JP4204731B2 - Contour correction circuit - Google Patents

Description

【００１６】
【表２】

【００１７】
このように０．２５画素ピッチだけ遅延させるディジタルフィルタと、０．７５画素ピッチだけ遅延させるディジタルフィルタとの係数は、左右の係数が入れ替わったものとなっているので、その特性が赤および青色チャンネルと緑色チャンネルとで同じとなるので好適である。
【００１８】
上述したように、カラーテレビジョンカメラ２１において空間画素ずらしが行なわれており、緑色用の固体撮像素子２２Ｇが、赤色および青色用の固体撮像素子２２Ｒおよび２２Ｂに対して０．５画素ピッチだけ空間的にずれているので、ディジタルフィルタ３１Ｒ、３１Ｇおよび３１Ｂにおいて、赤、緑および青色のディジタル３原色信号を、それぞれ０．２５画素ピッチ、０．７５画素ピッチおよび０．２５画素ピッチだけ遅延することによってこれらの３原色信号の位相を一致させることができる。
【００１９】
このように位相の一致した３原色信号をマトリックス回路３２に供給し、ここで所定の比率で混合して輝度信号Ｙを生成することができる。この３原色信号Ｒ、Ｇ、Ｂの混合比は、例えばＹ＝０．２５Ｒ＋０．５Ｇ＋０．２５Ｂとすることができる。このように、緑色信号と、赤および青色信号を合成した信号との比が０．５：０．５となるように設定することにより、信号中の偽信号となる折り返し成分を相殺することができる。マトリックス回路３２から出力される輝度信号Ｙを、ディジタルバンドパスフィルタ３３に供給して任意の周波数成分を輪郭信号として抽出し、この輪郭信号をディジタルゲイン調整回路３４に供給して、輪郭信号の振幅を本線の３原色信号のレベルに対して調整する。
【００２０】
本発明においては、このようにしてレベルを調整した輪郭信号を、加算回路３０Ｒ、３０Ｇおよび３０Ｂの他方の入力端子に直接供給せずに、ディジタルバンドパスフィルタ３３において抽出される周波数に応じて、アナログ回路の群遅延特性によって輪郭信号の位相がずれるのを補正するために、ディジタルゲイン調整回路３４から出力される輪郭信号を、１画素単位での位相の調整と１画素以内での位相の調整とを独立して行うことができる赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂと、同じく１画素単位での位相の調整と１画素以内での位相の調整とを独立して行うことができる緑色用のディジタル位相調整チャンネル３５Ｇとに供給し、赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂで位相が調整された輪郭信号を加算回路３０Ｒおよび３０Ｂの他方の入力端子に供給して本線の赤および青色信号と加算し、緑色用のディジタル位相調整チャンネル３５Ｇで位相が調整された輪郭信号を加算回路３０Ｇの他方の入力端子に供給して本線の緑色信号と加算する。
【００２１】
赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂは、１画素単位で位相を調整することができるディジタル遅延回路３６Ｒ／Ｂと、１画素以内で位相の調整を行うディジタルフィルタ３７Ｒ／Ｂとを具え、緑色用のディジタル位相調整チャンネル３５Ｇは、１画素単位で位相を調整することができるディジタル遅延回路３６Ｇと、１画素以内の位相の調整を行うディジタルフィルタ３７Ｇとを具えている。このようにして位相を補正した輪郭信号を３原色信号と加算することによって従来のような輪郭強調の非対称性を除くことができる。加算回路３０Ｒ、３０Ｇおよび３０Ｂから出力される輪郭補正された３原色信号を、さらにクリップ回路３８Ｒ、３８Ｇおよび３８Ｂへ供給して規定のレベルでクリップして出力端子３９Ｒ、３９Ｇおよび３９Ｂに、それぞれ輪郭補正が施された３原色信号が得られる。
【００２２】
図５は、赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂおよび緑色用のディジタル位相調整チャンネル３５Ｇに設けられた１画素単位で輪郭信号の位相を調整することができるディジタル遅延回路３６Ｒ／Ｂおよび３６Ｇの一例の構成を示すものである。１画素ピッチのクロックで動作する第１〜第３の３つのフリップフロップ４１〜４３を直列に接続し、第１〜第３のフリップフロップ４１〜４３の出力側からそれぞれ−１画素、０画素、＋１画素遅延した輪郭信号が得られるようにし、このように遅延した３つの輪郭信号を選択回路４４へ供給し、遅延選択信号に応じて何れかの輪郭信号を選択できるように構成する。フリップフロップの段数を増やし、遅延選択を増えたフリップフロップの段数に対応させることにより容易に位相調節の範囲を広げることができる。
【００２３】
