment includes Ye (yellow), Cy (cyan), W ( White) and G (green) color components are assigned periodically. Here, the W and G components are alternately arranged in odd rows, and the Ye and Cy components are alternately arranged in even rows. In the image signal obtained from the CCD 1 equipped with such a color filter, the W and G components are repeated in the reading of the odd rows, and the Ye and Cy components are repeated in the reading of the even rows.
[0008]
FIG. 11 is a block diagram showing the configuration of the digital signal processing unit 4 as the video signal processing device, and FIG. 12 is a timing diagram for explaining the operation thereof. This figure corresponds to the case where the mosaic type color filter shown in FIG.
[0009]
The color separation circuit 11 sorts the image data D1 in which each color component is continuous in the order corresponding to the arrangement of each segment of the color filter by color component data C [Ye], C [Cy], C [G]. , C [W]. As shown in FIG. 10, the image data D1 input from the A / D conversion circuit 3 has the G component and the W component alternately alternately in the odd row (ODD) read operation, and the even row (EVEN). In reading, the Ye component and the Cy component are alternately continued. Therefore, the color separation circuit 11 holds the image data D1 for at least one row, so that all the color component data C [Ye], C [Cy], C [G], C [ W] output is enabled. That is, when an odd-numbered row is being read out, the image data D1 in that row is sorted and the color component data C [G] and C [W] are output, and at the same time, the image data D1 in the previous row is sorted out. The color component data C [Ye] and C [Cy] are output. Further, the color component data C [Ye], C [Cy], C [G], and C [W] that are intermittently generated by distributing the serially output image data D1 are output twice in succession. Is interpolated.
[0010]
The color arithmetic circuit 12 applies, for example, the color component data C [Ye], C [Cy], C [G], and C [W] input from the color separation circuit 11, for example:
Ye-G = R
Cy-G = B
G = G
The primary color data P [R], P [G], and P [B] corresponding to the three primary colors of light (R: red, G: green, B: blue) are generated.
[0011]
The white balance control circuit 13 gives a unique gain set for each color to the primary color data P [R], P [G], and P [B] input from the color arithmetic circuit 12, thereby making each color Adjust the balance. That is, the white balance control circuit 13 individually compensates the gains of the primary color data P [R], P [G], and P [B] for each color in order to compensate for the difference in sensitivity for each color component generated in the light receiving pixel of the CCD. By setting to, the color reproducibility of the playback screen is improved.
[0012]
The color difference matrix circuit 14 generates color difference data U and V for the primary color data P [R], P [G], and P [B] input from the white balance control circuit 13. That is, luminance information is generated by combining the primary color data P [R], P [G], and P [B] at a ratio of 3: 6: 1, and this luminance information is converted into the primary color data corresponding to the B component. Color difference data U is generated by subtracting from P [B], and color difference data V is generated by subtracting from primary color data P [R] corresponding to the R component.
[0013]
The moiré balance control circuit 15 sets a specific gain for each component so as to align the average level for each color component included in the image data D1 input to the color separation circuit 11. That is, there is a variation in the light transmittance for each color component of the color filter, and due to this, the variation in the average level for each color component is included in the image data. Thus, a unique gain is given to each component. Thereby, the image data D2 in which the average level for each color component is aligned is generated.
[0014]
The luminance calculation circuit 16 generates luminance data Y by synthesizing four color components included in the image data D2 input from the moire balance control circuit 15. That is, if each component of Ye, Cy, G, W is synthesized,

Thus, luminance data Y in which R, G, and B components are combined at a ratio of 1: 2: 1 can be obtained. Originally, according to the NTSC standard, the luminance signal is generated by combining R, G, and B components at a ratio of 3: 6: 1. If generated, there is no practical problem.
[0015]
The gamma correction circuit 17 performs non-linear conversion in order to correct the deviation in linearity between the light emission luminance on the reproduction screen and the visual sensitivity. The aperture circuit 18 emphasizes a specific frequency component included in the luminance data Y, generates aperture data, and adds the aperture data to the luminance data Y. That is, in order to enhance the contour of the subject image, the image data D is filtered so as to enhance the frequency component of 1/4 of the sampling frequency when the image data D1 is obtained from the image signal I1, and the aperture data is obtained. Configured to generate. The luminance data Y generated in this way is supplied to an external display device or recording device together with the color difference data U and V.
