JP4132585B2 - Image processing apparatus and color image processing apparatus - Google Patents

Description

【００７９】
なお、ここで、ｉは主走査方向の畳込み変数、ｊは副走査方向の畳込み変数、ｋi，ｋj、ｎは任意の定数である。
【００８０】
そして、ｋi，ｋjは実際に印字されるドット径（最小ドット径）及びピッチ間隔によって最適な値が決まり、指数部ｎはドット形状、特にドットのエッジ形状によって最適な値が決まるようになっている。本式は用紙上に印字されるインクドットの光学的特性をパターン化した近似計算モデルである。
【００８１】
この時同じ値をとる位置が複数ある場合が想定される。これは初期パターンが周期的な組織的ディザであるがため起り易いが、この時どの位置を選択するかをランダムに選択しても最初に得られた位置としても好適な結果が得られる。
【００８２】
続いてステップＳ３にて、この検出された位置の画素に対してその順位を保存し、さらに、その位置のビットを０から１に変更したパターンを生成する。なお、この場合は対象が最小ディザ周期単位ではなく、全マトリクスサイズでの一連の順位が決定される。
【００８３】
これを０のビットパターンが無くなるまで繰り返し行い、全３０×３０画素の優先順位を決定する。なお、優先順位の割り当ては、既に周期的なパターンで再現する部分（２５×４＝１００個）を予め最初の順位割り振っておくと、計算行程では１０１〜９００までの優先順位が得られ、これを５〜３６の閾値に割り振ることにより行う。
【００８４】
この優先順位に従って、閾値を割り当てていく。サイズ３０×３０のマトリクスの場合は、２５個の画素づつ閾値が１つずつ増加していくような割り当てとなる。これにより残りの５〜３６の閾値を持つ画素の位置が決定され、最終的に３０×３０のマトリクス内全ての閾値が埋まる。これが３０×３０サイズのディザ基準閾値配列となる。
【００８５】
この時、さらに相対的に印字精度の低い方向にドットが連なるように閾値を生成するように、フィルタ演算の重みを主走査方向と副走査方向で相対的に変える。つまり上記数１式のｋｉ、ｋｊの値に重みを持たせる。詳しくは、ｋｉ＜ｋｊとすることにより、印字精度が低い主走査方向に連結し易いパターンを生成することができる。なお、この連結の強度は、ｋｉ、ｋｊの比率を変えることにより、図１６に示すように印字精度に応じて最適に設定することが望ましい。
【００８６】
これにより多値のプリンタの場合、非等方に生成される基準閾値配列は、単独では印字精度に対して大きな補正効果を持つことはないが、各閾値プレーン間のシーケンスと組み合わせることで、濃度ムラやスジといった印字誤差を大きく緩和する作用を持てるようになる。なお、２値のプリンタで２値ディザ処理を行う場合は、最初から解像度ピッチに対して大きなドットが隣接画素間で連結するため濃度ムラやスジに対して強くなる。従って、この基準閾値配列のみで効果を有する。
【００８７】
なお、この周期的な組織的ディザによる閾値の配置と、局所的に非周期的なディザ閾値の配置との切り替わりは、基本ディザ単位内の画素数のおよそ１／１０程度が良いことが分かった。これはあまり多くの画素を組織的ディザの閾値で固定してしまうと、空いている領域の自由度が著しく低くなってしまうために、特定階調で逆に不自然なテクスチャ等が発生してしまう可能性があるからである。実験的には、周期的な組織的ディザで構成する範囲は、基準閾値範囲の０〜２０％程度までが良い結果を得ている。
【００８８】
ここで上限２０％の場合とは、図１７に示すパターン（ＯＮ３２０／全画素１６００＝２０％）で構成されたようなものであり、周期性を持たせる限界のパターンである。残りのドット配置に自由度を持たせるのはこの２０％が限界である。もしこれ以上の部分を周期的に指定してしまうととたんに著しくテクスチャが発生することになる。
【００８９】
また、マトリクスサイズであるが、このサイズが余りにも小さいと周期性あるいは不要なテクスチャが見えてしまうので、冗長的ではない適当なサイズが必要である。この最適サイズは、各ドットの基本特性及び用紙との相性により変化する。各基本階調画素のドット設計上、極端に大きな非線型性を示すことが無いようであれば、１画素Ｍ階調の入力階調画像データを、多値ディザ処理して１画素Ｎ（Ｍ＞Ｎ＞２）階調のより小さい階調数の画像データへ変換する場合には、多値ディザ処理のマトリクスサイズをＫ×Ｌ、多値化した後の出力階調数をＮ階調とおくと、
【００９０】
【数２】