図６は、赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂおよび緑色用のディジタル位相調整チャンネル３５Ｇに設けられた１画素以内での位相調整を行うことができるディジタルフィルタ３７Ｒ／Ｂおよび３７Ｇの一例の構成を示すものである。本例では、入力端子Inのデジタル入力信号を直列接続した１６段のフリップフロップＦＦ０〜ＦＦ1５に順次に供給し、これらのフリップフロップの各々から出力される信号に、それぞれ乗算器M０〜M1５において所定の係数K０〜K1５を乗算し、その乗算結果を加算器Addにおいて加算し、その加算結果を割算器Divにおいて２５６で除算して出力端子Outからフィルタリング処理されたデジタル出力信号を出力するように構成する。このようなデジタルフィルタにおいては、係数K０〜K1５を適当に設定することによって所望のフィルタ特性を実現することができる。これについては、後に詳述する。
【００２４】
本例において、ディジタルバンドパスフィルタ３３で抽出する周波数が比較的低く、アナログ回路の群遅延特性がずれていない場合には、赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂおよび緑色用のディジタル位相調整チャンネル３５Ｇに設けられた１画素単位で輪郭信号の位相を調整することができるディジタル遅延回路３６Ｒ／Ｂおよび３６Ｇでの遅延を共に０画素とし、赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂに設けられた１画素以内での位相調整を行うことができるディジタルフィルタ３７Ｒ／Ｂでの遅延は０．７５画素とし、緑色用のディジタル位相調整チャンネル３５Ｇに設けられた１画素以内での位相調整を行うことができるディジタルフィルタ３７Ｇでの遅延は０．２５画素とする。
【００２５】
各素子の遅延量を上述したように設定した場合の位相関係を図７に示す。なお、説明を解り易くするために、マトリックス３２、ＢＰＦ３３、ディジタルゲイン調整回路３４では位相はずれないものとする。また、信号線上に括弧で示す数値は入力のＧチャンネルを基準としたときの遅延位相を表し、ブロック内に示す数値は各素子での遅延位相を表すものである。ディジタル遅延回路２９Ｒ、２９Ｇ、２９Ｂの遅延時間は１画素分の遅延時間とする。輪郭補正回路でのＧチャンネルの輪郭信号は、ディジタルフィルタ３１Ｇでの０．７５画素分とディジタルフィルタ３７Ｇでの０．２５画素分で合計１画素分遅れる。また、Ｒ／Ｂチャンネルの輪郭信号は、ディジタルフィルタ３１Ｒ、３１Ｂでの０．２５画素分とディジタルフィルタ３７Ｒ／Ｂでの０．７５画素分の合成１画素分遅れることになる。よって、加算回路３０Ｒ、３０Ｇおよび３０Ｂにおいて本線信号の位相と輪郭信号の位相とは合致したものとなる。
【００２６】
一方、ディジタルバンドパスフィルタ３３で抽出する周波数が比較的高く、アナログ回路の群遅延特性によって、ディジタルバンドパスフィルタ３３で抽出される輪郭信号の位相が、本線の３原色信号に対して、例えば０．４画素の遅れが生じてしまうような場合には、赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂに設けられた１画素単位で輪郭信号の位相を調整することができるディジタル遅延回路３６Ｒ／Ｂでの遅延を０画素とし、１画素以内での位相調整を行うことができるディジタルフィルタ３７Ｒ／Ｂでの遅延を０．３５画素とし、緑色用のディジタル位相調整チャンネル３５Ｇに設けられた１画素単位で輪郭信号の位相を調整することができるディジタル遅延回路３６Ｇでの遅延を−１画素とし、１画素以内での位相調整を行うことができるディジタルフィルタ３７Ｇでの遅延を０．８５画素とする。
【００２７】
これらの位相関係を図８に示す。この場合にもマトリックス３２、ＢＰＦ３３、ディジタルゲイン調整回路３４では位相はずれないものとする。ディジタル遅延回路２９Ｒ、２９Ｇ、２９Ｂの遅延時間は１画素分の遅延とする。輪郭補正回路でのＧチャンネルの輪郭信号はディジタルフィルタ３１での０．７５画素分とディジタルバンドパスフィルタ３３での０．４画素分と、ディジタル遅延回路３６Ｇでの−１画素分と、ディジタルフィルタ３７Ｇでの０．８５画素分で合計１画素分遅れる。また、Ｒ／Ｂチャンネルの輪郭信号は、ディジタルフィルタ３１Ｒでの０．２５画素分と、ディジタルバンドパスフィルタ３３での０．４画素分と、ディジタルフィルタ３７Ｒ／Ｂでの０．３５画素分との合計１画素分遅れることになる。このように遅延量を設定することにより、ディジタルバンドパスフィルタ３３で高い周波数を抽出する場合でも、加算回路３０Ｒ、３０Ｇおよび３０Ｂにおいて本線信号の位相と輪郭信号の位相とは合致したものとなる。
【００２８】
上述したように、０．８５画素だけ遅延させるデジタルフィルタは、次表３に示すように係数Ｋ０〜Ｋ１５を設定することによって実現することができ、また０．３５画素だけ遅延させるデジタルフィルタは、次表４に示すように係数Ｋ０〜Ｋ１５を設定することによって実現することができる。なお、このようなディジタルフィルタは、係数の合計が２５６となるときに、ゲインが１となるようなものである。
【００２９】
【表３】