[0016]
[Problems to be solved by the invention]
The gain set for each primary color data P [R], P [G], P [B] in the white balance control circuit 13 is a predetermined value for the primary color data P [R], P [G], P [B]. It is set to integrate each period and align the integral values. For example, the primary color data P [R], P [G], and P [B] are each integrated in units of one screen, and the primary color data P [R], P [G] is integrated with the integrated value I [G] of the primary color data P [G]. The gain for the primary color data P [R] is set to I [G] / I [R] so that the integral values I [R] and I [B] of P [B] match, and the gain for the primary color data P [B]. Is set as I [G] / I [R]. Further, the gain set for each component of the image data D1 in the moire balance control circuit 15 is also the component value of the predetermined period for each component of the image data D1, as is the gain set in the white balance control circuit 13. And is set based on the integral value. In these arithmetic processes, although the arithmetic processes are equivalent, the data to be handled is different, so that it is difficult to share them, which is an obstacle to circuit scale reduction.
[0017]
Accordingly, an object of the present invention is to make it possible to reduce the circuit scale by allowing the white balance control and the moire balance control to be shared.
[0018]
[Means for Solving the Problems]
The present invention has been made in order to solve the above-mentioned problems, and is characterized in that image data in which first to third complementary color components representing complementary colors for the three primary colors of light are repeated according to a predetermined rule. An image data processing apparatus that captures, generates luminance data, and first and second color difference data, and provides the image data with a unique gain for each complementary color component, and within a predetermined period of each complementary color component A color balance control circuit that adjusts the average level of the image, a separation circuit that separates the balance-adjusted image data into components to generate first to third complementary color component data, and the first to third color component data. A multiplication circuit that multiplies two of them to generate first to third products, and color difference data calculation that generates the first and second color difference data based on the first to third products Circuit and balance In that and a luminance data generating circuit for generating a luminance data by combining the integer is the image data.
[0019]
According to the present invention, after the color balance is adjusted, the primary color data is generated without adding or subtracting the respective color components to each other. Therefore, the color balance is also maintained for the primary color components, and the white balance adjustment is unnecessary. Become.
[0020]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a block diagram showing the configuration of the image data processing apparatus of the present invention, and FIG. 2 is a timing chart for explaining the operation thereof. Here, the image data D input to the apparatus is obtained from, for example, an image sensor provided with a stripe filter composed of complementary colors (Ye: yellow, Mg: magenta, Cy: cyan) of the three primary colors of light. It is assumed that three types of complementary color components are repeated.
[0021]
The color balance control circuit 20 gives a unique gain for each color component of the image data D1 based on the gain data G supplied from the gain calculation circuit 21, and sets the average level of each color component within a predetermined range within a predetermined range. Converge. The gain calculation circuit 21 performs an integration process for each color component on the image data D1, and calculates a gain for each color component so that the integration value falls within a predetermined appropriate range.
[0022]
The separation circuit 22 distributes the image data D2 input from the color balance control circuit 20 in units of pixels for each color component, and complementary color data C [Ye], C [Mg], C [corresponding to Ye, Mg, Cy. Cy] is generated. That is, in the image data D2, each complementary color component is repeated, for example, in units of one pixel in the order of Ye, Mg, and Cy. Therefore, the complementary color data C [Ye representing each complementary color component is distributed by dividing into three in a fixed order. ], C [Mg], C [Cy] are generated. The first to third latch circuits 23a to 23c are connected in parallel to the color separation circuit 22 and receive complementary color data C [Ye], C [Mg], and C [Cy] input from the color separation circuit 22. Each latch is latched, and the latched data is repeatedly output until the next data is output from the color separation circuit 21. In the case of complementary color data C [Ye], C [Mg], and C [Cy] obtained by separating the image data D2 into three, the data amount is 1/3 of the image data D, and 1 for every 3 pixels. Since it is updated twice, three consecutive pixels are represented by the same data.