【００９１】
の範囲の整数となるようにＫとＬを設定し、また特に正方マトリクス（Ｌ×Ｌ）の場合には、
【００９２】
【数３】

【００９３】
の範囲の整数となるように設定するようにすれば周期性及びテクスチャの発生を押さえることができる。この最小限界値の方は視覚的な我慢限界を示し、最大限界値の方は冗長的なサイズにならない限界を示す。なお、これによって導き出されるマトリクスサイズは、２値のストカスティックディザ一般にいわれている１２８×１２８以上の好適サイズに対して、もはや大規模マトリクスサイズとは呼べない小さなサイズとなり、より小さなハードウェア構成で実現することができる。
【００９４】
さらに、組織的ディザの基本周期で割り切れる整数値をとることを考えると、上記２つのマトリクスサイズの条件を満たす、例えば、図１４の右側に示すように３０×２４等の非正方のマトリクスでも良い。
【００９５】
ここでは最小ディザ単位を正方マトリクスとした例について述べたが、図１８に示すように最小ディザ単位を長方マトリクスにしても良い。図１８においては、８ｂｉｔ、２５６階調（０：白、２５５：黒）の入力階調画像データを擬似中間調処理して各色２ｂｉｔ、４階調（０：白、３：黒）に変換する場合を例として説明する。
【００９６】
この場合、多値ディザ処理で２５６階調を超えない最大再現階調数を実現することが可能な異なる閾値の最大個数ｘは、
２５６／｛ｘ＊（４−１）＋１｝≧１ ゆえにｘ≦８５
である。これは言い換えると３プレーンある閾値配列のうち各閾値プレーンが担当する階調数は８５階調分と言うことである。なお、必ずしも８５階調にする必要はなく、説明をわかりやすくするために、今回は８０階調分の出力パターンを各閾値プレーンが担当している。因みに、この時全閾値プレーンでは８０＊３＋１＝２４１階調の階調が再現できる。
【００９７】
この８０階調分を最小単位のマトリクスで構成した場合１０×８となり、この１０×８の閾値マトリクスの任意画素を１つずつオンすることによって８０階調分の階調を再現できる。ここで、図１８に示すように１０×８の閾値配列Ｂ1を最小ディザ単位とすると、４０×４０の全閾値配列Ｂ２は、Ｂ1の最小閾値配列が主走査方向に４つ、副走査方向に５つの計２０個がちょうど収まるサイズである。
【００９８】
このようにマトリクスサイズを最小ディザ単位の整数倍にすれば、組織的ディザによる繰り返し処理の繋ぎ目もスムーズに移行でき都合がよい。なお、図１８では主走査方向に１０画素、副走査方向に８画素として最小ディザ単位を構成しているが、この割り振り方を主／副走査方向に入れ替えても別に差し支えない。
【００９９】
また、前記ディザマトリクスのサイズは、以下の条件にも合致していれば
【０１００】
【数４】