【００３０】
【表４】

【００３１】
ディジタルバンドパスフィルタ３３で抽出される輪郭信号の周波数をさらに高くする場合には、アナログ回路の群遅延特性はさらにずれるので、それを補正できるように１画素単位で輪郭信号の位相を調整することができるディジタル遅延回路３６Ｒ／Ｂおよび３６Ｇでの遅延量および１画素以内での位相調整を行うことができるディジタルフィルタ３７Ｒ／Ｂおよび３７Ｇでの遅延量が得られるようにフィルタ係数を設定すれば良い。
【００３２】
上述した実施例では、入力３原色信号の間に画素ずらしがある場合であるが、本発明は画素ずらしがない場合にも適用できる。次に、このような場合について説明する。画素ずらしが行なわれていないカラーテレビジョンカメラ２１から出力される３原色信号に対しては、ディジタルフィルタ３１Ｒ、３１Ｇおよび３１Ｂでは位相をすらさないようにする。このように、位相をずらさない、すなわち０画素だけ遅延させる（すなわち遅延させない）デジタルフィルタは、次表５に示すように係数Ｋ０〜Ｋ１５を設定することによって実現することができる。
【００３３】
【表５】

【００３４】
また、マトリックス回路３２においては、３原色信号の間で空間画素ずらしを行なっていないので、通常の放送規格の輝度信号Ｙを生成するときと同じ混合比率で混合すればよい。すなわち、
Ｙ＝０．３Ｒ＋０．５９Ｇ＋０．１１Ｂ
の式にしたがって３原色信号を混合することによって輝度信号Ｙを生成する。このようにして生成された輝度信号Ｙをディジタルバンドパスフィルタ３３に供給して所望の周波数の輪郭信号を抽出し、そのレベルを調整した後、赤・青色用のディジタル位相調整チャンネル３５Ｒ／Ｂおよび緑色用のディジタル位相調整チャンネル３５Ｇに設けられた１画素単位で輪郭信号の位相を調整することができるディジタル遅延回路３６Ｒ／Ｂおよび３６Ｇおよび１画素以内での位相調整を行うことができるディジタルフィルタ３７Ｒ／Ｂおよび３７Ｇでの遅延量をアナログ回路の群遅延特性のずれによる位相のずれを補正できるような値に設定することにより、加算回路３０Ｒ、３０Ｇおよび３０Ｂにおいて互いに位相の合った本線信号と輪郭信号とを混合することができる。
【００３５】
上述したように本発明による輪郭補正回路においては、赤、緑および青色信号入力端子２８Ｒ、２８Ｇ、２８Ｂで受けるデジタル３原色信号の間で空間画素ずらしが行なわれていない場合には、前記輪郭信号を前記赤、緑および青色信号入力端子からのディジタル３原色信号に加える前記赤、緑および青色信号加算回路３０Ｒ、３０Ｇ、３０Ｂとの間にそれぞれ設けたディジタル遅延回路２９Ｒ、２９Ｇ、２９Ｂの遅延時間をＴとし、前記緑色用ディジタル位相調整回路３５Ｇの遅延時間をＸｇとし、前記赤／青色用ディジタル位相調整回路３５Ｒ／Ｂの遅延時間をＸｒとし、ディジタルバンドパスフィルタ３３で抽出される輪郭信号の遅れをＤとするとき、Ｘｇ＝Ｘｒ＝Ｔ−Ｄを満足するように各部の遅延時間を設定すれば良い。
【００３６】
一方、デジタル３原色信号の間で空間画素ずらしが行なわれている場合には、１画素分の遅延時間をＰとするとき、Ｘｇ＝Ｔ−Ｄ−０．７５ＰおよびＸｒ＝Ｔ−Ｄ−０．２５Ｐを満足するように各部の遅延時間を設定することによって、ディジタルバンドパスフィルタ３３において輝度信号から輪郭信号を抽出する際の周波数に応じて、アナログ回路の群遅延特性がずれるに基づく位相のずれを補正することができ、輪郭補正を立ち上がり部および立ち下がり部において対称的とすることができ、輪郭補正された画像の品質を向上することができる。
【００３７】
本発明は上述した実施例にのみ限定されるものではなく、幾多の変更や変形が可能である。例えば、上述した実施例においては、輪郭信号を緑色信号加算回路３０Ｇに加える緑色用のディジタル位相調整チャンネル３５Ｇに、１画素単位で位相を調整するディジタル遅延回路３６Ｇと１画素以内の位相を調整するディジタルフィルタ３７Ｇとを設け、輪郭信号を赤色および青信号加算回路３０Ｒおよび３０Ｂにそれぞれ加えるディジタル位相調整チャンネル３５Ｒ／Ｇに、１画素単位で位相を調整するディジタル遅延回路３６Ｒ／Ｂと１画素以内の位相を調整するディジタルフィルタ３７Ｒ／Ｂとを設けたが、１画素単位で位相を調整するディジタル遅延回路と１画素以内の位相を調整するディジタルフィルタとを一つのディジタルフィルタで構成することもできる。
【００３８】
図９は上述したように１画素単位で位相を調整するディジタル遅延回路の機能と１画素以内の位相を調整するディジタルフィルタの機能とを併せ持つディジタルフィルタの一例の構成を示すものである。入力端子Inのデジタル入力信号を直列接続した１８段のフリップフロップＦＦ０〜ＦＦ1７に順次に供給し、これらのフリップフロップの各々から出力される信号に、それぞれ乗算器M０〜M1７において所定の係数K０〜K1７を乗算し、その乗算結果を加算器Addにおいて加算し、その加算結果を割算器Divにおいて２５６で除算して出力端子Outからフィルタリング処理されたデジタル出力信号を出力するように構成する。このようなデジタルフィルタにおいて、輪郭信号を．例えば（０．２５−１）画素だけ遅延させる場合の係数K０〜K1７を表６に示し、（０．２５＋１）画素だけ遅延させる場合の係数Ｋ０〜Ｋ１７を表７に示す。
【００３９】
【表６】