[0023]
The first multiplication circuit 24a is connected to the first and second latch circuits 23a and 23b, and multiplies the complementary color data C [Ye] and C [Mg] obtained from the latch circuits 23a and 23b to obtain the first To generate a product P [r]. The second multiplication circuit 24b is connected to the first and third latch circuits 23a and 23c, and multiplies the complementary color data C [Y e] and C [Cy] obtained from the latch circuits 23a and 23c. , The second product P [g] is generated. Further, the third multiplication circuit 24c is connected to the second and third latch circuits 23b and 23c, and multiplies the complementary color data C [Mg] and C [Cy] obtained from the latch circuits 23b and 23c. A third product P [b] is generated. The first to third square root circuits 25a to 25c are connected to the first to third multiplication circuits 24a to 24c, respectively, and the first to third products P [r] inputted from the multiplication circuits 24a to 24c are respectively. , P [g] and P [b] are calculated, and the first to third roots P [R], P [G] and P [B] are output. These first to third roots P [R], P [G], and P [B] are supplied as they are to the color difference matrix circuit 26 as three types of primary color data corresponding to the three primary colors of light.
[0024]
The color difference matrix circuit 26 generates color difference data U and V for the primary color data P [R], P [B], and P [G] input from the white balance control circuit 25. Here, the primary color data P [R], P [B], and P [G] input to the color difference matrix circuit 26 are complementary color data C [Ye] in which the average values for a predetermined period are aligned in the color balance control circuit 20. , C [Mg], and C [Cy] are multiplied and square rooted, so that the gain balance is maintained. That is, since no addition process or subtraction process is used in the process of generating the primary color data P [R], P [B], P [G] from the complementary color data C [Ye], C [Mg], C [Cy]. The balance of gains arranged in the complementary color data C [Ye], C [Mg], and C [Cy] is also maintained in the primary color data P [R], P [B], and P [G]. Therefore, the white balance control circuit 13 as shown in FIG. 13 is not necessary.
The luminance calculation circuit 27 generates luminance data Y by synthesizing four color components included in the image data D2 input from the color balance control circuit 20. The gamma correction circuit 28 performs linear conversion of the luminance data Y so as to maintain the linearity between the light emission luminance on the reproduction screen and the visual sensitivity. Then, the aperture circuit 29 enhances the contour of the subject image by enhancing a specific frequency component included in the luminance data Y. The configuration from the luminance calculation circuit 27 to the aperture circuit 29 is the same as the configuration from the luminance calculation circuit 16 to the aperture circuit 18 shown in FIG.
[0025]
Here, a method of processing image data will be described.
[0026]
FIG. 3 is a diagram showing spectral characteristics of color filters associated with the three primary colors of light and their complementary colors.
[0027]
In general, color filters corresponding to the three primary colors of light (R: red, G: green, B: blue) exhibit spectral characteristics as shown by the wavy lines in FIG. The R filter has a maximum transmittance at a portion corresponding to red light having a long wavelength, and the transmittance decreases for light having a shorter wavelength than red. Conversely, the B filter has a maximum transmittance at a portion corresponding to blue light having a short wavelength, and the transmittance decreases for light having a longer wavelength than blue. The G filter exhibits an intermediate characteristic between the R filter and the G filter.
[0028]
On the other hand, the filters corresponding to the complementary colors (Ye, Mg, Cy) of R, G, B exhibit spectral characteristics as shown by the solid line in FIG. That is, the Cy filter which is a complementary color of R exhibits a characteristic opposite to that of the R filter, and has a high transmittance on the short wavelength side excluding a portion corresponding to red light. In addition, the Ye filter that is a complementary color of B exhibits characteristics that are contrary to those of the B filter, and has a high transmittance on the long wavelength side excluding a portion corresponding to blue light. The Mg filter which is a complementary color of G has a high transmittance except for a portion corresponding to green light.
[0029]
Here, when the Ye component obtained according to the light transmitted through the Ye filter and the Mg component obtained according to the light transmitted through the Mg filter are multiplied, the R component contained in both components is emphasized. The B component and the G component contained only in one component are attenuated. Similarly, by multiplying the Mg component obtained according to the light transmitted through the Mg filter and the Cy component obtained according to the light transmitted through the Cy filter, the B component contained in both components is obtained. The G component and the R component which are emphasized and included in only one component are attenuated. Further, by multiplying the Cy component obtained according to the light transmitted through the Cy filter and the Ye component obtained according to the light transmitted through the Ye filter, the G component contained in both components is enhanced. The R component and the B component contained in only one component are attenuated. The product of the respective complementary color components obtained in this way is associated with a specific primary color component, but is expressed as a squared value for the actual primary color component by multiplication of the complementary color component. Therefore, each primary color component can be approximately expressed by taking the square root of each product.
