JP3923293B2 - Image processing method, image processing apparatus, and image forming apparatus - Google Patents

Description

【０１０３】
上記空間フィルタ処理部２７は、黒生成下色除去部２６より入力されるＣＭＹＫ信号の画像データに対して、領域識別信号を基にデジタルフィルタによる空間フィルタ処理を行い、空間周波数特性を補正することによって出力画像のぼやけや粒状性劣化を防ぐように処理するものである。
【０１０４】
また、上記階調再現処理部２９は、上記空間フィルタ処理部２７と同様に、ＣＭＹＫ信号の画像データに対して、領域識別信号を基に所定の処理を施すものである。
【０１０５】
例えば、領域分離処理部２４にて文字に分離された領域は、特に黒文字或いは色文字の再現性を高めるために、空間フィルタ処理部２７による空間フィルタ処理における鮮鋭強調処理で高周波数の強調量が大きくされる。同時に、階調再現処理部２９においては、高域周波数の再現に適した高解像度のスクリーンでの二値化または多値化処理が選択される。
【０１０６】
また、領域分離処理部２４にて網点に分離された領域に関しては、空間フィルタ処理部２７において、入力網点成分を除去するためのローパス・フィルタ処理が施される。そして、出力階調補正部２８では、濃度信号などの信号をカラー画像形成装置の特性値である網点面積率に変換する出力階調補正処理を行った後、階調再現処理部２９で、最終的に画像を画素に分離してそれぞれの階調を再現できるように処理する階調再現処理（中間調生成）が施される。領域分離処理部２４にて写真に分離された領域に関しては、階調再現性を重視したスクリーンでの二値化または多値化処理が行われる。
【０１０７】
上述した各処理が施された画像データは、一旦記憶手段（図示せず）に記憶され、所定のタイミングで読み出されてカラー画像出力装置３に入力される。
【０１０８】
このカラー画像出力装置３は、画像データを記録媒体（例えば紙等）上に出力するもので、例えば、電子写真方式やインクジェット方式を用いたカラー画像出力を挙げることができるが、特に限定されるものではなく、画像処理後の画像を表示できる画像表示装置、例えばコンピュータや携帯情報機器に使用されている液晶ディスプレイ等であってもよい。この場合、色補正部２５では、入力されるＲＧＢ信号は画像表示装置の出力特性に応じてＲ’Ｇ’Ｂ’信号に変換される。また、黒生成下色除去部２６、階調再現処理部２９は不要であり、出力階調補正部２８では、空間フィルタ処理部２７から出力された画像データ１に対して、カラー画像出力装置３の特性に応じた階調の調整がなされる。
【０１０９】
ここで、上記カラー画像処理装置２内の色ずれ補正部２３について、図１および図３を参照しながら以下に説明する。
【０１１０】
上記色ずれ補正部２３は、図３に示すように、画像メモリ３０、画素濃度判定部（画素濃度判定手段）３１、フィッティング処理部（近似手段）３２、関数記憶部３３、色ずれ判定部（色ずれ判定手段）３４、閾値記憶部３５、補正処理部（補正手段）３６、フィルタ記憶部３７を含み、外部の制御部（制御手段）３８により駆動制御されるようになっている。
【０１１１】
すなわち、上記色ずれ補正部２３は、入力階調補正部２２（図１）からのＲＧＢ信号（入力画像データ）に対して、画像メモリ３０、画素濃度判定部３１、フィッティング処理部３２、色ずれ判定部３４、補正処理部３６による色ずれ補正処理を施し、色ずれ補正済の画像データとしてのＲＧＢ信号を後段の領域分離処理部２４（図１）に出力する一方、黒生成調整信号を後段の黒生成下色除去部２６（図１）に出力するようになっている。
【０１１２】
色ずれ補正部２３の処理について具体的に説明すると以下のようになる。
【０１１３】
入力階調補正部２２より送られてきたＲＧＢ信号（入力画像データ）に対してｍ×ｎの複数の画素よりなる画像データが画像メモリ３０に格納される（ｍ，ｎは正の整数である。後述するように、例えば１×７マスク領域の画像データを用いて画素濃度の判定を行う場合は、少なくとも３×７の画像データが格納される。）。
【０１１４】
次に、画像メモリ３０より、例えば１×７マスク領域の画像データが抽出されて、画素濃度判定部３１で濃度情報に基づく極大点の有無が判定される。ここで、抽出された画像データに極大点が存在すると判定された場合は１×７マスク領域のＲＧＢ信号はフィッティング処理部３２に出力される。
【０１１５】
なお、上記画素濃度判定部３１における画素濃度判定は必ずしも行わなくてもよいが（後述するように、色ずれ補正部２３を領域分離処理部２４の後段に配置する場合は必要ない）、この判定を行うことにより、文字や線画等の画像領域を速やかに見いだすことができ、極大点が存在しないと判定された場合は、その領域が下地領域あるいは均一な濃度の画像領域であると判断し、フィッティング処理を行う必要がないため、以降の処理の高速化が期待される。
【０１１６】
また、本実施の形態では、上記画素濃度判定部３１における画素濃度の判定基準として、上述したように、『極大点が存在するか否か』を判定基準として用いているが、これに限定されるものではなく、『画素濃度が単調増加傾向にあるか否か』や『所定の濃度以上の画素が所定数存在するか否か』等の判定基準を用いても構わない。
【０１１７】
上記フィッティング処理部３２においては、入力されたＲＧＢ信号の各々に対し、関数記憶部３３で記憶されている関数を用いてフィッティング処理がなされる。フィッティングの際には、画像メモリ３０より送られてくるＲＧＢ値（ＲＧＢ信号を数値化したもの）とフィッティング結果（関数）の２乗誤差が最小となるように算出される。
【０１１８】
上記色ずれ判定部３４では、フィッティング処理部３２によるフィッティング結果により得られた関数の各パラメータに対し閾値記憶部３５で記憶されている閾値を用いた判定が行われ、色ずれが発生しているかどうかが判定され、色ずれが発生していると判定された場合には、画像メモリ３０により送られてくるＲＧＢ信号を補正処理部３６に送る。一方、色ずれが発生していないと判定されれば、画像メモリ３０により送られてくるＲＧＢ信号は、そのまま色ずれ補正部２３の後段の領域分離処理部２４に送られる。
【０１１９】
上記色ずれ判定部３４では、画像メモリ３０より抽出された入力画像データの１×７マスク領域に対して色ずれの発生の有無を判定するようになっている。つまり、色ずれ判定部３４では、１×７マスク領域が色ずれが発生した領域（以下、色ずれ領域と称する）か否かを判定するようになっている。なお、色ずれ判定部３４における色ずれ判定処理の流れについては後述する。
【０１２０】
上記補正処理部３６では、色ずれ領域と判定された領域に対し、例えば、フィルタ記憶部３７で記憶されているフィルタを用いて空間フィルタ処理を行って色ずれを補正する。
【０１２１】
また、上記補正処理部３６は、黒生成下色除去部２６（図１）において黒生成量を制御する場合には、黒生成調整信号を該黒生成下色除去部２６に出力する。
【０１２２】
ここで、上記フィルタ記憶部３７に記憶されたフィルタとして、例えば図４に示すように、検出された色ずれ領域に対して、色ずれ領域内の注目画素ｅを中心とする３×３マスクの最小値フィルタを使用し、この最小値フィルタによりマスク内の最小値（ｅ’＝ｍｉｎ（ａ，ｂ，ｃ，ｄ，ｅ，ｆ，ｇ，ｈ，ｉ））を算出し、色ずれ補正結果として出力する。ここでは色ずれ補正前と補正後を示している。これにより、色ずれ領域は、原稿の下地濃度に近づくこととなり、その結果、文字または線画領域だけが鮮明になる。
【０１２３】
尚、本実施の形態では３×３マスクを用いて説明したが、マスクの大きさはこれに限定されるものではない。最小値フィルタでは、濃度が最小のものを選択して処理後の画素の濃度として採用するため、マスクサイズが大きくなるにしたがって色ずれ領域の濃度は原稿の下地濃度により近づくこととなる。しかし、マスクサイズが大きくなると回路規模が大きくなりコストアップを招来するので、この点を考慮に入れ適切なサイズのものを設計すればよい。
【０１２４】
色ずれ領域を補正する別の方法としては、黒生成下色除去部２６において、黒文字または線画領域の黒生成量をＦ１、色ずれ領域の黒生成量をＦ２とし、Ｆ１＞Ｆ２の関係を満たすように黒生成量を制御するようにしてもよい。これにより、色ずれ領域の黒生成量が、黒文字または線画領域での黒生成量より少なく設定されているので、黒文字または線画が太くなるのを防ぐと共に、色濁りのない良好な画像を形成することができる。
【０１２５】
上記色ずれ補正部２３における各種の処理は、ＣＰＵ（Central Processing Unit ）やＤＳＰ（Digital Signal Processor）をはじめとする制御部３８を介して行われる。
【０１２６】
また、上記制御部３８では、後述する図９のカラー画像処理装置４０において、原稿のプレスキャン（予備走査）を行い、スキャナの読み取り速度を高め、低解像度で読み込まれた入力画像データに対して、領域分離処理を施し、文字または線画領域を抽出して、ＲＧＢ信号の各色成分の差分の絶対値と予め定められる閾値とを比較し、その結果、色ずれが生じていないと判断された時、本スキャン時に色ずれ補正部２３での処理を行わないように制御するようになっている。この場合、プレスキャンが必要になるが、色ずれが生じたと判断された時のみ補正処理を行えば良いので効率よく画像処理を行うことができる。
【０１２７】
次に、上記色ずれ補正部２３における色ずれ判定処理の流れについて、図５に示すフローチャートを参照しながら以下に説明する。
【０１２８】
まず、画像入力が行われ、該入力画像から入力階調補正部２２よりＲＧＢ信号を得る（ステップＳ１）。
【０１２９】
次に、得られたＲＧＢ信号より所定のサイズのデータが画像メモリ３０に格納され、１×７マスク領域のデータが抽出される（ステップＳ２）。
【０１３０】
続いて、抽出されたデータに対して、画素濃度判定部３１で濃度情報に基づく極大点の有無の判定がなされる（ステップＳ３）。ここで、極大点の判定方法の１つとして、マスク領域内における隣り合うデータ（画素濃度）の差分情報を求め単調増加から単調減少、あるいは、ほとんど差が見られない状態へと変化する点があるか否かを調べる方法が挙げられる。
【０１３１】
ステップＳ３において、『極大点が存在する』と判定された場合は、フィッティング処理部３２において関数を用いて近似を行う（ステップＳ４）。
【０１３２】
一方、ステップＳ３において、『極大点が存在する』と判定されなかった場合は、その領域は下地領域あるいは均一な濃度の画像領域であると判定し、ステップＳ８に移行して色ずれ無し（色ずれの影響がない、あるいは目立たない）と判定される。
【０１３３】
ステップＳ４では、１×７マスク領域内のデータに対し関数を用いてフィッティングが行われる。フィッティングを行う場合は、ガウシアン関数をはじめとする各種関数や、スプライン補間、ラグランジュ補間といった補間手法があるが、本発明では、三角関数Ａｃｏｓⁿ （θ＋α）によるフィッティングを行う。このときＡは、振幅を表し、マスク領域内の画素濃度の極大点を示すことになる。また、ｎは、関数の尖度を指しており、ｎの値が大きくなるにつれ対象となる画像領域の線幅が小さくなることを意味している。θは、基準画素位置を、またαは、規準画素位置からの位相差を表しており、このパラメータが分かれば色ずれ量を求めることができる。
【０１３４】
上述のステップＳ４により１×７マスク領域内のデータに対し関数を用いて近似された後、ステップＳ５において、フィッティングにより、ＲＧＢ信号のそれぞれに対して、推定される関数との２乗誤差が最小となるＡ（Ａ_R ，Ａ_G ，Ａ_B ）、α（α_R ，α_G ，α_B ）、ｎ（ｎ_R ，ｎ_G ，ｎ_B ）を求める。
【０１３５】
そして、α（α_R ，α_G ，α_B ）に対し、条件判定を行い色ずれの有無を調べる。すなわち、α_R ＝α_G ＝α_B であるか否かを判定する（ステップＳ６）。ここで、α_R ＝α_G ＝α_B であると判定されれば、色ずれは無いものとしてステップＳ８に移行する。一方、α_R ＝α_G ＝α_B でないと判定されれば、ステップＳ７に移行して、補正処理が行われる。
【０１３６】
ここで、図５に示すフローチャートで示される色ずれ判定処理において、色ずれ補正を行うか否かの条件として、以下の３つの条件が考えられる。
【０１３７】
＜条件１＞
極大点が存在し、「α_R ＝α_G ＝α_B 」を満たすときは、色ずれが発生していないと判定する。つまり、ステップＳ３において、極大点を有すると判定され、ステップＳ６において、α_R ＝α_G ＝α_B を満たしていると判定されたとき、色ずれが発生していないと判定する。
【０１３８】
＜条件２＞
極大点が存在し、「α_R ＝α_G ＝α_B 」を満たさないときは、色ずれが発生していると判定し、色ずれ補正処理を行う。つまり、ステップＳ３において、極大点を有すると判定され、ステップＳ６において、α_R ＝α_G ＝α_B を満たしていないと判定されたとき、色ずれが発生していると判定する。
【０１３９】
＜条件３＞
極大点が存在しないときは、文字または線画領域ではない為、色ずれの影響がない、あるいは目立たないと判定し、色ずれ補正処理は行わない。つまり、ステップＳ３において、極大点を有しないと判定されたときには、色ずれの影響がない、あるいは目立たないと判定する。
【０１４０】
上記の条件２の場合、すなわち入力画像に色ずれが生じていると判定された場合には、ステップＳ７において、色ずれ補正処理、すなわち図３に示す補正処理部３６における補正処理が実行される。この補正処理については、前述したフィルタ記憶部３７に記憶されたフィルタを用いて行われる。
【０１４１】
そして、上記の色ずれ判定処理は、入力画像に対し全てのデータが終了するまで実行される（ステップＳ９）。
【０１４２】
ここで、上記色ずれ補正部２３内のフィッティング処理部３２におけるフィッティング処理について以下に説明する。
【０１４３】
まず、カラー信号におけるフィッティング処理について、図６（ａ）を参照しながら以下に説明する。ここで、カラー信号とは、ＲＧＢの色信号を示す。
【０１４４】
図６（ａ）は、所定の１つのカラー信号における画素位置と画素濃度の関係と、関数によるフィッティング結果とを示すグラフである。
【０１４５】
今、注目画素位置をｔ、画素濃度をＤとすると、関数でフィッティングさせた結果は、Ｄ＝１３２ｃｏｓ⁴⁰⁰⁰（ｔ）となる。本実施の形態では、１画素をπ／１８０［ｒａｄ］（≒０．０１７５［ｒａｄ］）としてフィッティングを行った。もし、注目画素位置ｔ＝０とし、注目画素位置において「０．１画素以上ずれているときは色ずれが発生している」と判定するならば、図５に示す色ずれ補正処理の流れにおいては、α＝０．００１７５という閾値を設ければよいことになる。すなわち、α≧０．００１７５の時、色ずれが生じていると判定されることになる。
【０１４６】
ここでは、１つのカラー信号について説明しているが、他のカラー信号についても同様のフィッティングを行い、得られたパラメータにより図５で示した色ずれ判定を行う。色ずれが発生しないときは、３つのカラー信号をフィッティングさせたときのパラメータαはほぼ一致することとなる。
【０１４７】
次に、色ずれが発生したときにおけるフィッティング処理を、図６（ｂ）を参照しながら以下に説明する。
【０１４８】
図６（ｂ）は、色ずれが発生した際の３つのカラー信号における画素位置と画素濃度の関係と、関数によるフィッティング結果とを示すグラフである。このグラフは、万線チャートをスキャナで読み込んだものであるが、黒い線画あるいは文字領域においては、色ずれをパラメータから確認することができる。
【０１４９】
続いて、太い文字・線画領域におけるフィッティング処理を、図７（ａ）（ｂ）および図８を参照しながら以下に説明する。
【０１５０】
図７（ａ）（ｂ）は、太い文字あるいは太い線画領域における画素位置と画素濃度の関係（図７（ａ））と、無彩色領域を除いて、その両端の領域ｘとｙをつなぎ、関数を利用したフィッティングを行った結果（図７（ｂ））とを表している。
【０１５１】
この結果、無彩色領域は、マスク領域内における隣り合うデータの濃度レベルの差分情報を取ったときにほぼ０になる領域とみなすことができる。この処理により、太い文字または線画領域においても小さな文字や細線の場合と同様の色ずれ判定を実行することが可能となる。
【０１５２】
図８は、太い文字あるいは太い線画領域において、小さな文字や細線の場合と同様の色ずれ判定を実行することが可能となる２番目の手法を示している。この手法でも、無彩色領域はマスク領域内における隣り合うデータの濃度レベルの差分情報から決定することができる。ここでは、画素濃度が飽和する画素までのデータを活用し、フィッティングを行う。これにより画像端部の領域をつなぐ手法（図７（ａ）（ｂ））を使うことなくフィッティングをすることが可能となる。
【０１５３】
図１に示すカラー画像処理装置２では、色ずれ補正を行った後、領域分離を行うようになっていたが、先ず、領域分離を行い、その後色ずれ補正を行うように構成してもよい。
【０１５４】
すなわち、領域分離情報を利用し、色ずれ補正を行う場合には、例えば図９に示すようなカラー画像処理装置４０が用いられる。なお、説明の便宜上、図１に示すカラー画像処理装置２における構成部材と同一機能を有する部材には、同一符号を付記し、その説明を省略する。
【０１５５】
上記カラー画像処理装置４０は、図９に示すように、図１に示すカラー画像処理装置２と、色ずれ補正部２３の配置位置が異なるだけで、他は全て同じである。すなわち、カラー画像処理装置４０は、色ずれ補正部２３の前段に領域分離処理部２４を配置した構造となっている。
【０１５６】
上記領域分離処理部２４は、ＲＧＢ信号より、入力画像中の各画素を文字または線画領域、網点領域、写真領域の何れかに分離するものであり、色ずれが目立ちやすくまた判定が容易であり、黒文字あるいは黒の線画領域を抽出して色ずれ補正部２３にて色ずれ補正を行うことも可能である。これにより所定の領域のみを抽出して色ずれ補正処理を行うので、処理速度の向上が期待される。色ずれ補正が施されたＲＧＢ信号は色補正部２５へ出力される。
【０１５７】
領域分離処理部２４における領域分離方法としては、例えば．「画像電子学会研究会予稿９０’０６’０４」に記載されている方法を用いることができる。以下に詳細を説明する。この方法では、注目画素を中心としたＭ×Ｎ（Ｍ、Ｎは自然数）画素のブロック内で以下のような一連の判定を行い、これらの判定結果を注目画素の領域識別信号とする。
【０１５８】
１．ブロック内の中央の９画素に対して信号レベルの平均値（Ｄ_ave ）を求め、その平均値を用いてブロック内の各画素を２値化する。また、最大画素信号レベル（Ｄ_max ）、最小画素信号レベル（Ｄ_min ）も同時に求める。
【０１５９】