【００４０】
【表７】

【００４１】
【発明の効果】
上述したように、本発明によれば、本線の赤および青色信号に加えるべき輪郭信号の位相と、本線の緑色信号に加算すべき輪郭信号の位相とを、１画素単位で位相を調整すると共に１画素以内の位相を調整するディジタル位相調整チャンネルで独立に調整できるようにしたので、ディジタルバンドパスフィルタにおいて輝度信号から輪郭信号を抽出する際の周波数に応じて、アナログ回路の群遅延特性がずれるに基づく位相のずれを補正することができ、したがって輪郭補正を立ち上がり部および立ち下がり部において対称的とすることができ、輪郭補正された画像の品質を向上することができる。
【００４２】
さらに、本発明において、ディジタル位相調整チャンネルにおける１画素単位での位相の調整および１画素以内での位相の調整を、入力される３原色信号の空間画素ずらしを考慮して設定することにより、空間画素ずらしが行なわれていない場合でも、空間画素ずらしが行なわれている場合にも適用することができる。
【図面の簡単な説明】
【図１】 図１従来の輪郭補正回路を具えるカラーテレビジョンカメラシステムの全体の構成を示すブロック図である。
【図２】 図２Ａ、ＢおよびＣは、従来の輪郭補正特性を示すグラフである。
【図３】 図３は、アナログ回路の群遅延特性を示すグラフである。
【図４】 図４は、本発明による輪郭補正回路を具えるカラーテレビジョンカメラシステムの全体の構成を示すブロック図である。
【図５】 図５は、画素単位での位相調整を行うディジタル遅延回路の一例の構成を示すブロック図である。
【図６】 図６は、１画素以内での位相調整を行うディジタルフィルタの一例の構成を示すブロック図である。
【図７】 図７は、アナログ回路の群遅延特性がずれていない場合の各部の信号の位相関係を示す線図である。
【図８】 図８は、アナログ回路の群遅延特性がずれている場合の各部の信号の位相関係を示す線図である。
【図９】 図９は、画素単位での位相調整および１画素以内での位相調整を行うディジタルフィルタの一例の構成を示すブロック図である。
【符号の説明】
２１ カラーテレビジョンカメラ、 ２２Ｒ、２２Ｇ、２２Ｂ 固体撮像素子、２３Ｒ、２３Ｇ、２３Ｂ プリアンプ、 ２４Ｒ、２４Ｇ、２４Ｂ 前置フィルタ、 ２５Ｒ、２５Ｇ、２５Ｂ Ａ／Ｄコンバータ、 ２６Ｒ、２６Ｇ、２６Ｂ ガンマ補正回路、 ２７ 輪郭補正回路、 ２８Ｒ、２８Ｇ、２８Ｂ 入力端子、 ２９Ｒ、２９Ｇ、２９Ｂ デジタル遅延回路、 ３０Ｒ、３０Ｇ、３０Ｂ 加算回路、 ３１Ｒ、３１Ｇ、３１Ｂ ディジタルフィルタ、 ３２ マトリックス回路、 ３３ ディジタルバンドパスフィルタ、 ３４ ディジタルゲイン調整回路、 ３５Ｒ／Ｂ、３５Ｇ ディジタル位相調整チャンネル、 ３６Ｒ／Ｂ、３６Ｇ ディジタル遅延回路、 ３７Ｒ／Ｂ、３７Ｇ ディジタルフィルタ、 ３８Ｒ、３８Ｇ、３８Ｂ クリップ回路[0001]
[Industrial application fields]
The present invention includes, for example, red, green and blue signal input terminals for receiving digital three primary color signals output from a color television camera and obtained by converting analog three primary color signals processed by an analog signal processing circuit. A matrix circuit that generates a luminance signal Y by mixing primary color signals, a contour signal generation circuit that extracts an arbitrary frequency component from the Y signal to generate a contour signal, and the contour signal as the red, green, and blue signals The present invention relates to a contour correction circuit including a red, green and blue signal adding circuit added to a digital three primary color signal from an input terminal.
[0002]
[Prior art]
FIG. 1 shows the overall configuration of a color television signal processing apparatus including the above-described contour correction circuit. As shown in FIG. 1, three primary color signals R, G, and B output from solid-state imaging devices 2R, 2G, and 2B such as red, green, and blue CCDs provided in a color television camera 1 are preamplifiers, respectively. Amplified by 3R, 3G, and 3B, band-limited to the Nyquist frequency of the CCD output sampling by the pre-filters 4R, 4G, and 4B, and then converted to digital three primary color signals by the A / D converters 5R, 5G, and 5B The These digital three primary color signals are further processed by gamma correction circuits 6R, 6G, and 6B for adjusting to the gamma characteristic of the monitor, and then supplied to the contour correction circuit 7.
[0003]
The input terminals 8R, 8G and 8B of the contour correction circuit 7 are supplied with red, green and blue digital three primary color signals, respectively. These digital three primary color signals are supplied to one input terminal of addition circuits 10R, 10G and 10B through digital delay circuits 9R, 9G and 9B, respectively. Further, the digital three primary color signals of red, green and blue received at the input terminals 8R, 8G and 8B are supplied to the matrix circuit 11, where they are mixed at a predetermined ratio to generate the luminance signal Y. Now, the color television camera 1 is performing spatial pixel shifting, and the green solid-state imaging device 2G is spatially shifted by 1/2 pixel with respect to the red and blue solid-state imaging devices 2R and 2B. In this case, the mixing ratio of the three primary color signals R, G, B can be Y = 0.25R + 0.5G + 0.25B, for example. That is, the ratio between the green signal and the signal obtained by combining the red and blue signals is set to 0.5: 0.5. By synthesizing in this way, aliasing components that are included in the signal and become false signals are canceled out.
[0004]
The luminance signal Y output from the matrix circuit 11 is supplied to the digital band pass filter 12 to extract an arbitrary frequency component as a contour signal. This contour signal is supplied to the digital gain adjustment circuit 13 so that the amplitude of the contour signal is adjusted to a predetermined level with respect to the level of the color signal of the main line. The contour signal whose level has been adjusted in this way is supplied to the other input terminal of each of the addition circuits 10R, 10G, and 10B to perform contour correction. The three primary color signals whose contours are corrected in this way are further supplied to the clipping circuits 14R, 14G and 14B and clipped at a prescribed level, and the three primary color signals which have been subjected to contour correction on the output terminals 15R, 15G and 15B, respectively. Will be obtained.
[0005]
[Problems to be Solved by the Invention]
In the conventional contour correction circuit described above, if an image signal having a waveform as shown in FIG. 2A is supplied from the color television camera 1, for example, the digital gain adjustment circuit 13 of the contour correction circuit 7 The contour signal as shown in FIG. This contour signal is added to the three primary color signals of the main line supplied from the digital delay circuits 9R, 9G and 9B by the adder circuits 10R, 10G and 10B. In this case, the contour signal is converted into a video waveform as shown in FIG. 2C. There is a problem that the rising part and the falling part are added asymmetrically. That is, the contour signal added to the rising start portion of the video signal is small, the contour signal added to the rising end portion is large, and the contour signal added to the falling start portion is small and added to the end portion. The contour signal to be generated becomes large.
[0006]
As described above, when the problem that the contour signal is added asymmetrically between the rising part and the falling part of the video waveform is analyzed, it is output from the solid-state imaging devices 2R, 2G, and 2B of the color television camera 1. Until the analog three primary color signals are converted into digital three primary color signals by the A / D converters 5R, 5G, and 5B, processing is performed by analog circuits such as preamplifiers 3R, 3G, and 3B and pre-filters 4R, 4G, and 4B. However, it has been confirmed that these analog circuits are caused by the fact that the group delay is not uniform over the entire frequency as shown in FIG.