[0030]
FIG. 4 is a block diagram illustrating an example of the gain calculation circuit 21. This figure shows a case where the components of Ye, Mg, and Cy correspond to the image signal D1 that is repeated according to a certain rule, and the gains of the Mg component and the Cy component are determined based on the Ye component. Yes.
[0031]
The gain calculation circuit 21 includes a separation circuit 31, first to third integration circuits 32a to 32c, and first and second division circuits 33a and 33b. The separation circuit 31 separates the input image data D1 for each color component, and generates first to third separation data D [Ye], D [Mg], and D [Cy]. The first to third integration circuits 32a to 32c are respectively connected to the separation circuit 31, and the first to third separation data D [Ye], D [Mg], and D [Cy] are given for a predetermined period, for example, Integration is performed every vertical scanning period to generate first to third integration data I [Ye], I [Mg], and I [Cy]. The first division circuit 33a divides the first integration data I [Ye] by the second integration data I [Mg] to generate first gain data G [Mg] corresponding to the Mg component. Then, the second division circuit 33b divides the first integration data I [Ye] by the third integration data I [Cy] to generate second gain data G [Cy] corresponding to the Cy component. . The first and second gain data G [Mg] and G [Cy] generated in this way align the average level of the Mg component and the Cy component included in the image data D1 with the Ye component.
[0032]
FIG. 5 is a plan view showing a configuration of a mosaic type color filter suitable for adopting the processing method of the present invention. FIG. 6 shows image signals obtained from an image sensor equipped with the color filter. It is a block diagram which shows the structure of the separation part which isolate | separates into a complementary color component. This separation unit is used in place of the color separation circuit 22 and the first to third latch circuits 23a to 23c in the processing apparatus shown in FIG.
[0033]
In the color filter, each segment is associated with each component of Ye, Mg, Cy, and G. Among these, the Mg component and the G component are alternately arranged in odd rows, and the Ye component and the Cy component are alternately arranged in even rows. Here, each component of Ye, Mg, and Cy is used for generation of luminance information in addition to generation of color information, and the G component is used only for generation of luminance information. In an image sensor equipped with such a color filter, the Mg component and the G component continue alternately in the image signal read from the odd rows, and the Ye component and the Cy component in the image signals read from the even rows. Will continue alternately.
[0034]
The separation unit includes first to third line memories 41 a to 41 c, a filter operation circuit 42, and a selector circuit 43. The first to third line memories 41a to 41c are connected in series, store the image data D for three rows, and simultaneously read the image data D [L1], D [L2], and D [L3] of each row. Thus, the filter operation circuit 42 can be supplied.
[0035]
The filter operation circuit 42 holds three columns of image data D [L1], D [L2], D [L3] input from the first to third line memories 41a to 41c, and has a unit of 3 rows × 3 columns. By this filter calculation, color component data C [1], C [2], C [3], C [4] corresponding to each segment of the color filter are generated. For example, in FIG. 5, for the nine pixels indicated by the wavy lines, the central pixel is set as the target pixel, and the Mg component of the target pixel is output as color component data C [1] as it is. Then, the average value of the two G components positioned on the left and right of the target pixel is output as color component data C [2], and the average value of the two pixels of the Ye component positioned above and below the target pixel is output. Is output as color component data C [3]. Furthermore, the average value of the four Cy component pixels located at the four corners is output as color component data C [4]. The correspondence between the color component data C [1], C [2], C [3], C [4] and the color components Ye, Mg, Cy, G is switched every time the target pixel is shifted.
[0036]
The selector circuit 43 always selects one of the color component data C [1], C [2], C [3], and C [4] that represents each component of Ye, Mg, and Cy, and the complementary color data C [ Output as Ye], C [Mg], C [Cy]. In this selection operation, four patterns are switched according to the color component of the target pixel. Accordingly, as in the first to third latch circuits 23a to 23c shown in FIG. 1, three types of complementary color data C [Ye], C [Mg], and C [Cy] that are not lost in all periods are stored. 1 is input to the first to third multiplication circuits 24a to 24c in the processing apparatus shown in FIG. According to such a separation unit, the processing apparatus of the present invention can be applied even to a mosaic type color filter.