２．網点領域では、小領域における画像信号の変動が大きいことや、背景に比べて濃度が高いことを利用し、網点領域を識別する。２値化されたデータに対して主走査、副走査方向でそれぞれ０から１への変化点数、１から０への変化点数を求めて、それぞれＫ_H 、Ｋ_V とし、閾値Ｔ_H 、Ｔ_V と比較して両者が共に閾値を上回ったら網点領域とする。また、背景との誤判定を防ぐために、Ｄ_max 、Ｄ_min 、Ｄ_ave を閾値Ｂ₁ 、Ｂ₂ と比較する。即ち、以下の条件で、網点領域と非網点領域とを分ける。
【０１６０】
【数２】

【０１６１】
３．文字領域では、最大信号レベルと最小信号レベルの差が大きく、濃度も高いと考えられることから、文字領域の識別を以下のように行う。非網点領域において先に求めていた最大、最小信号レベルとそれらの差分（Ｄ_sub ）を閾値Ｐ_A 、Ｐ_B 、Ｐ_C と比較し、どれか一つが上回ったならば文字領域、すべて閾値以下ならば写真領域とする。即ち、以下の条件で、非網点領域を、文字領域と写真領域とに分ける。
【０１６２】
【数３】

【０１６３】
以上のように、上記構成のデジタルカラー複写機によれば、カラー画像入力装置１により入力された画像データに生じる色ずれを、入力画像データ毎に判定し、この判定結果に基づいて、色ずれ補正を行うようになっているので、従来のように、予め設定した基準データ（色ずれ量）を用いた色ずれ補正の場合に比べて、手間が掛からず、精度よく色ずれ補正を行うことができる。
【０１６４】
また、本実施の形態においては、上述したような画像処理方法以外に以下の方法であっても同様の効果を奏することができる。
【０１６５】
すなわち、入力画像データの複数の色成分が、任意の画素位置において色ずれを生じていると判定された時、上記色ずれの補正処理を行う画像処理方法において、上記画像処理方法は、入力画像データの複数の色成分毎に各画素位置に対する濃度分布を関数を用いて近似する第１の工程と、上記複数の色成分毎に求められた近似式を比較し、色ずれが生じているか否かを判定する第２の工程と、色ずれが生じていると判定された時、この色ずれの補正処理を行う第３の工程とを含んでいてもよい。
【０１６６】
このように、関数によるフィッティングを行うことにより、あらかじめ規準データを読み取ることなく精度良く色ずれの有無を判定することができる。
【０１６７】
しかも、色ずれは入力画像における各色成分の画素位置がずれることによって生じるので、一次情報である各色成分の画素位置に基づいて、色ずれが生じているか否かを判定することによって、高い精度で色ずれ判定および色ずれ補正を行うことができる。
【０１６８】
これは、入力画像毎に、各色成分の画素位置に基づいて、色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行い、画像処理を行っているからである。
【０１６９】
本発明の別の実施の一形態について説明すれば、以下の通りである。
【０１７０】
本実施の形態に係るカラー画像処理装置５０を図１０に示す。なお、説明の便宜上、図１に示すカラー画像処理装置２における構成部材と同一機能を有する部材には、同一符号を付記し、その説明を省略する。
【０１７１】
上記カラー画像処理装置５０は、図１０に示すように、図１に示すカラー画像処理装置２と、色ずれ補正部５１が異なるだけで、他は全て同じである。すなわち、カラー画像処理装置５０は、色ずれ補正部２３の代わりに色ずれ補正部５１を有する構造となっている。
【０１７２】
上記色ずれ補正部５１は、図１１に示すように、画像メモリ３０、画素濃度判定部（画素濃度判定手段）３１、画素位置推定部（画素位置推定手段）５２、色ずれ判定部（色ずれ判定手段）５３、補正処理部（補正手段）３６、フィルタ記憶部３７を含み、外部の制御部３８により駆動制御されるようになっている。なお、説明の便宜上、図３に示す色ずれ補正部２３における構成部材と同一機能を有する部材である、画像メモリ３０、画素濃度判定部３１、補正処理部３６、フィルタ記憶部３７、および制御部３８には、同一符号を付記し、その説明を省略する。
【０１７３】
上記色ずれ補正部５１は、入力階調補正部２２（図１０）からのＲＧＢ信号（入力画像データ）に対して、画像メモリ３０、画素濃度判定部３１、画素位置推定部５２、色ずれ判定部５３、補正処理部３６による色ずれ補正処理を施し、色ずれ補正済の画像データとしてのＲＧＢ信号を後段の領域分離処理部２４（図１０）に出力する一方、黒生成調整信号を後段の黒生成下色除去部２６（図１０）に出力するようになっている。
【０１７４】
色ずれ補正部５１の処理について具体的に説明すると以下のようになる。
【０１７５】
入力階調補正部２２より送られてきたＲＧＢ信号（入力画像データ）に対してｍ×ｎの複数の画素よりなる画像データが画像メモリ３０に格納される（ｍ，ｎは正の整数である。後述するように、例えば１×７マスク領域の画像データを用いて画素濃度の判定を行う場合は、少なくとも３×７の画像データが格納される。）。
【０１７６】
次に、画像メモリ３０より、例えば１×７マスク領域の画像データが抽出されて、画素濃度判定部３１で濃度情報に基づく極大点の有無が判定される。ここで、抽出された画像データに極大点が存在すると判定された場合は１×７マスク領域のＲＧＢ信号は画素位置推定部５２に出力される。
【０１７７】
なお、上記画素濃度判定部３１における画素濃度判定は必ずしも行わなくてもよいが（前述したように、色ずれ補正部５１を領域分離処理部２４の後段に配置する場合は必要ない）、この判定を行うことにより、文字や線画等の画像領域を速やかに見いだすことができ、極大点が存在しないと判定された場合は、その領域が下地領域あるいは均一な濃度の画像領域であると判断し、色ずれ判定を行う必要がないため、以降の処理の高速化が期待される。
【０１７８】
画素位置推定部５２においては、入力されたＲＧＢ信号（１×７マスク領域のデータ）に対し、全色成分の平均濃度を算出し、そのときのＲＧＢ信号それぞれの画素位置β_R 、β_G 、β_B を推定する。なお、平均濃度をＲＧＢ信号ごとに算出し、各信号の平均濃度からβ_R 、β_G 、β_B を推定してもよい。
【０１７９】
各信号の平均濃度の画素位置β_R 、β_G 、β_B を推定する際には、画素濃度判定部３１より送られてくるＲＧＢ信号の平均濃度における画素位置を、この画素位置に隣接する画素の画素濃度を用いて線形近似することにより推定する。なお、ＲＧＢ信号の平均濃度における画素位置β_R 、β_G 、β_B を推定する方法の詳細については後述する。
【０１８０】
また、原稿からの反射光像が光学レンズを通る際に波長の違いに起因し、入力されたＲＧＢ信号のバランスが大きく崩れている場合がある。この場合には、濃度が低い色成分の信号に対し、濃度バランス補正を行った後に、全色成分の平均濃度を算出する。なお、この濃度バランス補正の詳細については後述する。
【０１８１】
色ずれ判定部５３では、画素位置推定部５２において得られた、ＲＧＢ信号それぞれの画素位置β_R 、β_G 、β_B が略一致するかどうかを判定する。色ずれ領域と判定された場合には、補正処理部３６へ送られる。一方、色ずれが発生していないと判定されれば、画像メモリ３０により送られてくるＲＧＢ信号は、そのまま色ずれ補正部５１の後段の領域分離処理部２４に送られる。色ずれ判定処理の流れについては、図１２を用いて後に説明する。
【０１８２】
補正処理部３６では、色ずれ領域と判定された領域に対し、例えば、フィルタ記憶部３７で記憶されているフィルタを用いて空間フィルタ処理を行い、色ずれを補正する。あるいは、図１０に示すように黒生成下色除去部２６に黒生成量を制御する場合には、補正処理部３６で黒生成調整信号を黒生成下色除去部２６に出力する。
【０１８３】
次に、上記色ずれ補正部５１における色ずれ判定処理の流れについて、図１２に示すフローチャートを参照しながら以下に説明する。
【０１８４】
まず、画像入力が行われ、該入力画像から入力階調補正部２２よりＲＧＢ信号を得る（ステップＳ１１）。
【０１８５】
次に、得られたＲＧＢ信号より所定のサイズのデータが画像メモリ３０に格納され、１×７マスク領域のデータが抽出される（ステップＳ１２）。
【０１８６】
続いて、抽出されたデータに対して、画素濃度判定部３１で濃度情報に基づく極大点の有無の判定がなされる（ステップＳ１３）。ここで、極大点の判定方法の１つとして、マスク領域内における隣り合うデータ（画素濃度）の差分情報を求め単調増加から単調減少、あるいは、ほとんど差が見られない状態へと変化する点があるか否かを調べる方法が挙げられる。
【０１８７】
ステップＳ１３において、『極大点が存在する』と判定された場合は、１×７マスク領域のデータに対し、平均濃度を算出し、そのときの画素位置β_R 、β_G 、β_B を推定する（ステップＳ１４）。
【０１８８】
平均濃度の画素位置β_R 、β_G 、β_B の推定には、平均濃度に対応する位置に隣接する画素の画素位置と画素濃度とから、線形補間により求める。そして、ＲＧＢ信号の平均濃度における画素位置β_R 、β_G 、β_B より色ずれ量を求めることができる。
【０１８９】
ここで、１×７マスク領域のデータに対し、ＲＧＢ信号のバランスが大きく崩れているときには、まず濃度バランス補正処理を行い（濃度バランス補正手段）、平均濃度を算出する。この場合、図１３に示すような濃度バランス補正処理ルーチンを、図１２に示すフローチャートのステップＳ１３とステップＳ１４との間に挿入する。
【０１９０】
図１３では、例えば、１×７マスク領域における各ＲＧＢ信号の最も高い濃度値をmax Ｒ、max Ｇ、max Ｂとし、最も低い濃度値をmin Ｒ、min Ｇ、min Ｂとする。最大濃度差の閾値をＴと設定し、次式により濃度バランス補正処理の必要の有無を調べる（ステップＳ１９）。
【０１９１】

上記条件式を満たすならば、ステップＳ２０に移行し、濃度バランス補正処理を行い、満たさなければ、ステップＳ１４に移行し、濃度バランス補正処理を行わずに画素位置推定処理を行う。
【０１９２】
濃度バランス補正処理は、１×７マスク領域において最も濃度の高い色成分の信号の最大値と、上記条件式を満たす色成分の信号の最大値を比較して補正倍率を求め、図１４に示すように、この補正倍率を上記条件式を満たす色成分の信号の各データに乗算することにより行う。
【０１９３】
図１５は、ＲＧＢ信号における画素位置と画素濃度の関係、および線形近似により平均濃度の画素位置を推定する方法を示す図である。画素濃度は各画素毎の離散データであるが、ここでは理解しやすくするために、画素間のデータを補間し曲線として示してある。
【０１９４】
例えば、モノクロの直線を読み取った場合、平均濃度Ａｖｅに近い画素濃度を有し、かつ隣接するＧ信号の画素位置をＰ、Ｐ＋１、それぞれの位置におけるＧ信号の画素濃度をＧ_P 、Ｇ_P+1 とする。このとき、Ｇ信号における平均濃度画素位置β_G は、次式で近似的に表すことができる。
【０１９５】
【数４】

【０１９６】
ここでは、一つのカラー信号についてのみ説明しているが、他のカラー信号であるＲ信号およびＢ信号についても、同様にして、β_R 、β_B を算出する。
【０１９７】
色ずれが発生しないときには、３つのＲＧＢ信号の平均濃度画素位置β_R 、β_G 、β_B は略一致する。よって、例えば、
max （｜β_R −β_G ｜，｜β_G −β_B ｜，｜β_B −β_R ｜）＜０．５
と設定し、上記の条件を満たすか否かにより色ずれの有無を判定する。そして、この条件を満たすとき、
β_R ＝β_G ＝β_B
であるとし、色ずれが発生していないと判定する。したがって、図１２のフローチャートのステップＳ１５において、β_R ＝β_G ＝β_B であるか否かを判定する。
【０１９８】
ここで、図１２に示すフローチャートで示される色ずれ判定処理において、色ずれ補正を行うか否かの条件として、以下の３つの条件を考える。
【０１９９】
＜条件１＞
極大点が存在し、「β_R ＝β_G ＝β_B 」を満たすときは、色ずれが発生していないと判定する。つまり、ステップＳ１３において、極大点を有すると判定され、ステップＳ１５において、β_R ＝β_G ＝β_B を満たしていると判定されたとき、色ずれが発生していないと判定する。
【０２００】
＜条件２＞
極大点が存在し、「β_R ＝β_G ＝β_B 」を満たさないときは、色ずれが発生していると判定し、色ずれ補正処理を行う（ステップＳ１６）。つまり、ステップＳ１３において、極大点を有すると判定され、ステップＳ１５において、β_R ＝β_G ＝β_B を満たしていないと判定されたとき、色ずれが発生していると判定する。
【０２０１】
＜条件３＞
極大点が存在しないときは、文字または線画領域ではない為、色ずれの影響がない、あるいは目立たないと判定し、色ずれ補正処理は行わない。つまり、ステップＳ１３において、極大点を有しないと判定されたときには、色ずれの影響がない、あるいは目立たないと判定する。
【０２０２】
上記の条件２の場合、すなわち入力画像に色ずれが生じていると判定された場合には、ステップＳ１６において、色ずれ補正処理、すなわち図１０に示す補正処理部３６における補正処理が実行される。この補正処理については、前述したフィルタ記憶部３７に記憶されたフィルタを用いて行われる。
【０２０３】
そして、上記の色ずれ判定処理は、入力画像に対し全てのデータが終了するまで実行される（ステップＳ１８）。
【０２０４】
なお、図１０に示すカラー画像処理装置５０では、色ずれ補正を行った後、領域分離を行うようになっていたが、先ず、領域分離を行い、その後色ずれ補正を行うように構成してもよい。
【０２０５】
【発明の効果】
本発明の画像処理方法は、以上のように、入力画像の色ずれを補正して出力画像を得る画像処理方法において、入力画像毎に、各色成分の画素位置に基づいて、色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行う構成である。
【０２０６】
それゆえ、入力画像毎に色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行うことで、入力画像毎に色ずれ補正を行うことができる。
【０２０７】
このように、入力画像毎に色ずれを判定し、得られた色ずれ判定結果に基づいて色ずれ補正を行うことにより、スキャナ等の駆動系における振動の経時変化が生じても、色ずれ補正の精度が低下することはなく、常に高い精度で色ずれ補正を行うことができる。
【０２０８】
また、従来のように、基準原稿である万線チャートまたは基準パターンを使用する必要はないので、これらを使用する手間を省くことができる。
【０２０９】
さらに、上述のように、入力画像毎に色ずれの有無を判定するようにすれば、色ずれのない場合には、色ずれ補正処理を行わないで画像処理を行うことができるので、従来のように、入力された画像に対して、色ずれの有無に関わり無く色ずれ処理が常に行われる場合に比べて、画像処理に係る時間を大幅に短くすることができる。
【０２１０】
しかも、色ずれは入力画像における各色成分の画素位置がずれることによって生じるので、一次情報である各色成分の画素位置に基づいて、色ずれが生じているか否かを判定することによって、高い精度で色ずれ判定および色ずれ補正を行うことができるという効果を奏する。
【０２１１】
本発明の画像処理方法は、また、以上のように、入力画像の色ずれを補正して出力画像を得る画像処理方法において、入力画像毎に、各色成分の各画素位置の位置ずれ量を検出し、この検出結果に基づいて色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行う構成である。
【０２１２】
それゆえ、入力画像毎に色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行うことで、入力画像毎に色ずれ補正を行うことができる。
【０２１３】
このように、入力画像毎に色ずれを判定し、得られた色ずれ判定結果に基づいて色ずれ補正を行うことにより、スキャナ等の駆動系における振動の経時変化が生じても、色ずれ補正の精度が低下することはなく、常に高い精度で色ずれ補正を行うことができる。
【０２１４】
さらに、上述のように、入力画像毎に色ずれの有無を判定するようにすれば、色ずれのない場合には、色ずれ補正処理を行わないで画像処理を行うことができるので、従来のように、入力された画像に対して、色ずれの有無に関わり無く色ずれ処理が常に行われる場合に比べて、画像処理に係る時間を大幅に短くすることができる。
【０２１５】
しかも、また、入力画像の各色成分の各画素位置の位置ずれ量を検出することで、色ずれが生じているか否かを判定するようになっているので、簡単に色ずれを判定することができる。
【０２１６】
また、従来のように、基準原稿である万線チャートまたは基準パターンを使用する必要はないので、これらを用いた場合のように線形補間を使用することによる誤差を考慮しないで精度の高い色ずれ補正を行うことができるという効果を奏する。
【０２１７】
さらに、入力画像の色成分毎に、各画素位置に対する濃度分布を関数により近似し、この近似結果に基づいて、色ずれが生じているか否かを判定する構成であってもよい。
【０２１８】
この場合、入力画像の色成分毎に、各画素位置に対する濃度分布を関数により近似することで、入力画像の各色成分の画素位置のずれ量を正確に求めることができる。
【０２１９】
したがって、この入力画像の各色成分の画素位置のずれ量に基づいて、入力画像の色ずれを判定すれば、より精度よく、入力画像の色ずれの有無を判定することができる。
【０２２０】
これにより、さらに、色ずれ補正の精度を向上させることができ、この結果、入力画像を忠実に再現した出力画像を得ることができるという効果を奏する。
【０２２１】
本発明の画像処理方法は、また、以上のように、入力画像の色ずれを補正して出力画像を得る画像処理方法において、入力画像毎に、各色成分の平均濃度における画素位置を推定し、この推定結果に基づいて、色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行う構成である。
【０２２２】
それゆえ、入力画像毎に色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行うことで、入力画像毎に色ずれ補正を行うことができる。
【０２２３】
このように、入力画像毎に色ずれを判定し、得られた色ずれ判定結果に基づいて色ずれ補正を行うことにより、スキャナ等の駆動系における振動の経時変化が生じても、色ずれ補正の精度が低下することはなく、常に高い精度で色ずれ補正を行うことができる。
【０２２４】
さらに、上述のように、入力画像毎に色ずれの有無を判定するようにすれば、色ずれのない場合には、色ずれ補正処理を行わないで画像処理を行うことができるので、従来のように、入力された画像に対して、色ずれの有無に関わり無く色ずれ処理が常に行われる場合に比べて、画像処理に係る時間を大幅に短くすることができる。
【０２２５】
しかも、また、入力画像の各色成分の平均濃度における画素位置を推定することで、色ずれが生じているか否かを判定するようになっているので、簡単に色ずれを判定することができるという効果を奏する。
【０２２６】
さらに、入力画像の各色成分の濃度バランスが、予め定められる範囲を越えていると判断したとき、上記各色成分の濃度バランスが、予め定められた範囲に収まるように補正する構成であってもよい。