[0007]
In the contour correction circuit, the frequency extracted by the digital bandpass filter 12 is changed according to the user's instruction. However, there is no deviation when extracting a low frequency, but the extracted signal becomes higher and the contour signal becomes the main line. This is slightly different from the three primary color signals. The higher the frequency to be extracted, the greater the phase shift, and asymmetry as shown in FIG. 2C will appear. Such a problem is often seen when a high frequency signal such as an HDTV signal is handled.
[0008]
An object of the present invention is to process an analog three primary color signal output from a color television camera by an analog circuit, convert it to a digital three primary color signal, and perform an outline correction based on the digital three primary color signal. Therefore, it is an object of the present invention to provide a contour correction circuit that can effectively suppress the asymmetry of the contour signal added to the edge of the image due to the frequency dependence of the group delay.
[0009]
[Means for Solving the Problems]
According to the present invention, a red, green and blue signal input terminal for receiving a digital three primary color signal obtained by converting an analog three primary color signal processed by an analog signal processing circuit, and the digital three primary color signal are mixed to obtain a luminance signal Y. A matrix circuit that generates a contour signal by extracting an arbitrary frequency component from the Y signal, and this contour signal is converted into the digital three primary color signals from the red, green, and blue signal input terminals. In a contour correction circuit including a red, green, and blue signal adding circuit to be added, the phase is adjusted in units of one pixel and 1 in a channel to which the contour signal output from the contour signal generating circuit is added to the green signal adding circuit. Provided is a green digital phase adjusting means for adjusting a phase within a pixel, and the contour signal is converted into a red signal adding circuit and a blue signal adding circuit. Each channel to be added is provided with a red / blue digital phase adjusting means for adjusting the phase in units of one pixel and adjusting the phase within one pixel, and the phase in the digital phase adjusting means for green and red / blue is adjusted. The adjustment is made to compensate for the phase shift of the digital three primary color signals due to the group delay characteristic in the analog signal processing circuit.
[0010]
In the contour correction circuit according to the present invention, the green digital phase adjusting means is provided with a green digital delay circuit for adjusting the phase in units of one pixel and a green digital filter for adjusting the phase within one pixel, The digital phase adjusting means for / blue can be provided with a red / blue digital delay circuit for adjusting the phase in units of one pixel and a red / blue digital filter for adjusting the phase within one pixel.
[0011]
  Further, when the three primary color signals subjected to the spatial pixel shift are supplied to the red, green and blue signal input terminals, the digital three primary color signals are provided between the red, green and blue signal input terminals and the matrix circuit. Phase matching digital filter means for aligning the phases can be provided, and the phase in the phase adjustment means can be adjusted in consideration of the phase introduced by the phase matching digital filter means..
[0012]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 4 shows the overall configuration of a color television camera system having a contour correction circuit according to the present invention. The color television camera 21 is provided with solid-state image sensors 22R, 22G, and 22B such as red, green, and blue CCDs. The three primary color signals R, G, and B output from these solid-state image sensors 22R, 22G, and 22B are amplified by preamplifiers 23R, 23G, and 23B, respectively, and further, CCD outputs are sampled by pre-filters 24R, 24G, and 24B. After band limiting to the Nyquist frequency or lower, the signals are converted into digital three primary color signals by A / D converters 25R, 25G, and 25B, respectively. These digital three primary color signals are supplied to the gamma correction circuits 26R, 26G, and 26B, processed so as to match the gamma characteristics of the monitor, and then supplied to the contour correction circuit 27.
[0013]
The three primary color signals of red, green and blue received at the input terminals 28R, 28G and 28B of the contour correction circuit 27 are passed through the digital delay circuits 29R, 29G and 29B, respectively, and one input terminal of the adder circuits 30R, 30G and 30B. To supply each. In this example, spatial pixel shifting is performed with respect to the solid-state imaging devices 22R, 22G, and 22B of the color television camera 21, and the red and blue signals R and B have a 0.5 pixel pitch with respect to the green signal G. Is just what is delayed. Therefore, a digital filter which gives delays corresponding to 0.25 pixel pitch, 0.75 pixel pitch and 0.25 pixel pitch to the red, green and blue digital three primary color signals received at input terminals 28R, 28G and 28B, respectively. The signals are supplied to the matrix circuit 32 through 31R, 31G and 31B, and mixed at a predetermined ratio to generate the luminance signal Y.
[0014]
As a digital filter that gives such a delay, 16 stages of flip-flops that operate with a clock of one pixel pitch are connected in series, and signals output from each of them are multiplied by predetermined coefficients K0 to K15, respectively. By adding the multiplication results and dividing the addition result by 256, it is possible to output a filtered digital output signal. In such a digital filter, coefficients K0 to K15 when red and blue signals R and B are delayed by 0.25 pixel pitch are shown in Table 1, and coefficients when green signal G is delayed by 0.75 pixel pitch. Table 2 shows K0 to K15.
[0015]
[Table 1]