[0037]
FIG. 7 is a block diagram showing an example of the gain calculation circuit 21 configured to correspond to the image data D1 obtained corresponding to the mosaic filter shown in FIG. This figure shows a case where the gains of the Ye component, the Mg component, and the Cy component are determined based on the G component.
[0038]
The gain calculation circuit 21 includes a separation circuit 51, first to fourth integration circuits 52a to 52d, a multiplication circuit 53, and first to third division circuits 54a to 54c. The separation circuit 51 separates input image data D1 for each color component, and generates first to fourth separation data, D [Ye], D [Mg], D [Cy], and D [G]. . The first to fourth integration circuits 52a to 52d are respectively connected to the separation circuit 51, and the first to fourth separation data D [G], D [Ye], D [Mg], and D [Cy] are predetermined. For example, the first to fourth integration data I [G], I [Ye], I [Mg], and I [Cy] are generated by integrating for each vertical scanning period. The multiplier circuit 53 doubles the integral data I [G]. The multiplication processing by the multiplication circuit 53 is for compensating for the lack of sensitivity of the primary color (G) filter, which has a light transmittance of about half that of the complementary color filter. The first division circuit 54a divides the doubled first integration data 2I [G] by the second integration data I [Ye], and obtains the first gain data G [Ye] corresponding to the Ye component. Generate. The second division circuit 54b divides the doubled first integral data 2I [G] by the third integral data I [Mg], and obtains the second gain data G [Mg] corresponding to the Mg component. Generate. Then, the third division circuit 54b divides the doubled first integration data 2I [Ye] by the fourth integration data I [Cy] to obtain third gain data G [Cy corresponding to the Cy component. ] Is generated. The first to third gain data G [Ye], G [Mg], and G [Cy] generated in this way indicate the average levels of the Ye component, Mg component, and Cy component included in the image data D1 as G. Align to twice the ingredients.
[0039]
FIG. 8 shows another circuit configuration for generating the color difference data U and V from the first to third products P [r], P [g] and P [b] generated by the multiplication circuits 24a to 24c. FIG. The color difference data generation circuit shown in this figure is replaced with the first to third square root circuits 25a to 25c and the color difference matrix circuit 26 in FIG.
[0040]
The color difference data generation circuit includes first and second subtraction circuits 61a and 61b, first and second absolute value circuits 62a and 62b, first and second square root circuits 63a and 63b, and first and second codes. The circuit includes an additional circuit 64 a and 64 b and a color difference matrix circuit 65. The first subtraction circuit 61a subtracts the second product P [g] from the first product P [r] to generate a first difference S [r]. The second subtraction circuit 61b subtracts the second product P [g] from the third product P [b] to generate a second difference S [b].
[0041]
The first and second absolute value circuits 62a and 62b are connected to the first and second subtraction circuits 61a and 61b, respectively, and generate the first and second absolute values A [r] and A [b]. . At the same time, the first and second absolute value circuits 62a and 62b have first and second signs Q [r] indicating positive / negative of the first and second differences S [r] and S [b], respectively. Q [b] is taken out and supplied to sign adding circuits 64a and 64b described later. The first and second square root circuits 63a and 63b are connected to the first and second absolute value circuits 62a and 62b, respectively, and the first and second absolute values A inputted from the respective absolute value circuits 62a and 62b. The square roots of [r] and A [b] are calculated, and first and second roots R1 [r] and R1 [b] are generated. The first and second sign addition circuits 64a and 64b are connected to the first and second roots R1 [r] and R1 [b] input from the first and second square root circuits 63a and 63b, respectively. First and second roots R2 [r] having polarity by adding first and second codes Q [r] and Q [b] supplied from the first and second absolute value circuits 62a and 62b, R2 [b] is generated.
[0042]
The chrominance matrix circuit 65 is connected to the first and second sign addition circuits 64a and 64b, and synthesizes the first and second roots R2 [r] and R2 [b] to which the polarities are added, thereby obtaining the color difference data U. , V is generated. In the color difference matrix circuit 64, for example, the second root R2 [b] is multiplied by a predetermined multiplier, and the product is added to the first root R2 [r] to generate the color difference data V. The color difference data U is generated by multiplying the root R2 [r] by a predetermined multiplier and adding the product to the second root R2 [g]. As a result, the color difference data U and V can be obtained similarly in the case of the circuit configuration shown in FIG.