【０２２７】
この場合、さらに、入力画像の各色成分の平均濃度における画素位置を推定する前に、入力画像の各色成分の濃度バランスが、予め定められる範囲を越えていると判断した時、補正を行うことにより、精度よく画素位置を推定することができる。
【０２２８】
これにより、さらに、色ずれ補正の精度を向上させることができ、この結果、入力画像を忠実に再現した出力画像を得ることができるという効果を奏する。
【０２２９】
本発明の画像処理装置は、以上のように、入力画像に生じた色ずれを補正する色ずれ補正手段を備えた画像処理装置において、上記色ずれ補正手段は、入力画像の色成分毎の画素位置に基づいて、色ずれが生じているか否かを判定する色ずれ判定手段と、上記色ずれ判定手段により色ずれが生じていると判定されたとき、該色ずれを補正する補正手段とを備えている構成である。
【０２３０】
それゆえ、入力画像毎に色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行うことで、入力画像毎に色ずれ補正を行うことができる。
【０２３１】
このように、入力画像毎に色ずれを判定し、得られた色ずれ判定結果に基づいて色ずれ補正を行うことにより、スキャナ等の駆動系における振動の経時変化が生じても、色ずれ補正の精度が低下することはなく、常に高い精度で色ずれ補正を行うことができる。
【０２３２】
また、従来のように、基準原稿である万線チャートまたは基準パターンを使用する必要はないので、これらを使用する手間を省くことができる。
【０２３３】
さらに、上述のように、入力画像毎に色ずれの有無を判定するようにすれば、色ずれのない場合には、色ずれ補正処理を行わないで画像処理を行うことができるので、従来のように、入力された画像に対して、色ずれの有無に関わり無く色ずれ処理が常に行われる場合に比べて、画像処理に係る時間を大幅に短くすることができる。
【０２３４】
しかも、色ずれは入力画像における各色成分の画素位置がずれることによって生じるので、一次情報である各色成分の画素位置に基づいて、色ずれが生じているか否かを判定することによって、高い精度で色ずれ判定および色ずれ補正を行うことができるという効果を奏する。
【０２３５】
本発明の画像処理装置は、また、以上のように、入力画像に生じた色ずれを補正する色ずれ補正手段を備えた画像処理装置において、上記色ずれ補正手段は、入力画像の色成分毎に、各画素位置に対する濃度分布を関数により近似する近似手段と、上記近似手段による近似結果に基づいて、色ずれが生じているか否かを判定する色ずれ判定手段と、上記色ずれ判定手段により色ずれが生じていると判定されたとき、該色ずれを補正する補正手段とを備えている構成である。
【０２３６】
それゆえ、入力画像毎に色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行うことで、入力画像毎に色ずれ補正を行うことができる。
【０２３７】
このように、入力画像毎に色ずれを判定し、得られた色ずれ判定結果に基づいて色ずれ補正を行うことにより、スキャナ等の駆動系における振動の経時変化が生じても、色ずれ補正の精度が低下することはなく、常に高い精度で色ずれ補正を行うことができる。
【０２３８】
さらに、上述のように、入力画像毎に色ずれの有無を判定するようにすれば、色ずれのない場合には、色ずれ補正処理を行わないで画像処理を行うことができるので、従来のように、入力された画像に対して、色ずれの有無に関わり無く色ずれ処理が常に行われる場合に比べて、画像処理に係る時間を大幅に短くすることができる。
【０２３９】
しかも、また、入力画像の色成分毎に、各画素位置に対する濃度分布を関数により近似することで、入力画像の各色成分の画素位置のずれ量を正確に求めることができる。
【０２４０】
したがって、この入力画像の各色成分の画素位置のずれ量に基づいて、入力画像の色ずれを判定すれば、より精度よく、入力画像の色ずれの有無を判定することができる。
【０２４１】
また、従来のように、基準原稿である万線チャートまたは基準パターンを使用する必要はないので、これらを用いた場合のように線形補間を使用することによる誤差を考慮しないで精度の高い色ずれ補正を行うことができるという効果を奏する。
【０２４２】
本発明の画像処理装置は、また、以上のように、入力画像に生じた色ずれを補正する色ずれ補正手段を備えた画像処理装置において、上記色ずれ補正手段は、入力画像の色成分毎の濃度分布から平均濃度の画素位置を推定する画素位置推定手段と、上記画素位置推定手段の推定結果に基づき、色ずれが生じているか否かを判定する色ずれ判定手段と、上記色ずれ判定手段により色ずれが生じていると判定されたとき、該色ずれを補正する補正手段とを備えている構成である。
【０２４３】
それゆえ、入力画像毎に色ずれが生じているか否かを判定し、色ずれが生じていると判定した場合に色ずれ補正を行うことで、入力画像毎に色ずれ補正を行うことができる。
【０２４４】
このように、入力画像毎に色ずれを判定し、得られた色ずれ判定結果に基づいて色ずれ補正を行うことにより、スキャナ等の駆動系における振動の経時変化が生じても、色ずれ補正の精度が低下することはなく、常に高い精度で色ずれ補正を行うことができる。
【０２４５】
さらに、上述のように、入力画像毎に色ずれの有無を判定するようにすれば、色ずれのない場合には、色ずれ補正処理を行わないで画像処理を行うことができるので、従来のように、入力された画像に対して、色ずれの有無に関わり無く色ずれ処理が常に行われる場合に比べて、画像処理に係る時間を大幅に短くすることができる。
【０２４６】
しかも、また、入力画像の色成分毎の濃度分布から平均濃度の画素位置を推定することができる。
【０２４７】
したがって、この入力画像の各色成分毎の平均濃度の画素位置に基づいて、入力画像の色ずれを判定すれば、より精度よく、入力画像の色ずれの有無を判定することができるという効果を奏する。
【０２４８】
上記画素位置推定手段には、さらに、入力画像の各色成分の濃度バランスが予め定められる範囲を越えていると判断したとき、上記各色成分の濃度バランスが、予め定められた範囲に収まるように補正する濃度バランス補正手段を設ける構成であってもよい。
【０２４９】
この場合、さらに、入力画像の各色成分の平均濃度における画素位置を推定する前に、入力画像の各色成分の濃度バランスが、予め定められる範囲を越えていると判断した時、補正を行うことにより、精度よく画素位置を推定することができる。
【０２５０】
これにより、さらに、色ずれ補正の精度を向上させることができ、この結果、入力画像を忠実に再現した出力画像を得ることができるという効果を奏する。
【０２５１】
上記色ずれ補正手段には、さらに各画素毎の濃度を基に所定の領域内の濃度分布を求め、この濃度分布に基づいて、色ずれ判定の対象となる領域であるか否かを判定する画素濃度判定手段を、上記色ずれ判定手段の前段に設ける構成であってもよい。
【０２５２】
この場合、色ずれ判定手段の前段に画素濃度判定手段が設けられていることにより、指定された領域でのみ色ずれ判定を行えばよいことになり、色ずれ補正処理の処理速度を向上させることができるという効果を奏する。
【０２５３】
また、入力画像を文字・網点・写真の各領域に分離する領域分離手段を備え、上記色ずれ補正手段を、上記領域分離手段の前段に設ける構成であってもよい。
【０２５４】
この場合、領域分離手段による入力画像の領域分離の前に、色ずれ補正手段により、予め色ずれ補正処理を行うことにより、該領域分離手段による領域分離処理における領域分離の誤判定を防ぎ、領域分離処理における領域分離の精度を向上させることができるという効果を奏する。
【０２５５】
また、上記色ずれ補正手段を、上記領域分離手段の後段に設ける構成であってもよい。
【０２５６】
この場合、領域分離手段により領域分離された領域、すなわち抽出された文字または線画領域に対し、色ずれ補正手段による色ずれ補正処理を行うことにより、文字または線画領域周辺に発生する色ずれを検知・補正することが可能となる。
【０２５７】
特に、色ずれが目立ち易い文字や線画領域を領域分離処理により抽出し、色ずれ補正処理を行うので、処理の高速化を図ることができるという効果を奏する。
【０２５８】
また、原稿の予備走査時に読み込まれた入力画像に対して、上記領域分離手段による領域分離処理を行わせ、色ずれが生じているか否かを判定し、色ずれが生じていないと判定されたときに、該入力画像に対する色ずれ補正処理を行わないように上記色ずれ補正手段を制御する制御手段を設ける構成であってもよい。
【０２５９】
この場合、予備走査時に読み込まれる低解像度の入力画像にて大略的に色ずれの有無の判定を行い、色ずれが生じていると判定された時にのみ、色ずれ補正手段による色ずれ補正処理を行うようになるので、無駄なく効率よく画像処理を行うことができるという効果を奏する。
【０２６０】
本発明の画像形成装置は、以上のように、入力画像に対して所定の処理を施して出力画像を形成するものであって、上述した本発明の画像処理装置を備えている構成である。
【０２６１】
それゆえ、上記した画像処理装置によれば、色ずれが生じているか否かを精度よく判定し、色ずれ補正を行うことができるので、高品質の画像を出力することができる画像形成装置を提供することができるという効果を奏する。
【０２６２】
この画像形成装置としては、例えば電子写真方式やインクジェット方式を用いたカラー画像形成装置等が挙げられる。また、画像処理後の画像を表示できる画像表示装置、例えばコンピュータや携帯情報機器の表示手段として用いられる液晶ディスプレイ等であってもよい。
【図面の簡単な説明】
【図１】本発明の画像処理装置を備えた画像形成装置としてのデジタルカラー複写機の概略構成ブロック図である。
【図２】図１に示すデジタルカラー複写機に備えられたカラー画像入力装置を示す概略構成図である。
【図３】図１に示すデジタルカラー複写機に備えられたカラー画像処理装置の色ずれ補正部を示す概略構成ブロック図である。
【図４】図３に示す色ずれ補正部で行われる色ずれ補正処理として、フィルタ処理を行う場合の説明図である。
【図５】図３に示す色ずれ補正部において行われる色ずれ補正処理の流れを示すフローチャートである。
【図６】（ａ）は関数を用いたフィッティングの例を示すグラフであり、（ｂ）は色ずれが生じている時の関数によるフィッティング結果を示したグラフである。
【図７】（ａ）（ｂ）は太い文字や線画領域を関数を用いてフィッティングし、色ずれが生じているか否かを判定する方法を説明するグラフである。
【図８】太い文字や線画領域を関数を用いてフィッティングし、色ずれが生じているか否かを判定する別の方法を説明するグラフである。
【図９】本発明の他のカラー画像処理装置を備えるデジタルカラー複写機の概略構成ブロック図である。
【図１０】本発明の更に別のカラー画像処理装置を備えるデジタルカラー複写機の概略構成ブロック図である。
【図１１】図１０に示すデジタルカラー複写機に備えられたカラー画像処理装置の色ずれ補正部を示す概略構成ブロック図である。
【図１２】図１１に示す色ずれ補正部において行われる色ずれ補正処理の流れを示すフローチャートである。
【図１３】濃度バランス補正処理の流れを示すフローチャートである。
【図１４】濃度バランス補正処理を行う方法を説明するグラフである。
【図１５】平均濃度の画素位置を線形補間により推測し、色ずれ量を求める方法を説明するグラフである。
【符号の説明】
１ カラー画像入力装置
２ カラー画像処理装置（画像処理装置）
３ カラー画像出力装置
２３ 色ずれ補正部（色ずれ補正手段）
２４ 領域分離処理部（領域分離手段）
３１ 画素濃度判定部（画素濃度判定手段）
３２ フィッティング処理部（近似手段）
３４ 色ずれ判定部（色ずれ判定手段）
３６ 補正処理部（補正手段）
３８ 制御部（制御手段）
４０ カラー画像処理装置（画像処理装置）
５０ カラー画像処理装置（画像処理装置）
５１ 色ずれ補正部（色ずれ補正手段）
５２ 画像位置推定部（画像位置推定手段）
５３ 色ずれ判定部（色ずれ判定手段）[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing method, an image processing apparatus, and an image forming apparatus for correcting a color shift generated in an input image.
[0002]
[Prior art]
In recent years, digitalization of office automation equipment has rapidly progressed, and the demand for color image output has increased. As a result, output devices such as electrophotographic digital color copiers and inkjet / thermal transfer color printers have become widely available. It has become widespread.
[0003]
For example, image information input from an input device such as a digital camera or a scanner, or image information created on a computer is output using the output device as described above. In these input / output devices, it is necessary to always output an image with stable color reproduction for input image information, and digital image processing technology plays an important role.
[0004]
In an image input device such as a scanner, for example, three CCD (Charge Coupled Device) line sensors of R (red), G (green), and B (blue) are used, and the positional deviation of each sensor is corrected. Therefore, a method of delaying the outputs of the two preceding sensors and outputting a signal in accordance with the sensor to be read last is used.
[0005]
When vibration occurs in the drive system of such an image input apparatus, the time difference when the three CCD line sensors read the same portion of the document may be increased or decreased by the vibration. As a result, for example, when a black area of an image is read, the position of the image read by the three CCD line sensors may be shifted, and the edge portion may be colored. Such a phenomenon is called image color misregistration.
[0006]
In order to solve such a problem of image color misregistration, it is disclosed in Japanese Patent Application Laid-Open No. 10-42157 (hereinafter referred to as Document 1) and Japanese Patent Application Laid-Open No. 2001-16401 (hereinafter referred to as Document 2). There is a technology.
[0007]
In this document 1, color misregistration of an image is eliminated by the following process.
[0008]
(1) A line chart is read, an intersection with a predetermined threshold value is obtained for each component of the RGB signal, and a relative color shift amount with other colors is calculated using a certain color as reference data.
[0009]
(2) The color misregistration amount thus obtained is stored in the color misregistration amount table in association with the position of the intersection.
[0010]
(3) The color misregistration amount stored in the color misregistration amount table is read together with the target pixel and the adjacent pixel value, and the color misregistration correction calculator calculates the color misregistration correction data value when correcting the color misregistration of the input image. Used for operations.
[0011]
Further, in Document 2, the color misregistration of the image is eliminated through the following process.
[0012]