[0016]
[Table 2]

[0017]
Since the coefficients of the digital filter delayed by 0.25 pixel pitch and the digital filter delayed by 0.75 pixel pitch in this way are the left and right coefficients interchanged, their characteristics are red and blue channels. And the green channel are the same.
[0018]
As described above, spatial pixel shifting is performed in the color television camera 21, and the solid-state image pickup device 22G for green has a space of 0.5 pixel pitch with respect to the solid-state image pickup devices 22R and 22B for red and blue. Therefore, in the digital filters 31R, 31G, and 31B, the digital three primary color signals of red, green, and blue are delayed by 0.25 pixel pitch, 0.75 pixel pitch, and 0.25 pixel pitch, respectively. Thus, the phases of these three primary color signals can be matched.
[0019]
Thus, the three primary color signals having the same phase can be supplied to the matrix circuit 32 and mixed at a predetermined ratio to generate the luminance signal Y. The mixing ratio of the three primary color signals R, G, and B can be set to Y = 0.25R + 0.5G + 0.25B, for example. Thus, by setting the ratio of the green signal and the signal obtained by combining the red and blue signals to be 0.5: 0.5, the aliasing component that becomes a false signal in the signal can be canceled out. it can. The luminance signal Y output from the matrix circuit 32 is supplied to the digital band-pass filter 33 to extract an arbitrary frequency component as a contour signal, and this contour signal is supplied to the digital gain adjustment circuit 34 to thereby detect the amplitude of the contour signal. Is adjusted with respect to the level of the three primary color signals of the main line.
[0020]
In the present invention, the contour signal whose level is adjusted in this way is not directly supplied to the other input terminals of the adder circuits 30R, 30G, and 30B, but according to the frequency extracted by the digital bandpass filter 33, In order to correct the phase shift of the contour signal due to the group delay characteristic of the analog circuit, the contour signal output from the digital gain adjustment circuit 34 is adjusted in phase in units of pixels and in phase within one pixel. The digital phase adjustment channel 35R / B for red and blue can be independently performed, and the phase adjustment in units of one pixel and the phase adjustment within one pixel can be independently performed. The contour signal whose phase is adjusted by the digital phase adjustment channel 35R / B for red and blue is supplied to the digital phase adjustment channel 35G for green. Is added to the other input terminals of the adder circuits 30R and 30B and added to the red and blue signals of the main line, and the contour signal whose phase is adjusted by the digital phase adjustment channel 35G for green is supplied to the other input terminal of the adder circuit 30G. Is added to the green signal of the main line.
[0021]
The digital phase adjustment channel 35R / B for red and blue includes a digital delay circuit 36R / B that can adjust the phase in units of one pixel and a digital filter 37R / B that adjusts the phase within one pixel. The green digital phase adjustment channel 35G includes a digital delay circuit 36G that can adjust the phase in units of one pixel and a digital filter 37G that adjusts the phase within one pixel. By adding the contour signal whose phase is corrected in this way to the three primary color signals, the conventional asymmetry of contour emphasis can be eliminated. The contour-corrected three primary color signals output from the adder circuits 30R, 30G, and 30B are further supplied to the clip circuits 38R, 38G, and 38B, clipped at a specified level, and output to the output terminals 39R, 39G, and 39B, respectively. The corrected three primary color signals are obtained.
[0022]
FIG. 5 shows a digital delay circuit 36R / B and a digital delay circuit 36R / B that can adjust the phase of the contour signal in units of one pixel provided in the digital phase adjustment channel 35R / B for red and blue and the digital phase adjustment channel 35G for green. The structure of an example of 36G is shown. The first to third flip-flops 41 to 43 that operate with a clock of one pixel pitch are connected in series, and from the output side of the first to third flip-flops 41 to 43, −1 pixel, 0 pixel, A contour signal delayed by +1 pixel is obtained, and the three contour signals delayed in this way are supplied to the selection circuit 44, and any one of the contour signals can be selected in accordance with the delay selection signal. The range of phase adjustment can be easily expanded by increasing the number of flip-flop stages and making the delay selection correspond to the increased number of flip-flop stages.
[0023]
FIG. 6 shows an example of digital filters 37R / B and 37G provided in the digital phase adjustment channel 35R / B for red and blue and the digital phase adjustment channel 35G for green and capable of performing phase adjustment within one pixel. The structure of is shown. In this example, a digital input signal of the input terminal In is sequentially supplied to 16 stages of flip-flops FF0 to FF15 connected in series, and signals output from each of these flip-flops are respectively predetermined by multipliers M0 to M15. Are multiplied by the adder Add, and the addition result is divided by 256 in the divider Div to output the filtered digital output signal from the output terminal Out. Constitute. In such a digital filter, desired filter characteristics can be realized by appropriately setting the coefficients K0 to K15. This will be described in detail later.
[0024]
In this example, when the frequency extracted by the digital bandpass filter 33 is relatively low and the group delay characteristics of the analog circuit are not shifted, the digital phase adjustment channel 35R / B for red and blue and the digital phase for green are used. The digital delay circuits 36R / B and 36G that can adjust the phase of the contour signal in units of one pixel provided in the adjustment channel 35G are set to 0 pixels in delay, and the digital phase adjustment channel 35R / B for red and blue is used. The delay in the digital filter 37R / B capable of performing phase adjustment within one pixel provided in the filter is 0.75 pixel, and the phase adjustment within one pixel provided in the green digital phase adjustment channel 35G The delay in the digital filter 37G capable of performing the above is 0.25 pixels.
[0025]
FIG. 7 shows the phase relationship when the delay amount of each element is set as described above. For easy understanding, it is assumed that the matrix 32, the BPF 33, and the digital gain adjustment circuit 34 are out of phase. A numerical value shown in parentheses on the signal line represents a delay phase when the input G channel is used as a reference, and a numerical value shown in the block represents a delay phase in each element. The delay time of the digital delay circuits 29R, 29G, and 29B is a delay time for one pixel. The contour signal of the G channel in the contour correction circuit is delayed by one pixel in total for 0.75 pixels in the digital filter 31G and 0.25 pixels in the digital filter 37G. The contour signal of the R / B channel is delayed by one synthesized pixel corresponding to 0.25 pixels in the digital filters 31R and 31B and 0.75 pixels in the digital filter 37R / B. Therefore, in the adder circuits 30R, 30G, and 30B, the phase of the main line signal matches the phase of the contour signal.
[0026]
On the other hand, the frequency extracted by the digital bandpass filter 33 is relatively high, and the phase of the contour signal extracted by the digital bandpass filter 33 is, for example, 0 with respect to the three primary color signals of the main line due to the group delay characteristics of the analog circuit. When a delay of 4 pixels occurs, the digital delay circuit 36R / that can adjust the phase of the contour signal in units of one pixel provided in the digital phase adjustment channel 35R / B for red and blue The delay in B is 0 pixel, the delay in the digital filter 37R / B capable of performing phase adjustment within 1 pixel is 0.35 pixel, and one pixel provided in the green digital phase adjustment channel 35G The delay in the digital delay circuit 36G that can adjust the phase of the contour signal in units is -1 pixel, and the phase within one pixel The delay in the digital filter 37G which can perform integer and 0.85 pixels.
[0027]
These phase relationships are shown in FIG. Also in this case, it is assumed that the matrix 32, the BPF 33, and the digital gain adjustment circuit 34 are out of phase. The delay time of the digital delay circuits 29R, 29G, and 29B is a delay of one pixel. The G channel contour signal in the contour correction circuit is 0.75 pixels in the digital filter 31, 0.4 pixels in the digital bandpass filter 33, -1 pixel in the digital delay circuit 36G, and the digital filter. A total of one pixel is delayed by 0.85 pixels at 37G. Further, the contour signal of the R / B channel includes 0.25 pixels for the digital filter 31R, 0.4 pixels for the digital bandpass filter 33, and 0.35 pixels for the digital filter 37R / B. Will be delayed by a total of one pixel. By setting the delay amount in this way, even when a high frequency is extracted by the digital bandpass filter 33, the phase of the main signal and the phase of the contour signal are matched in the adder circuits 30R, 30G, and 30B.
[0028]
As described above, the digital filter delayed by 0.85 pixels can be realized by setting the coefficients K0 to K15 as shown in the following Table 3, and the digital filter delayed by 0.35 pixels is This can be realized by setting coefficients K0 to K15 as shown in Table 4 below. Such a digital filter has a gain of 1 when the sum of the coefficients is 256.
[0029]
[Table 3]