[0043]
In the above embodiment, the CCD is exemplified as the image sensor, but it is also possible to use a MOS sensor as the image sensor. The configuration of the color filter is applicable as long as it has at least three types of complementary colors (Ye, Mg, Cy), and can be combined with other color components.
[0044]
【The invention's effect】
According to the present invention, it is possible to adjust the color balance at the image data input stage, and then generate the luminance data and the color difference data while maintaining the balance. Therefore, the white balance control in the color difference data generation process and the moire balance control in the luminance data generation process can be shared, and the white balance control circuit and the circuit for calculating the gain thereof are omitted. Is possible.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of an image data processing apparatus of the present invention.
FIG. 2 is a timing chart for explaining the operation of the image data processing apparatus of the present invention.
FIG. 3 is a diagram illustrating spectral characteristics of a color filter.
FIG. 4 is a block diagram illustrating a first example of a configuration of a gain calculation circuit.
FIG. 5 is a configuration diagram illustrating an example of a mosaic type color filter.
FIG. 6 is a block diagram illustrating a configuration of a color separation unit corresponding to a mosaic filter.
FIG. 7 is a block diagram showing a second example of the configuration of the gain calculation circuit.
FIG. 8 is a block diagram illustrating another configuration example of the color difference data generation circuit.
FIG. 9 is a block diagram illustrating a configuration of a conventional solid-state imaging device.
FIG. 10 is a configuration diagram illustrating an example of a mosaic type color filter.
FIG. 11 is a block diagram illustrating a configuration of a conventional image data processing apparatus.
FIG. 12 is a timing chart for explaining the operation of a conventional image data processing apparatus.
[Explanation of symbols]
1 CCD image sensor
2 Analog processing circuit
3 A / D conversion circuit
4 Digital processing circuit
5 Drive circuit
6 Timing control circuit
11 Color separation circuit
12 color arithmetic circuit
13 White balance control circuit
14 Color difference matrix circuit
15 Moire balance control circuit
16, 27 Luminance calculation circuit
17, 28 Gamma correction circuit
18, 29 Aperture circuit
20 color balance control circuit
21 Gain calculation circuit
22 Separation circuit
23a to 23c latch circuit
24a-24c multiplication circuit
25a-25c, 63a, 63b Square root circuit
26, 65 Color difference matrix circuit
31, 53 Separation circuit
32a to 32c, 52a to 52d Integration circuit
33a, 33b, 54a to 54c Dividing circuit
41a-41c line memory
42 Filter operation circuit
43 Selector circuit

Claims (3)

An image data processing apparatus that captures image data in which first to third complementary color components representing complementary colors for three primary colors of light are repeated according to a predetermined rule, and generates luminance data and first and second color difference data, A color balance control circuit for adjusting the average level of each complementary color component within a predetermined period by giving a unique gain for each complementary color component to the image data, and separating the image data after the balance adjustment for each component. A separation circuit for generating first to third complementary color component data; a multiplication circuit for multiplying two of the first to third complementary color component data by each other to generate first to third products; A color difference data calculation circuit that generates the first and second color difference data based on the first to third products, and a luminance data generation circuit that generates luminance data by combining the balance-adjusted image data; Image data processing apparatus characterized by comprising a.   The color difference data calculating circuit calculates a square root of the first to third products to generate the first to third roots, and the first to third roots represent the three primary colors of light. The image data processing apparatus according to claim 1, further comprising: a matrix circuit that regards the first to third primary color data and generates first and second color difference data by combining them at a predetermined ratio. .   The color difference data calculation circuit includes a subtraction circuit that subtracts the second product from the first and third products to generate a first and second difference, and an absolute value of the first and second differences. A square root circuit for calculating a square root of the value to generate first and second roots, a sign adding circuit for adding the first and second roots to the first and second roots, and 2. The image data processing according to claim 1, further comprising: a color difference matrix circuit that generates the first and second color difference data by combining the first and second roots to which codes are added. apparatus.

Images