(1) A plurality of straight line edge portions included in the reference pattern are read, and a relative color shift amount is calculated from the difference between the read output value of the RGB sensor and the average value of the read output values.
[0013]
(2) The read output value of each sensor is corrected in accordance with the color misregistration amount thus obtained.
[0014]
That is, in Document 1 and Document 2, a line chart or reference pattern serving as a reference document is read by a scanner, and a positional shift amount (color shift amount) with respect to each pixel position caused by vibration in the drive system of the scanner is calculated. The color shift at each pixel position of the input image is corrected based on the obtained color shift amount.
[0015]
[Problems to be solved by the invention]
However, since vibrations in a drive system such as a scanner usually change with time, when the amount of color misregistration is obtained in advance and the color misregistration of the input image is corrected based on this amount of color misregistration. In order to perform color misregistration correction with high accuracy, it is necessary to use a line chart or a reference pattern every time in accordance with the change with time of the drive system, which causes a problem that much labor is required.
[0016]
Further, the methods of the above-mentioned literature 1 and literature 2 have no concept of detecting the presence or absence of color shift of the input image, and therefore color shift correction is always performed on the input image. For this reason, even when there is little color misregistration occurring in the input image and color misregistration correction is not necessary, color misregistration correction is performed, and waste is caused in the entire image processing. There arises a problem that the entire processing time becomes long.
[0017]
Furthermore, in Document 1 and Document 2 described above, when color shift correction is performed by reading a line chart or a reference pattern, if the line chart or reference pattern is lost, the color shift can be obtained. The problem of being unable to do so occurs.
[0018]
The present invention has been made to solve the above-described problems, and its purpose is to determine whether or not color misregistration has occurred for each input image, and to perform color misregistration correction based on the determination result. Accordingly, it is an object of the present invention to provide an image processing method, an image processing apparatus, and an image forming apparatus capable of always performing accurate color misregistration correction without taking time and effort.
[0019]
[Means for Solving the Problems]
In order to solve the above problems, the image processing method of the present invention corrects color misregistration of an input image and obtains an output image. It is characterized in that it is determined whether or not there is a shift, and color shift correction is performed when it is determined that a color shift has occurred.
[0020]
According to the above method, it is determined whether or not color misregistration has occurred for each input image, and color misregistration correction is performed for each input image by performing color misregistration correction when it is determined that color misregistration has occurred. It can be carried out.
[0021]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. Therefore, the color misregistration correction can always be performed with high accuracy.
[0022]
Further, unlike the prior art, it is not necessary to use a line chart or a reference pattern which is a reference document, so that it is possible to save time and effort to use these.
[0023]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0024]
In addition, since color misregistration occurs when the pixel position of each color component in the input image deviates, it is determined with high accuracy by determining whether or not color misregistration occurs based on the pixel position of each color component that is primary information. Color misregistration determination and color misregistration correction can be performed.
[0025]
In order to solve the above problems, the image processing method of the present invention is also an image processing method for obtaining an output image by correcting color misregistration of an input image. For each input image, the position of each pixel position of each color component It is characterized in that a shift amount is detected, it is determined whether a color shift has occurred based on the detection result, and a color shift correction is performed when it is determined that a color shift has occurred.
[0026]
According to the above method, it is determined whether or not color misregistration has occurred for each input image, and color misregistration correction is performed for each input image by performing color misregistration correction when it is determined that color misregistration has occurred. It can be carried out.
[0027]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0028]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0029]
In addition, since it is determined whether or not a color shift has occurred by detecting a positional shift amount of each pixel position of each color component of the input image, it is possible to easily determine the color shift. it can.
[0030]
Furthermore, for each color component of the input image, the density distribution with respect to each pixel position may be approximated by a function, and it may be determined whether or not a color shift has occurred based on this approximation result.
[0031]
In this case, by approximating the density distribution for each pixel position with a function for each color component of the input image, the shift amount of the pixel position of each color component of the input image can be accurately obtained.
[0032]
Therefore, if the color shift of the input image is determined based on the shift amount of the pixel position of each color component of the input image, the presence or absence of the color shift of the input image can be determined more accurately.
[0033]
As a result, the accuracy of color misregistration correction can be further improved, and as a result, an output image faithfully reproducing the input image can be obtained.
[0034]
By the way, in the above-mentioned document 1, since the linear interpolation is used when obtaining the intersection between the value obtained by reading the line chart and the threshold value, the accuracy may be lowered due to an error occurring there.
[0035]
Further, in the above-mentioned document 2, since linear interpolation is used when obtaining the deviation amount of the reading position from the discrete data of the deviation amount when the reference pattern is read, the accuracy is reduced due to the error generated there. Conceivable.
[0036]
However, in the above method, the density distribution for each pixel position is approximated by a function for each color component of the input image, and it is determined whether or not color misregistration has occurred based on this approximation result. ing. Therefore, it is not necessary to use a line chart or a reference pattern as a reference document as in the prior art, so that high-accuracy color misregistration can be achieved without considering errors caused by using linear interpolation as in the case of using these. Correction can be performed.
[0037]
In order to solve the above problems, the image processing method of the present invention is also an image processing method for obtaining an output image by correcting a color shift of an input image. For each input image, a pixel position at an average density of each color component. , And based on the estimation result, it is determined whether or not color misregistration has occurred, and color misregistration correction is performed when it is determined that color misregistration has occurred.
[0038]
According to the above method, it is determined whether or not color misregistration has occurred for each input image, and color misregistration correction is performed for each input image by performing color misregistration correction when it is determined that color misregistration has occurred. It can be carried out.