[0030]
[Table 4]

[0031]
When the frequency of the contour signal extracted by the digital bandpass filter 33 is further increased, the group delay characteristic of the analog circuit is further shifted. Therefore, the phase of the contour signal is adjusted in units of one pixel so that it can be corrected. The filter coefficient may be set so that the delay amount in the digital delay circuits 36R / B and 36G capable of performing the delay and the delay amount in the digital filters 37R / B and 37G capable of performing phase adjustment within one pixel can be obtained. .
[0032]
In the embodiment described above, there is a case where there is a pixel shift between the input three primary color signals, but the present invention can also be applied to a case where there is no pixel shift. Next, such a case will be described. The digital filters 31R, 31G, and 31B do not even phase the three primary color signals output from the color television camera 21 that has not undergone pixel shifting. Thus, a digital filter that does not shift the phase, that is, delays by 0 pixels (that is, does not delay) can be realized by setting the coefficients K0 to K15 as shown in Table 5 below.
[0033]
[Table 5]

[0034]
Further, since the matrix circuit 32 does not perform spatial pixel shifting between the three primary color signals, the matrix circuit 32 may be mixed at the same mixing ratio as that when the luminance signal Y of the normal broadcasting standard is generated. That is,
Y = 0.3R + 0.59G + 0.11B
The luminance signal Y is generated by mixing the three primary color signals according to the following equation. The luminance signal Y generated in this way is supplied to the digital bandpass filter 33 to extract a contour signal of a desired frequency, and after adjusting its level, the digital phase adjustment channels 35R / B for red and blue and Digital delay circuits 36R / B and 36G capable of adjusting the phase of the contour signal in units of one pixel provided in the digital phase adjustment channel 35G for green and a digital filter 37R capable of adjusting the phase within one pixel. By setting the delay amount at / B and 37G to a value that can correct the phase shift due to the shift in the group delay characteristics of the analog circuit, the main circuit signal and the contour in phase with each other in the adder circuits 30R, 30G, and 30B The signal can be mixed.
[0035]
As described above, in the contour correction circuit according to the present invention, when the spatial pixel shift is not performed between the digital three primary color signals received at the red, green, and blue signal input terminals 28R, 28G, 28B, the contour signal Delay time of digital delay circuits 29R, 29G, and 29B provided between the red, green, and blue signal adder circuits 30R, 30G, and 30B, respectively. Is set to T, the delay time of the green digital phase adjustment circuit 35G is set to Xg, the delay time of the red / blue digital phase adjustment circuit 35R / B is set to Xr, and the contour signal extracted by the digital bandpass filter 33 is When the delay is D, the delay time of each part may be set so as to satisfy Xg = Xr = TD.
[0036]
On the other hand, when spatial pixel shifting is performed between the digital three primary color signals, assuming that the delay time for one pixel is P, Xg = TD-0.75P and Xr = TD-0 By setting the delay time of each part so as to satisfy .25P, the phase of the phase based on the deviation of the group delay characteristic of the analog circuit according to the frequency when the contour signal is extracted from the luminance signal in the digital bandpass filter 33 is obtained. The shift can be corrected, the contour correction can be made symmetric at the rising and falling portions, and the quality of the contour-corrected image can be improved.
[0037]
The present invention is not limited to the above-described embodiments, and many changes and modifications can be made. For example, in the above-described embodiment, the digital phase adjustment channel 35G for applying the contour signal to the green signal addition circuit 30G is adjusted to the digital delay circuit 36G for adjusting the phase in units of one pixel and the phase within one pixel. A digital delay circuit 36R / B that adjusts the phase in units of one pixel and a phase within one pixel, and a digital phase adjustment channel 35R / G that provides a digital filter 37G and applies contour signals to the red and blue signal addition circuits 30R and 30B, respectively. Although a digital filter 37R / B for adjusting the phase is provided, the digital delay circuit for adjusting the phase in units of one pixel and the digital filter for adjusting the phase within one pixel can be constituted by one digital filter.
[0038]
FIG. 9 shows a configuration of an example of a digital filter having both the function of the digital delay circuit for adjusting the phase in units of one pixel and the function of the digital filter for adjusting the phase within one pixel as described above. The digital input signal of the input terminal In is sequentially supplied to 18 stages of flip-flops FF0 to FF17 connected in series, and the signals output from each of these flip-flops are respectively given predetermined coefficients K0 to K0 in multipliers M0 to M17. The multiplication result is multiplied by K17, and the multiplication result is added by the adder Add, and the addition result is divided by 256 by the divider Div, and the filtered digital output signal is output from the output terminal Out. In such a digital filter, the contour signal is. For example, the coefficients K0 to K17 for delaying by (0.25-1) pixels are shown in Table 6, and the coefficients K0 to K17 for delaying by (0.25 + 1) pixels are shown in Table 7.
[0039]
[Table 6]

[0040]
[Table 7]