[0039]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. Therefore, the color misregistration correction can always be performed with high accuracy.
[0040]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0041]
In addition, since it is determined whether or not a color shift has occurred by estimating the pixel position at the average density of each color component of the input image, the color shift can be easily determined.
[0042]
Further, when it is determined that the density balance of each color component of the input image exceeds a predetermined range, the density balance of each color component may be corrected so as to be within a predetermined range.
[0043]
In this case, furthermore, before estimating the pixel position at the average density of each color component of the input image, when it is determined that the density balance of each color component of the input image exceeds a predetermined range, correction is performed. The pixel position can be estimated with high accuracy.
[0044]
As a result, the accuracy of color misregistration correction can be further improved, and as a result, an output image faithfully reproducing the input image can be obtained.
[0045]
In order to solve the above problems, the image processing apparatus of the present invention is an image processing apparatus including color misregistration correction means that corrects color misregistration generated in an input image. Color misregistration determining means for determining whether or not color misregistration has occurred based on the pixel position for each component, and correcting the color misregistration when the color misregistration determining means determines that color misregistration has occurred. And a correction means.
[0046]
According to the above configuration, it is determined whether or not color misregistration occurs for each input image, and color misregistration correction is performed for each input image by performing color misregistration correction when it is determined that color misregistration occurs. It can be carried out.
[0047]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0048]
Further, unlike the prior art, it is not necessary to use a line chart or a reference pattern which is a reference document, so that it is possible to save the trouble of using these.
[0049]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0050]
In addition, since color misregistration is caused by the displacement of the pixel position of each color component in the input image, it is determined with high accuracy by determining whether or not color misregistration has occurred based on the pixel position of each color component as primary information. Color misregistration determination and color misregistration correction can be performed.
[0051]
In order to solve the above-described problems, the image processing apparatus of the present invention is an image processing apparatus including color misregistration correction means for correcting color misregistration generated in an input image. For each color component, approximation means for approximating the density distribution for each pixel position with a function, color deviation determination means for determining whether color deviation has occurred based on an approximation result by the approximation means, and the color And a correction unit that corrects the color shift when it is determined by the shift determination unit that the color shift has occurred.
[0052]
According to the above configuration, it is determined whether or not color misregistration occurs for each input image, and color misregistration correction is performed for each input image by performing color misregistration correction when it is determined that color misregistration occurs. It can be carried out.
[0053]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0054]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0055]
In addition, by approximating the density distribution for each pixel position with a function for each color component of the input image, the shift amount of the pixel position of each color component of the input image can be accurately obtained.
[0056]
Therefore, if the color shift of the input image is determined based on the shift amount of the pixel position of each color component of the input image, the presence or absence of the color shift of the input image can be determined with higher accuracy.
[0057]
In addition, since it is not necessary to use a line chart or a reference pattern as a reference document as in the past, high-accuracy color misregistration does not take into account errors caused by using linear interpolation as in the case of using these. Correction can be performed.
[0058]
In order to solve the above-described problems, the image processing apparatus of the present invention is an image processing apparatus including color misregistration correction means for correcting color misregistration generated in an input image. Pixel position estimating means for estimating the pixel position of the average density from the density distribution for each color component, color misregistration determining means for determining whether color misregistration has occurred based on the estimation result of the pixel position estimating means, And a correction unit that corrects the color shift when the color shift determination unit determines that the color shift has occurred.
[0059]
According to the above configuration, it is determined whether or not color misregistration occurs for each input image, and color misregistration correction is performed for each input image by performing color misregistration correction when it is determined that color misregistration occurs. It can be carried out.
[0060]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0061]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0062]
In addition, the pixel position of the average density can be estimated from the density distribution for each color component of the input image.
[0063]
Therefore, if the color shift of the input image is determined based on the pixel position of the average density for each color component of the input image, the presence or absence of the color shift of the input image can be determined more accurately.
[0064]
The pixel position estimating means further corrects the density balance of each color component to fall within a predetermined range when it is determined that the density balance of each color component of the input image exceeds a predetermined range. A density balance correction unit may be provided.
[0065]
In this case, furthermore, before estimating the pixel position at the average density of each color component of the input image, when it is determined that the density balance of each color component of the input image exceeds a predetermined range, correction is performed. The pixel position can be estimated with high accuracy.
[0066]
As a result, the accuracy of color misregistration correction can be further improved, and as a result, an output image faithfully reproducing the input image can be obtained.
[0067]
The color misregistration correction means further obtains a density distribution in a predetermined area based on the density for each pixel, and determines whether or not the area is a target of color misregistration determination based on the density distribution. Pixel density determination means may be provided in the preceding stage of the color misregistration determination means.
[0068]
In this case, since the pixel density determination unit is provided before the color misregistration determination unit, it is only necessary to perform the color misregistration determination in the designated region, and the processing speed of the color misregistration correction processing is improved. Can do.
[0069]
Further, an area separation unit that separates the input image into character, halftone dot, and photo areas may be provided, and the color misregistration correction unit may be provided in the preceding stage of the area separation unit.
[0070]
In this case, by performing color misregistration correction processing by the color misregistration correction unit in advance before the region separation of the input image by the region separation unit, erroneous determination of region separation in the region separation processing by the region separation unit is prevented, The accuracy of region separation in the separation process can be improved.
[0071]
Further, the color misregistration correction means may be provided at the subsequent stage of the area separation means.
[0072]
In this case, color misalignment that occurs around the character or line drawing area is detected by performing color misregistration correction processing by the color misregistration correcting means on the area separated by the area separating means, that is, the extracted character or line drawing area.・ It can be corrected.
[0073]
In particular, since characters and line drawing areas where color misregistration is conspicuous are extracted by area separation processing and color misregistration correction processing is performed, the processing speed can be increased.
[0074]
Also, the input image read during the preliminary scan of the document is subjected to the region separation process by the region separation unit to determine whether or not color misregistration has occurred, and it is determined that no color misregistration has occurred. Sometimes, control means for controlling the color misregistration correction means so as not to perform the color misregistration correction processing on the input image may be provided.
[0075]
In this case, the presence or absence of color misregistration is roughly determined in the low-resolution input image read during preliminary scanning, and color misregistration correction processing by the color misregistration correction means is performed only when it is determined that color misregistration has occurred. Therefore, it is possible to efficiently perform image processing without waste.
[0076]
The above-described image processing apparatus may be used for an image forming apparatus that performs a predetermined process on an input image to form an output image.
[0077]
In this case, according to the above-described image processing apparatus, it is possible to accurately determine whether or not color misregistration has occurred and perform color misregistration correction. Therefore, an image forming apparatus capable of outputting a high-quality image is provided. Can be provided.
[0078]
Examples of the image forming apparatus include a color image forming apparatus using an electrophotographic system or an ink jet system. Further, it may be an image display device capable of displaying an image after image processing, such as a liquid crystal display used as a display means of a computer or a portable information device.
[0079]
DETAILED DESCRIPTION OF THE INVENTION
An embodiment of the present invention will be described as follows. In this embodiment, a case where the image forming apparatus of the present invention is applied to a digital color copying machine will be described.
[0080]
As shown in FIG. 1, the digital color copying machine according to the present embodiment has a color image input device 1 for inputting color image information obtained from a document or the like, and the color image input device 1 inputs the color image information. Color image processing apparatus (image processing apparatus) 2 for performing various image processing on color image information, and color for outputting color image information subjected to various image processing by the color image processing apparatus 2 The image output apparatus 3 is comprised.
[0081]
As shown in FIG. 2, the color image input apparatus 1 includes a document supply unit 10 and a scanner unit 11 for reading image information from a document supplied from the document supply unit 10.
[0082]
The document supply unit 10 is provided on the scanner unit 11 and is supported to be openable and closable with respect to the document table 12 constituting the scanner unit 11. It consists of a double-sided automatic document feeder (RADF) mounted with a positional relationship.
[0083]
The document supply unit 10 first transports the document so that one side of the document faces the scanner unit 11 at a predetermined position on the document table 12, and after the image reading on the one side is completed, the other side However, the document is reversed and conveyed toward the document table 12 so as to face the scanner unit 11 at a predetermined position on the document table 12.
[0084]
Then, the document supply unit 10 discharges the document after the double-sided image reading is completed for one document, and performs a duplex conveyance operation for the next document.
[0085]
The above document transport and front / back reversing operations are controlled in relation to the operation of the entire digital color copying machine.
[0086]
The scanner unit 11 is provided with a document table 12, an operation panel (not shown), an optical system 13, and the like.
[0087]
An optical system 13 constituting the scanner unit 11 is provided below the document table 12 in order to read the document conveyed on the document table by the document supply unit 10 or the image of the document placed on the document table 12 by the user. Is arranged.
[0088]
The optical system 13 includes a first scanning unit 14 and a second scanning unit 15 as two document scanning bodies that reciprocate in parallel along the lower surface of the document table 12, an optical lens 16, and a photoelectric conversion element. A certain CCD (Charge Coupled Device) line sensor 17 is included.
[0089]
The first scanning unit 14 includes an exposure lamp 14 a that exposes the surface of the document image, and a first mirror 14 b that deflects a reflected light image from the document in a predetermined direction. On the other hand, it reciprocates in parallel at a predetermined scanning speed while maintaining a certain distance.
[0090]
The second scanning unit 15 includes a second mirror 15a and a third mirror 15b that deflect the reflected light image from the original deflected by the first mirror 14b of the first scanning unit 14 in a predetermined direction. , And reciprocates in parallel with the first scanning unit 14 while maintaining a constant speed relationship.
[0091]
The optical lens 16 reduces the reflected light image from the original deflected by the third mirror 15b of the second scanning unit 15, and forms the reduced light image at a predetermined position on the CCD line sensor 17. It is. The CCD line sensor 17 sequentially photoelectrically converts the formed light image and outputs it as an electrical signal.
[0092]
The CCD line sensor 17 is a three-line color CCD that can read a black and white image or a color image and output line data separated into R (red), G (green), and B (blue) color components. is there. The document image information converted into an electrical signal by the CCD line sensor 17 is transferred to the color image processing apparatus 2 as analog color image information and subjected to predetermined image data processing.
[0093]
The color image processing apparatus 2 is an image processing apparatus for carrying out the image processing method of the present invention. As shown in FIG. 1, an A / D (analog / digital) conversion unit 20, a shading correction unit 21, an input Tone correction unit 22, color misregistration correction unit (color misregistration correction means) 23, region separation processing unit (region separation means) 24, color correction unit 25, black generation and under color removal unit 26, spatial filter processing unit 27, output floor The tone correction unit 28 and the gradation reproduction processing unit 29 are included.
[0094]
That is, an analog signal read by the color image input device 1 is converted into an A / D conversion unit 20, a shading correction unit 21, an input tone correction unit 22, a color misregistration correction unit 23, and the like in the color image processing device 2. The CMYK digital color signal is sent in the order of the region separation processing unit 24, the color correction unit 25, the black generation and under color removal unit 26, the spatial filter processing unit 27, the output gradation correction unit 28, and the gradation reproduction processing unit 29. Is output to the color image output device 3.
[0095]
The A / D conversion unit 20 converts an analog RGB signal into a digital RGB signal, and the shading correction unit 21 applies the digital RGB signal sent from the A / D conversion unit 20 to the digital RGB signal. Various distortions generated in the illumination system (first scanning unit 14, second scanning unit 15, etc.), imaging system (optical lens 16, etc.), and imaging system (CCD line sensor 17, etc.) of the color image input apparatus 1 The removal process is performed.
[0096]
The input tone correction unit 22 adjusts the color balance of the RGB signal (RGB reflectance signal) from which various distortions have been removed by the shading correction unit 21, and at the same time, a color image processing device such as a density signal. 2 is subjected to processing for conversion into a signal that can be easily handled by the image processing system employed in the second embodiment.
[0097]
The color misregistration correction unit 23 detects the presence / absence of color misregistration generated by the color image input apparatus 1 with respect to the RGB signal sent from the input tone correction unit 22, and color misregistration occurs due to the detection result. If it is determined that the color shift has occurred, the color misregistration is corrected. The color misregistration correction unit 23 may input a black generation adjustment signal to the black generation and under color removal unit 26 based on the color misregistration determination result obtained by the internal processing. The detailed configuration of the color misregistration correction unit 23 will be described later.
[0098]
The region separation processing unit 24 separates each pixel in the input image into a character or line drawing region, a halftone dot region, or a photographic region from the RGB signal that has been color misregistration corrected by the color misregistration correction unit 23. . The region separation processing unit 24 outputs, to the black generation and under color removal unit 26, the spatial filter processing unit 27, and the gradation reproduction processing unit 29, a region identification signal indicating which region the pixel belongs to based on the separation result. In addition, the input signal output from the input tone correction unit 22 is output to the subsequent color correction unit 25 as it is.
[0099]
The color correction unit 25 performs processing for removing color turbidity based on the spectral characteristics of CMY (C: cyan, M: magenta, Y: yellow) color materials including unnecessary absorption components in order to realize faithful color reproduction. Is what you do.
[0100]
The black generation and under color removal unit 26 generates a black (K) signal from the CMY three-color signal after color correction, and subtracts the K signal obtained by the black generation from the original CMY signal to generate a new CMY signal. The CMY three-color signal is converted into a CMYK four-color signal.
[0101]
As an example of the black generation process, there is a method (general method) for generating black by skeleton black. In this method, the input / output characteristic of the skeleton curve is y = f (x), the input data is C, M, Y, the output data is C ′, M ′, Y ′, K ′, UCR (Under Color When the removal ratio is α (0 <α <1), the black generation and under color removal processing is expressed by the following equation (1).
[0102]
[Expression 1]

[0103]
The spatial filter processing unit 27 performs spatial filter processing using a digital filter on the image data of the CMYK signal input from the black generation and under color removal unit 26 to correct the spatial frequency characteristics. Thus, processing is performed so as to prevent blurring of the output image and deterioration of graininess.
[0104]
The gradation reproduction processing unit 29 performs predetermined processing on the image data of the CMYK signal based on the region identification signal, like the spatial filter processing unit 27.
[0105]
For example, the region separated into characters by the region separation processing unit 24 has a high frequency enhancement amount in the sharp enhancement processing in the spatial filter processing by the spatial filter processing unit 27 in order to improve the reproducibility of black characters or color characters in particular. Increased. At the same time, the gradation reproduction processing unit 29 selects binarization or multi-value processing on a high-resolution screen suitable for high frequency reproduction.
[0106]
In addition, with respect to the region separated into halftone dots by the region separation processing unit 24, the spatial filter processing unit 27 performs low-pass filter processing for removing the input halftone component. The output tone correction unit 28 performs output tone correction processing for converting a signal such as a density signal into a halftone dot area ratio that is a characteristic value of the color image forming apparatus. Gradation reproduction processing (halftone generation) is performed so that the image is finally separated into pixels and each gradation is reproduced. With respect to the region separated into photographs by the region separation processing unit 24, binarization or multi-value processing is performed on the screen with an emphasis on gradation reproducibility.
[0107]
The image data subjected to the above-described processes is temporarily stored in a storage means (not shown), read at a predetermined timing, and input to the color image output device 3.
[0108]
The color image output device 3 outputs image data onto a recording medium (for example, paper). For example, a color image output using an electrophotographic method or an ink jet method can be mentioned, but it is particularly limited. It may be an image display device that can display an image after image processing, such as a liquid crystal display used in a computer or a portable information device. In this case, the color correction unit 25 converts the input RGB signal into an R′G′B ′ signal according to the output characteristics of the image display device. Further, the black generation and under color removal unit 26 and the gradation reproduction processing unit 29 are unnecessary, and the output gradation correction unit 28 applies the color image output device 3 to the image data 1 output from the spatial filter processing unit 27. The gradation is adjusted according to the characteristics.
[0109]
Here, the color misregistration correction unit 23 in the color image processing apparatus 2 will be described below with reference to FIGS. 1 and 3.
[0110]
As shown in FIG. 3, the color misregistration correction unit 23 includes an image memory 30, a pixel density determination unit (pixel density determination unit) 31, a fitting processing unit (approximation unit) 32, a function storage unit 33, and a color misregistration determination unit ( A color misregistration determination unit) 34, a threshold storage unit 35, a correction processing unit (correction unit) 36, and a filter storage unit 37 are included, and are driven and controlled by an external control unit (control unit) 38.
[0111]
That is, the color misregistration correction unit 23 performs image memory 30, pixel density determination unit 31, fitting processing unit 32, color misregistration on the RGB signal (input image data) from the input tone correction unit 22 (FIG. 1). The determination unit 34 and the correction processing unit 36 perform color misregistration correction processing, and output RGB signals as color misregistration corrected image data to the subsequent region separation processing unit 24 (FIG. 1), while the black generation adjustment signal is output to the subsequent step. Are output to the black generation and under color removal unit 26 (FIG. 1).
[0112]
The processing of the color misregistration correction unit 23 will be specifically described as follows.
[0113]
Image data composed of a plurality of m × n pixels is stored in the image memory 30 with respect to the RGB signal (input image data) sent from the input tone correction unit 22 (m and n are positive integers). As will be described later, for example, when the pixel density is determined using the image data of the 1 × 7 mask area, at least 3 × 7 image data is stored.
[0114]
Next, for example, image data of a 1 × 7 mask area is extracted from the image memory 30, and the pixel density determination unit 31 determines the presence / absence of a local maximum point based on the density information. Here, when it is determined that a maximum point exists in the extracted image data, the RGB signal of the 1 × 7 mask area is output to the fitting processing unit 32.
[0115]
Note that the pixel density determination in the pixel density determination unit 31 is not necessarily performed (not necessary when the color misregistration correction unit 23 is arranged in the subsequent stage of the region separation processing unit 24 as described later). By performing the above, it is possible to quickly find an image area such as a character or a line drawing, and when it is determined that there is no maximum point, it is determined that the area is a ground area or an image area having a uniform density, Since it is not necessary to perform the fitting process, it is expected to speed up the subsequent processes.
[0116]
In the present embodiment, as described above, “whether or not there is a maximum point” is used as the determination criterion as the determination criterion of the pixel density in the pixel concentration determination unit 31. However, the present invention is not limited to this. Instead, a criterion such as “whether the pixel density tends to increase monotonously” or “whether there is a predetermined number of pixels having a predetermined density or higher” may be used.
[0117]
The fitting processing unit 32 performs a fitting process on each of the input RGB signals using the function stored in the function storage unit 33. At the time of fitting, calculation is performed so that the square error between the RGB value (a numerical value of the RGB signal) sent from the image memory 30 and the fitting result (function) is minimized.
[0118]
In the color misregistration determination unit 34, determination is made using the threshold value stored in the threshold value storage unit 35 for each parameter of the function obtained from the fitting result by the fitting processing unit 32, and color misregistration has occurred. If it is determined that color misregistration has occurred, the RGB signal sent from the image memory 30 is sent to the correction processing unit 36. On the other hand, if it is determined that no color misregistration has occurred, the RGB signal sent from the image memory 30 is sent to the area separation processing unit 24 at the subsequent stage of the color misregistration correction unit 23 as it is.
[0119]
The color misregistration determination unit 34 determines whether or not color misregistration has occurred in the 1 × 7 mask area of the input image data extracted from the image memory 30. That is, the color misregistration determination unit 34 determines whether or not the 1 × 7 mask area is an area where color misregistration has occurred (hereinafter referred to as a color misregistration area). The flow of color misregistration determination processing in the color misregistration determination unit 34 will be described later.
[0120]
The correction processing unit 36 corrects the color misregistration by performing a spatial filter process on the region determined to be the color misregistration region using, for example, a filter stored in the filter storage unit 37.
[0121]
The correction processing unit 36 outputs a black generation adjustment signal to the black generation / under color removal unit 26 when the black generation amount is controlled in the black generation / under color removal unit 26 (FIG. 1).
[0122]
Here, as a filter stored in the filter storage unit 37, for example, as shown in FIG. 4, with respect to the detected color misregistration area, a 3 × 3 mask centered on the target pixel e in the color misregistration area is used. Using the minimum value filter, the minimum value (e ′ = min (a, b, c, d, e, f, g, h, i)) in the mask is calculated by this minimum value filter, and the color misregistration correction result Output as. Here, before and after color misregistration correction are shown. As a result, the color misregistration area approaches the background density of the document, and as a result, only the character or line drawing area becomes clear.
[0123]
Although the present embodiment has been described using a 3 × 3 mask, the size of the mask is not limited to this. In the minimum value filter, the one having the lowest density is selected and used as the density of the processed pixel. Therefore, as the mask size increases, the density of the color shift area becomes closer to the background density of the document. However, as the mask size increases, the circuit scale increases, resulting in an increase in cost. Therefore, an appropriate size may be designed in consideration of this point.
[0124]
As another method of correcting the color misregistration area, the black generation and under color removal unit 26 satisfies F1> F2 where F1 is the black generation amount of the black character or line drawing area and F2 is the black generation amount of the color misregistration area. In this way, the black generation amount may be controlled. As a result, the black generation amount in the color misregistration area is set to be smaller than the black generation amount in the black character or line drawing area, so that the black character or line drawing is prevented from becoming thick and a good image with no color turbidity is formed. be able to.
[0125]
Various processes in the color misregistration correction unit 23 are performed via a control unit 38 including a CPU (Central Processing Unit) and a DSP (Digital Signal Processor).
[0126]
Further, the control unit 38 performs pre-scanning (preliminary scanning) of the original in the color image processing apparatus 40 shown in FIG. 9 to be described later, thereby increasing the reading speed of the scanner and processing the input image data read at a low resolution. When the area separation process is performed, the character or line drawing area is extracted, the absolute value of the difference between each color component of the RGB signal is compared with a predetermined threshold value, and as a result, it is determined that no color misregistration has occurred The color misregistration correction unit 23 is controlled not to perform processing during the main scan. In this case, pre-scanning is necessary, but correction processing only needs to be performed when it is determined that color misregistration has occurred, so that image processing can be performed efficiently.
[0127]
Next, the flow of color misregistration determination processing in the color misregistration correction unit 23 will be described below with reference to the flowchart shown in FIG.
[0128]
First, an image is input, and an RGB signal is obtained from the input tone correction unit 22 from the input image (step S1).
[0129]
Next, data of a predetermined size is stored in the image memory 30 from the obtained RGB signals, and 1 × 7 mask area data is extracted (step S2).
[0130]
Subsequently, the pixel density determination unit 31 determines the presence or absence of a local maximum point based on the density information for the extracted data (step S3). Here, as one of the methods for determining the maximum point, the difference information between adjacent data (pixel density) in the mask area is obtained and the point changes from a monotonic increase to a monotonic decrease or a state in which almost no difference is observed. There is a method of checking whether or not there is.
[0131]
If it is determined in step S3 that “a local maximum exists”, approximation is performed using a function in the fitting processing unit 32 (step S4).
[0132]
On the other hand, if it is not determined in step S3 that “a local maximum exists”, it is determined that the area is a background area or an image area having a uniform density, and the process proceeds to step S8 and there is no color shift (color). It is determined that there is no influence of the deviation or is not noticeable.
[0133]
In step S4, fitting is performed on the data in the 1 × 7 mask area using a function. When performing the fitting, there are various functions including a Gaussian function, and interpolation methods such as spline interpolation and Lagrange interpolation. In the present invention, the trigonometric function Acos is used.ⁿFitting by (θ + α) is performed. At this time, A represents the amplitude and represents the maximum point of the pixel density in the mask region. Further, n indicates the kurtosis of the function, and means that the line width of the target image region decreases as the value of n increases. θ represents the reference pixel position, and α represents the phase difference from the reference pixel position. If this parameter is known, the color misregistration amount can be obtained.
[0134]
After the data in the 1 × 7 mask area is approximated by using a function in step S4 described above, the square error with the estimated function is minimized for each of the RGB signals by fitting in step S5. A (A_R, A_G, A_B), Α (α_R, Α_G, Α_B), N (n_R, N_G, N_B)
[0135]
And α (α_R, Α_G, Α_B) To determine whether there is a color shift. That is, α_R= Α_G= Α_BIt is determined whether or not (step S6). Where α_R= Α_G= Α_BIf it is determined that there is no color shift, the process proceeds to step S8. Meanwhile, α_R= Α_G= Α_BIf not, the process proceeds to step S7 and correction processing is performed.
[0136]
Here, in the color misregistration determination process shown in the flowchart shown in FIG. 5, the following three conditions are conceivable as conditions for determining whether or not to perform color misregistration correction.
[0137]
<Condition 1>
There is a local maximum,_R= Α_G= Α_B"Is satisfied, it is determined that no color misregistration has occurred. That is, in step S3, it is determined that it has a maximum point, and in step S6, α_R= Α_G= Α_BWhen it is determined that the above condition is satisfied, it is determined that no color misregistration has occurred.
[0138]
<Condition 2>
There is a local maximum,_R= Α_G= Α_B"Is not satisfied, it is determined that color misregistration has occurred, and color misregistration correction processing is performed. That is, in step S3, it is determined that it has a maximum point, and in step S6, α_R= Α_G= Α_BWhen it is determined that the above condition is not satisfied, it is determined that color misregistration has occurred.
[0139]
<Condition 3>
When there is no maximum point, it is not a character or line drawing area, so it is determined that there is no influence of color misregistration or is not noticeable, and color misregistration correction processing is not performed. That is, when it is determined in step S3 that there is no maximum point, it is determined that there is no influence of color misregistration or is not noticeable.
[0140]
In the case of the above condition 2, that is, when it is determined that a color shift has occurred in the input image, a color shift correction process, that is, a correction process in the correction processing unit 36 shown in FIG. . This correction process is performed using the filter stored in the filter storage unit 37 described above.
[0141]
The color misregistration determination process is executed until all data for the input image is completed (step S9).
[0142]
Here, the fitting process in the fitting processing unit 32 in the color misregistration correction unit 23 will be described below.
[0143]
First, a fitting process for a color signal will be described below with reference to FIG. Here, the color signal indicates an RGB color signal.
[0144]
FIG. 6A is a graph showing the relationship between the pixel position and the pixel density in one predetermined color signal, and the fitting result by the function.
[0145]
Now, assuming that the target pixel position is t and the pixel density is D, the result of fitting with the function is D = 132 cos⁴⁰⁰⁰(T). In the present embodiment, the fitting is performed by setting one pixel to π / 180 [rad] (≈0.0175 [rad]). If the target pixel position t = 0 and it is determined that “the color shift occurs when the target pixel position is shifted by 0.1 pixel or more”, the color shift correction process shown in FIG. Therefore, a threshold value of α = 0.00175 may be provided. That is, when α ≧ 0.00175, it is determined that a color shift has occurred.
[0146]
Although one color signal is described here, the same fitting is performed for the other color signals, and the color misregistration determination shown in FIG. 5 is performed based on the obtained parameters. When color misregistration does not occur, the parameters α when the three color signals are fitted substantially match.
[0147]
Next, fitting processing when color misregistration occurs will be described below with reference to FIG.
[0148]
FIG. 6B is a graph showing the relationship between the pixel position and the pixel density in the three color signals when color misregistration occurs, and the fitting result by the function. This graph is obtained by reading a line chart with a scanner, but in a black line drawing or a character area, color misregistration can be confirmed from parameters.
[0149]
Next, a fitting process in a thick character / line drawing area will be described below with reference to FIGS. 7 (a) and 7 (b) and FIG.
[0150]
FIGS. 7A and 7B connect the relationship between the pixel position and the pixel density in a thick character or thick line drawing area (FIG. 7A) and the areas x and y at both ends except for the achromatic area, The result (FIG.7 (b)) which performed the fitting using a function is represented.
[0151]
As a result, the achromatic region can be regarded as a region that becomes almost zero when the difference information of the density levels of adjacent data in the mask region is taken. By this processing, it is possible to execute the same color misregistration determination as in the case of a small character or thin line even in a thick character or line drawing area.
[0152]
FIG. 8 shows a second method that makes it possible to perform the same color misregistration determination as in the case of a small character or thin line in a thick character or thick line drawing area. Even in this method, the achromatic region can be determined from the difference information of the density levels of adjacent data in the mask region. In this case, fitting is performed using data up to pixels where the pixel density is saturated. As a result, fitting can be performed without using the technique (FIGS. 7A and 7B) for connecting the image end regions.
[0153]
In the color image processing apparatus 2 shown in FIG. 1, the region separation is performed after the color misregistration correction is performed. However, the region separation may be performed first, and then the color misregistration correction may be performed. .
[0154]
That is, when color misregistration correction is performed using region separation information, for example, a color image processing apparatus 40 as shown in FIG. 9 is used. For convenience of explanation, members having the same functions as those of the structural members in the color image processing apparatus 2 shown in FIG.