[0041]
【The invention's effect】
As described above, according to the present invention, the phase of the contour signal to be added to the main line red and blue signals and the phase of the contour signal to be added to the main line green signal are adjusted in units of one pixel. Since the digital phase adjustment channel for adjusting the phase within one pixel can be adjusted independently, the group delay characteristic of the analog circuit is shifted in accordance with the frequency when the contour signal is extracted from the luminance signal in the digital band pass filter. Therefore, the contour correction can be made symmetrical in the rising and falling portions, and the quality of the contour-corrected image can be improved.
[0042]
Furthermore, in the present invention, the phase adjustment in units of one pixel and the phase adjustment within one pixel in the digital phase adjustment channel are set in consideration of the spatial pixel shift of the three primary color signals to be input. Even when the pixel shift is not performed, the present invention can be applied even when the spatial pixel shift is performed.
[Brief description of the drawings]
1 is a block diagram showing the overall configuration of a color television camera system having a conventional contour correction circuit.
FIGS. 2A, 2B, and 2C are graphs showing conventional contour correction characteristics.
FIG. 3 is a graph illustrating group delay characteristics of an analog circuit.
FIG. 4 is a block diagram showing the overall configuration of a color television camera system including a contour correction circuit according to the present invention.
FIG. 5 is a block diagram illustrating a configuration of an example of a digital delay circuit that performs phase adjustment in units of pixels.
FIG. 6 is a block diagram illustrating a configuration of an example of a digital filter that performs phase adjustment within one pixel.
FIG. 7 is a diagram showing the phase relationship of signals at various parts when the group delay characteristics of the analog circuit are not shifted.
FIG. 8 is a diagram showing the phase relationship of signals at various parts when the group delay characteristics of the analog circuit are shifted.
FIG. 9 is a block diagram illustrating a configuration of an example of a digital filter that performs phase adjustment in units of pixels and phase adjustment within one pixel.
[Explanation of symbols]
21 color television camera, 22R, 22G, 22B solid-state imaging device, 23R, 23G, 23B preamplifier, 24R, 24G, 24B pre-filter, 25R, 25G, 25B A / D converter, 26R, 26G, 26B gamma correction circuit, 27 Contour Correction Circuit, 28R, 28G, 28B Input Terminal, 29R, 29G, 29B Digital Delay Circuit, 30R, 30G, 30B Adder Circuit, 31R, 31G, 31B Digital Filter, 32 Matrix Circuit, 33 Digital Bandpass Filter, 34 Digital Gain adjustment circuit, 35R / B, 35G digital phase adjustment channel, 36R / B, 36G digital delay circuit, 37R / B, 37G digital filter, 38R, 38G, 38B clip times

Claims (6)

  Red, green and blue signal input terminals for receiving digital three primary color signals obtained by converting analog three primary color signals processed by the analog signal processing circuit, and a matrix for generating a luminance signal Y by mixing these digital three primary color signals A circuit, a contour signal generating circuit for extracting an arbitrary frequency component from the Y signal to generate a contour signal, and red and green for adding the contour signal to the digital three primary color signals from the red, green and blue signal input terminals. And a blue signal adding circuit, wherein the phase is adjusted in units of one pixel and the phase is within one pixel in a channel for adding the contour signal output from the contour signal generating circuit to the green signal adding circuit. A digital phase adjusting means for green is provided, and the contour signal is supplied to the red signal adding circuit and the blue signal adding circuit, respectively. The red / blue digital phase adjusting means for adjusting the phase in units of one pixel and adjusting the phase within one pixel is provided in each channel, and the phases of these digital phase adjusting means for green and red / blue are adjusted. An outline correction circuit configured to compensate for a phase shift of the digital three primary color signals due to group delay characteristics in the analog signal processing circuit.   The green digital phase adjusting means is provided with a green digital delay circuit for adjusting the phase in units of one pixel and a green digital filter for adjusting the phase within one pixel, and the red / blue digital phase adjusting means 2. The contour correction circuit according to claim 1, further comprising a red / blue digital delay circuit for adjusting a phase in units of one pixel and a red / blue digital filter for adjusting a phase within one pixel.   When the three primary color signals subjected to spatial pixel shifting are supplied to the red, green and blue signal input terminals, the phase of the digital three primary color signal is between the red, green and blue signal input terminals and the matrix circuit. Phase matching digital filter means for adjusting the phase and adjusting the phase in units of one pixel and adjusting the phase within one pixel, and adjusting the phase in units of one pixel and phase within one pixel 3. The phase of the red / blue digital phase adjusting means for adjusting the phase is adjusted in consideration of the phase introduced by the phase matching digital filter means. The contour correction circuit described. A digital bandpass filter that extracts a component in a desired frequency range from the luminance signal output from the matrix circuit, and a digital gain adjustment circuit that adjusts the gain of the output signal of the digital bandpass filter; contour correction circuit according to any one of claims 1 to 3, characterized in that the provided. Red, green and blue signal input terminals for receiving digital three primary color signals not subjected to spatial pixel shifting, and the red, green and blue signals for adding the contour signal to the digital three primary color signals from the red, green and blue signal input terminals. A digital delay circuit having a delay time T is provided between the blue signal adding circuit and the delay time of the green digital phase adjusting means Xg, a delay time of the red / blue digital phase adjusting means Xr, 2. The contour correction circuit according to claim 1, wherein Xg = Xr = TD is set, where D is a delay of a contour signal extracted by a digital bandpass filter provided in the contour signal generation circuit. . Red, green and blue signal input terminals for receiving digital three primary color signals subjected to spatial pixel shifting, and the red, green and blue signals for adding the contour signal to the digital three primary color signals from the red, green and blue signal input terminals. A digital delay circuit having a delay time T is provided between the blue signal adding circuit and the delay time of the green digital phase adjusting means Xg, a delay time of the red / blue digital phase adjusting means Xr, When the delay of the contour signal extracted by the digital bandpass filter provided in the contour signal generation circuit is D and the delay time for one pixel is P, Xg = TD−0.75P and Xr = T− 2. The contour correction circuit according to claim 1, wherein the contour correction circuit is set to D-0.25P .

Images