[0155]
As shown in FIG. 9, the color image processing apparatus 40 is the same as the color image processing apparatus 2 shown in FIG. 1 except for the arrangement position of the color misregistration correction unit 23. In other words, the color image processing apparatus 40 has a structure in which the region separation processing unit 24 is arranged in front of the color misregistration correction unit 23.
[0156]
The region separation processing unit 24 separates each pixel in the input image from the RGB signal into one of a character or line drawing region, a halftone dot region, and a photographic region, and color misregistration is conspicuous and easy to determine. It is also possible to extract a black character or a black line drawing region and perform color misregistration correction by the color misregistration correction unit 23. As a result, only a predetermined region is extracted and color misregistration correction processing is performed, so that an improvement in processing speed is expected. The RGB signal subjected to the color misregistration correction is output to the color correction unit 25.
[0157]
Examples of the region separation method in the region separation processing unit 24 include: The method described in “Image Electronics Society of Japan Proceedings 90′06′04” can be used. Details will be described below. In this method, the following series of determinations are performed within a block of M × N (M and N are natural numbers) pixels centered on the target pixel, and these determination results are used as a region identification signal of the target pixel.
[0158]
1. Average signal level (D) for the central 9 pixels in the block_ave) And binarize each pixel in the block using the average value. The maximum pixel signal level (D_max), Minimum pixel signal level (D_min) At the same time.
[0159]
2. In the halftone dot region, the halftone dot region is identified by utilizing the fact that the fluctuation of the image signal in the small region is large and the density is higher than the background. With respect to the binarized data, the number of change points from 0 to 1 in the main scanning and sub-scanning directions is obtained, and the number of change points from 1 to 0 is obtained._H, K_VAnd threshold T_H, T_VIf both of them exceed the threshold, the halftone dot region is set. In order to prevent misjudgment with the background, D_max, D_min, D_aveTo threshold B₁, B₂Compare with That is, the halftone dot region and the non-halftone dot region are separated under the following conditions.
[0160]
[Expression 2]

[0161]
3. In the character area, since the difference between the maximum signal level and the minimum signal level is large and the density is considered high, the character area is identified as follows. The maximum and minimum signal levels previously obtained in the non-halftone area and the difference between them (D_sub) For threshold P_A, P_B, P_CIf any one of them exceeds the character region, the character region is used. That is, the non-halftone area is divided into a character area and a photograph area under the following conditions.
[0162]
[Equation 3]

[0163]
As described above, according to the digital color copying machine having the above configuration, the color shift generated in the image data input by the color image input apparatus 1 is determined for each input image data, and based on the determination result, the color shift is determined. Since correction is performed, it is less time-consuming and performs color misregistration with higher accuracy than in the case of conventional color misregistration correction using preset reference data (color misregistration amount). Can do.
[0164]
Further, in the present embodiment, the same effect can be obtained even by the following method other than the image processing method as described above.
[0165]
That is, in the image processing method for correcting the color misregistration when it is determined that a plurality of color components of the input image data have color misregistration at an arbitrary pixel position, the image processing method includes: The first step of approximating the density distribution for each pixel position for each of a plurality of color components of data using a function and the approximation formula obtained for each of the plurality of color components are compared, and whether or not color misregistration has occurred A second step of determining whether or not color misregistration has occurred, and a third step of performing correction processing of the color misregistration when it is determined that color misregistration has occurred.
[0166]
As described above, by performing fitting using a function, it is possible to accurately determine the presence or absence of color misregistration without reading reference data in advance.
[0167]
In addition, since color misregistration occurs when the pixel position of each color component in the input image deviates, it is determined with high accuracy by determining whether or not color misregistration occurs based on the pixel position of each color component that is primary information. Color misregistration determination and color misregistration correction can be performed.
[0168]
For each input image, it is determined whether or not color misregistration has occurred based on the pixel position of each color component. When it is determined that color misregistration has occurred, color misregistration correction is performed and image processing is performed. Because.
[0169]
Another embodiment of the present invention will be described as follows.
[0170]
A color image processing apparatus 50 according to the present embodiment is shown in FIG. For convenience of explanation, members having the same functions as those of the structural members in the color image processing apparatus 2 shown in FIG.
[0171]
As shown in FIG. 10, the color image processing apparatus 50 is the same as the color image processing apparatus 2 shown in FIG. That is, the color image processing apparatus 50 has a structure including a color misregistration correction unit 51 instead of the color misregistration correction unit 23.
[0172]
As shown in FIG. 11, the color misregistration correction unit 51 includes an image memory 30, a pixel density determination unit (pixel density determination unit) 31, a pixel position estimation unit (pixel position estimation unit) 52, and a color misregistration determination unit (color misregistration). A determination unit 53, a correction processing unit (correction unit) 36, and a filter storage unit 37 are included, and are driven and controlled by an external control unit 38. For convenience of explanation, the image memory 30, the pixel density determination unit 31, the correction processing unit 36, the filter storage unit 37, and the control unit are members having the same functions as the constituent members in the color misregistration correction unit 23 shown in FIG. The same reference numerals are attached to 38 and the description thereof is omitted.
[0173]
The color misregistration correction unit 51 performs image memory 30, pixel density determination unit 31, pixel position estimation unit 52, color misregistration determination on the RGB signal (input image data) from the input tone correction unit 22 (FIG. 10). The color misregistration correction processing is performed by the unit 53 and the correction processing unit 36, and the RGB signal as the color misregistration corrected image data is output to the subsequent area separation processing unit 24 (FIG. 10), while the black generation adjustment signal is output to the subsequent stage. This is output to the black generation and under color removal unit 26 (FIG. 10).
[0174]
The processing of the color misregistration correction unit 51 will be specifically described as follows.
[0175]
Image data composed of a plurality of m × n pixels is stored in the image memory 30 with respect to the RGB signal (input image data) sent from the input tone correction unit 22 (m and n are positive integers). As will be described later, for example, when the pixel density is determined using the image data of the 1 × 7 mask area, at least 3 × 7 image data is stored.
[0176]
Next, for example, image data of a 1 × 7 mask area is extracted from the image memory 30, and the pixel density determination unit 31 determines the presence / absence of a local maximum point based on the density information. Here, when it is determined that a maximum point exists in the extracted image data, the RGB signal of the 1 × 7 mask region is output to the pixel position estimation unit 52.
[0177]
Note that the pixel density determination in the pixel density determination unit 31 does not necessarily have to be performed (as described above, it is not necessary when the color misregistration correction unit 51 is arranged at the subsequent stage of the region separation processing unit 24). By performing the above, it is possible to quickly find an image area such as a character or a line drawing, and when it is determined that there is no maximum point, it is determined that the area is a ground area or an image area having a uniform density, Since it is not necessary to perform color misregistration determination, the subsequent processing is expected to be speeded up.
[0178]
The pixel position estimation unit 52 calculates the average density of all color components for the input RGB signal (1 × 7 mask area data), and the pixel position β of each RGB signal at that time._R, Β_G, Β_BIs estimated. The average density is calculated for each RGB signal, and β is calculated from the average density of each signal._R, Β_G, Β_BMay be estimated.
[0179]
Pixel position β of average density of each signal_R, Β_G, Β_BIs estimated by linearly approximating the pixel position at the average density of the RGB signals sent from the pixel density determination unit 31 using the pixel density of the pixels adjacent to this pixel position. Note that the pixel position β at the average density of the RGB signals_R, Β_G, Β_BDetails of the method for estimating the value will be described later.
[0180]
Further, when the reflected light image from the original passes through the optical lens, the balance of the input RGB signals may be greatly lost due to the difference in wavelength. In this case, after performing density balance correction on the signal of the color component having a low density, the average density of all the color components is calculated. Details of this density balance correction will be described later.
[0181]
In the color misregistration determination unit 53, the pixel position β of each of the RGB signals obtained by the pixel position estimation unit 52._R, Β_G, Β_BAre substantially matched. If it is determined that the region is a color misregistration region, it is sent to the correction processing unit 36. On the other hand, if it is determined that no color misregistration has occurred, the RGB signal sent from the image memory 30 is sent to the area separation processing unit 24 at the subsequent stage of the color misregistration correction unit 51 as it is. The flow of the color misregistration determination process will be described later with reference to FIG.
[0182]
In the correction processing unit 36, for example, a spatial filter process is performed on the area determined as the color misregistration area using a filter stored in the filter storage unit 37 to correct the color misregistration. Alternatively, when the black generation amount is controlled by the black generation and under color removal unit 26 as shown in FIG. 10, the correction processing unit 36 outputs a black generation adjustment signal to the black generation and under color removal unit 26.
[0183]
Next, the flow of color misregistration determination processing in the color misregistration correction unit 51 will be described below with reference to the flowchart shown in FIG.
[0184]
First, an image is input, and an RGB signal is obtained from the input tone correction unit 22 from the input image (step S11).
[0185]
Next, data of a predetermined size is stored in the image memory 30 from the obtained RGB signals, and 1 × 7 mask area data is extracted (step S12).
[0186]
Subsequently, the pixel density determination unit 31 determines the presence / absence of a local maximum point based on the density information for the extracted data (step S13). Here, as one of the methods for determining the maximum point, the difference information between adjacent data (pixel density) in the mask area is obtained and the point changes from a monotonic increase to a monotonic decrease or a state in which almost no difference is observed. There is a method of checking whether or not there is.
[0187]
If it is determined in step S13 that “a local maximum exists”, an average density is calculated for the data in the 1 × 7 mask area, and the pixel position β at that time is calculated._R, Β_G, Β_BIs estimated (step S14).
[0188]
Average density pixel position β_R, Β_G, Β_BIs estimated by linear interpolation from the pixel position of the pixel adjacent to the position corresponding to the average density and the pixel density. And the pixel position β at the average density of the RGB signal_R, Β_G, Β_BMore color misregistration amount can be obtained.
[0189]
Here, when the balance of the RGB signals is largely lost with respect to the data of the 1 × 7 mask area, density balance correction processing is first performed (density balance correction means), and an average density is calculated. In this case, a density balance correction processing routine as shown in FIG. 13 is inserted between step S13 and step S14 in the flowchart shown in FIG.
[0190]
In FIG. 13, for example, the highest density value of each RGB signal in the 1 × 7 mask area is set as max R, max G, max B, and the lowest density value is set as min R, min G, min B. The threshold value of the maximum density difference is set as T, and it is checked whether or not density balance correction processing is necessary by the following equation (step S19).
[0191]

If the above conditional expression is satisfied, the process proceeds to step S20 and the density balance correction process is performed. If not satisfied, the process proceeds to step S14 and the pixel position estimation process is performed without performing the density balance correction process.
[0192]
In the density balance correction process, the maximum magnification of the color component signal having the highest density in the 1 × 7 mask region is compared with the maximum value of the color component signal satisfying the above conditional expression to obtain a correction magnification, and is shown in FIG. Thus, the correction magnification is multiplied by each data of the color component signal satisfying the conditional expression.
[0193]
FIG. 15 is a diagram illustrating a relationship between a pixel position and a pixel density in an RGB signal, and a method for estimating a pixel position of an average density by linear approximation. The pixel density is discrete data for each pixel, but for the sake of easy understanding, data between pixels is shown as a curve by interpolation.
[0194]
For example, when a monochrome straight line is read, the pixel density is close to the average density Ave, the pixel positions of adjacent G signals are P and P + 1, and the pixel density of the G signal at each position is G._P, G_{P + 1}And At this time, the average density pixel position β in the G signal_GCan be approximately expressed by the following equation.
[0195]
[Expression 4]

[0196]
Here, only one color signal is described, but the other color signals R and B signals are similarly expressed by β_R, Β_BIs calculated.
[0197]
When no color shift occurs, the average density pixel position β of the three RGB signals_R, Β_G, Β_BAre almost identical. So, for example,
max (| β_R-Β_G|, | Β_G-Β_B|, | Β_B-Β_R|) <0.5
And whether or not there is a color shift is determined based on whether or not the above condition is satisfied. And when this condition is met,
β_R= Β_G= Β_B
It is determined that no color misregistration has occurred. Therefore, in step S15 of the flowchart of FIG._R= Β_G= Β_BIt is determined whether or not.
[0198]
Here, in the color misregistration determination process shown in the flowchart shown in FIG. 12, the following three conditions are considered as conditions for whether or not to perform color misregistration correction.
[0199]
<Condition 1>
There is a maximal point, “β_R= Β_G= Β_B"Is satisfied, it is determined that no color misregistration has occurred. That is, it is determined in step S13 that it has a maximum point, and in step S15, β_R= Β_G= Β_BWhen it is determined that the above condition is satisfied, it is determined that no color misregistration has occurred.
[0200]
<Condition 2>
There is a maximal point, “β_R= Β_G= Β_B"Is not satisfied, it is determined that color misregistration has occurred, and color misregistration correction processing is performed (step S16). That is, it is determined in step S13 that it has a maximum point, and in step S15, β_R= Β_G= Β_BWhen it is determined that the above condition is not satisfied, it is determined that color misregistration has occurred.
[0201]
<Condition 3>
When there is no maximum point, it is not a character or line drawing area, so it is determined that there is no influence of color misregistration or is not noticeable, and color misregistration correction processing is not performed. That is, when it is determined in step S13 that there is no maximum point, it is determined that there is no influence of color misregistration or is not noticeable.
[0202]
In the case of the above condition 2, that is, when it is determined that a color shift has occurred in the input image, a color shift correction process, that is, a correction process in the correction processing unit 36 shown in FIG. 10 is executed in step S16. . This correction process is performed using the filter stored in the filter storage unit 37 described above.
[0203]
The color misregistration determination process is executed until all data for the input image is completed (step S18).
[0204]
In the color image processing apparatus 50 shown in FIG. 10, the region separation is performed after the color misregistration correction. However, the region separation is performed first, and then the color misregistration correction is performed. Also good.
[0205]
【The invention's effect】
As described above, according to the image processing method of the present invention, in the image processing method for correcting the color shift of the input image to obtain an output image, the color shift occurs for each input image based on the pixel position of each color component. The color misregistration correction is performed when it is determined whether or not the color misregistration has occurred.
[0206]
Therefore, it is possible to perform color misregistration correction for each input image by determining whether color misregistration has occurred for each input image and performing color misregistration correction when it is determined that color misregistration has occurred. .
[0207]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0208]
Further, unlike the prior art, it is not necessary to use a line chart or a reference pattern which is a reference document, so that it is possible to save the trouble of using these.
[0209]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0210]
In addition, since color misregistration occurs when the pixel position of each color component in the input image deviates, it is determined with high accuracy by determining whether or not color misregistration occurs based on the pixel position of each color component that is primary information. There is an effect that color misregistration determination and color misregistration correction can be performed.
[0211]
As described above, the image processing method of the present invention also detects the amount of displacement of each pixel position of each color component for each input image in the image processing method for obtaining an output image by correcting the color shift of the input image. Then, based on this detection result, it is determined whether or not color misregistration has occurred, and when it is determined that color misregistration has occurred, color misregistration correction is performed.
[0212]
Therefore, it is possible to perform color misregistration correction for each input image by determining whether color misregistration has occurred for each input image and performing color misregistration correction when it is determined that color misregistration has occurred. .
[0213]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0214]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0215]
In addition, since it is determined whether or not a color shift has occurred by detecting a positional shift amount of each pixel position of each color component of the input image, it is possible to easily determine the color shift. it can.
[0216]
In addition, since it is not necessary to use a line chart or a reference pattern as a reference document as in the past, high-accuracy color misregistration does not take into account errors caused by using linear interpolation as in the case of using these. There is an effect that correction can be performed.
[0217]
Furthermore, for each color component of the input image, the density distribution with respect to each pixel position may be approximated by a function, and based on this approximation result, it may be determined whether color misregistration has occurred.
[0218]
In this case, by approximating the density distribution for each pixel position with a function for each color component of the input image, the shift amount of the pixel position of each color component of the input image can be accurately obtained.
[0219]
Therefore, if the color shift of the input image is determined based on the shift amount of the pixel position of each color component of the input image, the presence or absence of the color shift of the input image can be determined with higher accuracy.
[0220]
As a result, the accuracy of color misregistration correction can be further improved. As a result, an output image that faithfully reproduces the input image can be obtained.
[0221]
The image processing method of the present invention also estimates the pixel position at the average density of each color component for each input image in the image processing method for obtaining an output image by correcting the color shift of the input image as described above. Based on this estimation result, it is determined whether or not color misregistration has occurred, and color misregistration correction is performed when it is determined that color misregistration has occurred.
[0222]
Therefore, it is possible to perform color misregistration correction for each input image by determining whether color misregistration has occurred for each input image and performing color misregistration correction when it is determined that color misregistration has occurred. .
[0223]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0224]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0225]
In addition, it is possible to easily determine the color misregistration because it is determined whether or not the color misregistration occurs by estimating the pixel position at the average density of each color component of the input image. There is an effect.
[0226]
Further, when it is determined that the density balance of each color component of the input image exceeds a predetermined range, the density balance of each color component may be corrected so as to be within a predetermined range. .
[0227]
In this case, furthermore, before estimating the pixel position at the average density of each color component of the input image, when it is determined that the density balance of each color component of the input image exceeds a predetermined range, correction is performed. The pixel position can be estimated with high accuracy.
[0228]
As a result, the accuracy of color misregistration correction can be further improved. As a result, an output image that faithfully reproduces the input image can be obtained.
[0229]
As described above, the image processing apparatus of the present invention is an image processing apparatus including a color misregistration correction unit that corrects a color misregistration generated in an input image. The color misregistration correction unit includes pixels for each color component of the input image. Color misregistration determining means for determining whether color misregistration has occurred based on the position, and correction means for correcting the color misregistration when the color misregistration determining means determines that color misregistration has occurred. It is the composition which is provided.
[0230]
Therefore, it is possible to perform color misregistration correction for each input image by determining whether color misregistration has occurred for each input image and performing color misregistration correction when it is determined that color misregistration has occurred. .
[0231]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0232]
Further, unlike the prior art, it is not necessary to use a line chart or a reference pattern which is a reference document, so that it is possible to save the trouble of using these.
[0233]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0234]
In addition, since color misregistration occurs when the pixel position of each color component in the input image deviates, it is determined with high accuracy by determining whether or not color misregistration occurs based on the pixel position of each color component that is primary information. There is an effect that color misregistration determination and color misregistration correction can be performed.
[0235]
As described above, the image processing apparatus of the present invention is also provided with a color misregistration correction unit that corrects a color misregistration generated in the input image. The color misregistration correction unit includes each color component of the input image. Further, an approximation means for approximating the density distribution with respect to each pixel position by a function, a color misregistration determination means for judging whether color misregistration has occurred based on an approximation result by the approximating means, and the color misregistration determination means. When it is determined that color misregistration has occurred, the image forming apparatus includes a correcting unit that corrects the color misregistration.
[0236]
Therefore, it is possible to perform color misregistration correction for each input image by determining whether color misregistration has occurred for each input image and performing color misregistration correction when it is determined that color misregistration has occurred. .
[0237]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0238]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0239]
In addition, by approximating the density distribution for each pixel position with a function for each color component of the input image, the shift amount of the pixel position of each color component of the input image can be accurately obtained.
[0240]
Therefore, if the color shift of the input image is determined based on the shift amount of the pixel position of each color component of the input image, the presence or absence of the color shift of the input image can be determined with higher accuracy.
[0241]
In addition, since it is not necessary to use a line chart or a reference pattern as a reference document as in the past, high-accuracy color misregistration does not take into account errors caused by using linear interpolation as in the case of using these. There is an effect that correction can be performed.
[0242]
As described above, the image processing apparatus of the present invention is also provided with a color misregistration correction unit that corrects a color misregistration generated in the input image. The color misregistration correction unit includes each color component of the input image. A pixel position estimating unit that estimates a pixel position of an average density from the density distribution of the color, a color misregistration determining unit that determines whether color misregistration has occurred based on an estimation result of the pixel position estimating unit, and the color misregistration determination And a correction unit that corrects the color misregistration when it is determined that the color misregistration has occurred.
[0243]
Therefore, it is possible to perform color misregistration correction for each input image by determining whether color misregistration has occurred for each input image and performing color misregistration correction when it is determined that color misregistration has occurred. .
[0244]
In this way, color misregistration is determined for each input image, and color misregistration correction is performed based on the obtained color misregistration determination result. The color misregistration correction can always be performed with high accuracy.
[0245]
Further, as described above, if the presence or absence of color misregistration is determined for each input image, if there is no color misregistration, image processing can be performed without performing color misregistration correction processing. As described above, the time required for image processing can be significantly shortened as compared to the case where color misregistration processing is always performed on an input image regardless of the presence or absence of color misregistration.
[0246]
In addition, the pixel position of the average density can be estimated from the density distribution for each color component of the input image.
[0247]
Therefore, if the color shift of the input image is determined based on the pixel position of the average density for each color component of the input image, it is possible to more accurately determine the presence or absence of the color shift of the input image. .
[0248]
The pixel position estimating means further corrects the density balance of each color component to fall within a predetermined range when it is determined that the density balance of each color component of the input image exceeds a predetermined range. A configuration may be provided in which density balance correction means is provided.
[0249]
In this case, furthermore, before estimating the pixel position at the average density of each color component of the input image, when it is determined that the density balance of each color component of the input image exceeds a predetermined range, correction is performed. The pixel position can be estimated with high accuracy.
[0250]
As a result, the accuracy of color misregistration correction can be further improved. As a result, an output image that faithfully reproduces the input image can be obtained.
[0251]
The color misregistration correction means further obtains a density distribution in a predetermined area based on the density for each pixel, and determines whether or not the area is a target of color misregistration determination based on the density distribution. The pixel density determination unit may be provided in the preceding stage of the color misregistration determination unit.
[0252]
In this case, since the pixel density determination unit is provided before the color misregistration determination unit, it is only necessary to perform the color misregistration determination in the designated region, and the processing speed of the color misregistration correction processing is improved. There is an effect that can be.
[0253]
Further, the image processing apparatus may include a region separating unit that separates the input image into each region of characters, halftone dots, and photographs, and the color misregistration correcting unit may be provided before the region separating unit.
[0254]
In this case, by performing color misregistration correction processing by the color misregistration correction unit in advance before the region separation of the input image by the region separation unit, erroneous determination of region separation in the region separation processing by the region separation unit is prevented, There is an effect that the accuracy of the region separation in the separation processing can be improved.
[0255]
Further, the color misregistration correction unit may be provided after the region separation unit.
[0256]
In this case, color misalignment that occurs around the character or line drawing area is detected by performing color misregistration correction processing by the color misregistration correcting means on the area separated by the area separating means, that is, the extracted character or line drawing area.・ It can be corrected.
[0257]
In particular, since the character and line drawing areas where color misregistration is conspicuous are extracted by the area separation process and the color misregistration correction process is performed, there is an effect that the processing speed can be increased.
[0258]
Also, the input image read during the preliminary scan of the document is subjected to the region separation process by the region separation unit to determine whether or not color misregistration has occurred, and it is determined that no color misregistration has occurred. In some cases, a control unit that controls the color misregistration correction unit so as not to perform the color misregistration correction process on the input image may be provided.
[0259]
In this case, it is determined whether or not there is a color shift in the low-resolution input image read at the time of preliminary scanning. As a result, it is possible to efficiently perform image processing without waste.
[0260]
As described above, the image forming apparatus of the present invention forms an output image by performing predetermined processing on an input image, and includes the above-described image processing apparatus of the present invention.
[0261]
Therefore, according to the above-described image processing apparatus, it is possible to accurately determine whether or not color misregistration has occurred and perform color misregistration correction. Therefore, an image forming apparatus that can output a high-quality image is provided. There is an effect that it can be provided.
[0262]
Examples of the image forming apparatus include a color image forming apparatus using an electrophotographic system or an ink jet system. Further, it may be an image display device capable of displaying an image after image processing, such as a liquid crystal display used as a display means of a computer or a portable information device.
[Brief description of the drawings]
FIG. 1 is a schematic block diagram of a digital color copying machine as an image forming apparatus provided with an image processing apparatus of the present invention.
2 is a schematic configuration diagram showing a color image input device provided in the digital color copying machine shown in FIG. 1;
3 is a schematic block diagram showing a color misregistration correction unit of the color image processing apparatus provided in the digital color copying machine shown in FIG. 1;
4 is an explanatory diagram when a filter process is performed as the color misregistration correction process performed by the color misregistration correction unit illustrated in FIG. 3;
5 is a flowchart showing the flow of color misregistration correction processing performed in the color misregistration correction unit shown in FIG. 3;
6A is a graph showing an example of fitting using a function, and FIG. 6B is a graph showing a fitting result by a function when color misregistration occurs.
FIGS. 7A and 7B are graphs for explaining a method of fitting a thick character or line drawing area using a function to determine whether or not a color shift has occurred.
FIG. 8 is a graph for explaining another method for fitting a thick character or line drawing area using a function to determine whether or not a color shift has occurred.
FIG. 9 is a schematic block diagram of a digital color copying machine including another color image processing apparatus of the present invention.
FIG. 10 is a block diagram of a schematic configuration of a digital color copying machine including still another color image processing apparatus according to the present invention.
11 is a schematic block diagram showing a color misregistration correction unit of the color image processing apparatus provided in the digital color copying machine shown in FIG.
12 is a flowchart showing the flow of color misregistration correction processing performed in the color misregistration correction unit shown in FIG.
FIG. 13 is a flowchart illustrating a flow of density balance correction processing.
FIG. 14 is a graph illustrating a method for performing density balance correction processing;
FIG. 15 is a graph illustrating a method for estimating a pixel position of average density by linear interpolation and obtaining a color shift amount.
[Explanation of symbols]
1 Color image input device
2 Color image processing device (image processing device)
3 Color image output device
23 Color misregistration correction unit (color misregistration correction means)
24 Region separation processing unit (region separation means)
31 Pixel density determination unit (pixel density determination means)
32 Fitting processing unit (approximation means)
34 Color shift determination unit (color shift determination means)
36 Correction processing unit (correction means)
38 Control unit (control means)
40 Color image processing device (image processing device)
50 color image processing device (image processing device)
51 Color misregistration correction unit (color misregistration correction means)
52 Image position estimation unit (image position estimation means)
53 Color shift determination unit (color shift determination means)

Claims (6)

In an image processing method for obtaining an output image by correcting a color shift of an input image,
A first step for determining, for each input image, the presence or absence of a local maximum point based on density information for the data extracted from the input image ;
A second step of calculating an average density for the extracted data and estimating a pixel position of the calculated average density for each RGB when it is determined in the first step that there is a maximum point;
It said each pixel position beta _R of RGB that have been estimated in the second step, beta _G, is beta _B, a third step of determining whether to satisfy the _{β R = β G = β B} ,
When it is determined in the third step that β _R = β _G = β _B is not satisfied , the extracted data is subjected to spatial filter processing using a prestored filter to correct color misregistration . An image processing method comprising: a fourth step . In an image processing apparatus provided with color misregistration correction means for correcting color misregistration of an input image,
The color misregistration correction means is
For each input image , a pixel density determination unit that determines the presence or absence of a local maximum point based on density information for data extracted from the input image ;
A pixel position estimation unit that calculates an average density for the extracted data and estimates a pixel position of the calculated average density for each RGB when the pixel density determination unit determines that there is a maximum point; ,
A color misregistration determining unit that determines whether or not each of the RGB pixel positions β _R , β _G , β _B estimated by the pixel position estimating unit satisfies β _R = β _G = β _B ;
If the color misregistration determining unit determines that β _R = β _G = β _B is not satisfied , the extracted data is subjected to spatial filter processing using a pre-stored filter, and color misregistration is performed. An image processing apparatus comprising: a correction processing unit that performs correction. 3. The apparatus according to claim 2 , further comprising region separation means for separating the input image into character, halftone dot, and photograph regions, wherein the color misregistration correction means is provided in a preceding stage of the region separation means. Image processing device. 4. The image processing apparatus according to claim 2, further comprising an area separation unit that separates an input image into areas of characters, halftone dots, and photographs, wherein the color misregistration correction unit is provided after the area separation unit. The image processing apparatus according to any one of the above. When the input image read during the preliminary scanning of the document is subjected to the region separation processing by the region separation unit, it is determined whether or not color misregistration has occurred, and when it is determined that no color misregistration has occurred 5. The image processing according to claim 4 , further comprising control means for controlling the color misregistration correction means so as not to perform color misregistration correction processing on the input image read during the main scan of the original. apparatus. In an image forming apparatus that performs predetermined processing on an input image to form an output image,
An image forming apparatus comprising the image processing apparatus according to any one of claims 2 to 5.

Images