【０１０１】
により、例えば、４０×４８等の非正方のマトリクスサイズとしても良い。
【０１０２】
図１４、及び図１８の例では、最小画素数構成のディザマトリクスに対して強制的に斜め成分を持たせる構成とした例であるが、上述した手法を使えば、例えば、図１９に示すように、適当な階調数、適当な角度、適当なマトリクスサイズを持つスクリーンディザマトリクスに関して、この低階調部のパターンのみ使って同様に全マトリクス内の閾値を生成できる。この場合も当然前記２つの各マトリクスサイズの条件は満足する。
【０１０３】
また、再現する全階調数は常に２５６階調にする必要はなく、スクリーンディザマトリクスを使用する場合でも固有の階調再現数が決定されてしまうので、視覚を満足させる適当な階調数を再現できれば良い。
このように低階調部においては、周期的な組織的ディザの法則に沿ったパターンを用いて基準閾値配列を算出することになる。
【０１０４】
次に、他の好適な例について述べると、低階調部においては局所的に非周期的なランダムな出力特性を持つパターンとなるように、ディザの基準閾値配列を設定し、中間階調部から高階調部にかけては、相対的に印字精度が低い走査方向にドットが優先的に連なる非等方的な出力パターンとなるようにディザの基準閾値配列を設定する。
【０１０５】
これは低階調部において、周期的にドットを再現させるよりも、局所的に非周期的に分散させてドットを再現させた方が視覚特性上奇麗な出力が得られる場合が当てはまる。この場合も基準閾値配列の０〜２０％の範囲が、印字精度によって中間階調部から高階調部での出力特性とは異なる特性を示す範囲となる。すなわち、中間階調部から高階調部では、印字精度に関わらず濃度ムラやスジが常に見え易いが、低階調部では印字精度によって相対的に濃度ムラやスジが見えにくい状態が発生しやすいという特性がある。
【０１０６】
ここで実際に印字精度の変動による濃度ムラやスジが比較的目立たない低階調部においても、プリンタにより印字された出力の見え具合により、以下に示すように低階調部の閾値の設定を切り替えても良い。
【０１０７】
すなわち、低濃度部において印字ムラやスジが目立たないプリンタの特性の場合での閾値設定は、図１４に示すパターンの代わりに、完全に等方的で、確率統計的にマトリクスの主走査方向／副走査方向の各行／各列に均一数の出力ドットが発生するような出力パターンを求めこれを初期パターンとして用いても良いし、低階調領域の任意の均一階調を、最適化した誤差拡散アルゴリズムで処理して得られたパターンを初期パターンとして用いても良い。
【０１０８】
そしてこの初期パターンを用いて、それ以上の高階調部の閾値をランダムに作成するか、畳み込みフィルタを使って、相対的に印字精度が低い走査方向にドットが優先的に連なる非等方的な出力パターンとなる閾値を求める。また同様に、この中間階調部から高階調部にかけての連結強度は、印字精度に応じて最適に設定することが望ましい。また、低階調部側での閾値の割り振りは、組織的ディザのパターンのように最初から順序が決定されているわけではないので、予め順序を確率統計的に各低階調部においてマトリクスの各行／各列に均一数の出力ドットが得られる閾値を求めても良いし、高階調側の閾値算出手法を低階調側に当てはめて算出しても良い。マトリクスサイズも比較的小さく、低階調側の階調数はたかだか知れているので手動で最適化を行うことも容易である。
【０１０９】
また、低階調部においても実際に比較的濃度ムラやスジが目立てしまうプリンタの特性の場合は、低階調部から率先して相対的に印字精度が低い走査方向にドットが優先的に連なる非等方的な出力パターンとなるようにディザの基準閾値配列を設定する。この場合の閾値設定は、図１４に示すパターンの代わりに、完全に非等方的で、強制的にマトリクスの主走査方向に連なる出力ドットが発生するような出力パターンを求め、これを初期パターンとして用いても良いし、任意均一階調を主走査方向に出力が連続するように誤差拡散マトリクスの係数を最適化した誤差拡散で処理した出力パターンを初期パターンとして用いても良い。そしてこの初期パターンを用いて、それ以上の高階調部の閾値を前記手法と同様に求める。なお、別に低階調部において主走査方向に連結する特性を示す組織的ディザの閾値配列を用いても良い。
【０１１０】
これにより低階調部において、周期的にドットを再現させるよりも、局所的に非周期的にドットを再現させた方が視覚特性上好ましい場合においても、全階調域で最適な出力が得られるようになる基準ディザ閾値が求められる。また、中間階調部から高階調部では印字ムラやスジを目立たなくさせるために非等方的なパターンとなる基準ディザ閾値を得ることが出来る。
【０１１１】
先に述べたように、これにより生成した基準閾値配列の各種は、多値出力装置では印字精度に対して単独では大きな補正効果を持つことはないが、次に述べる各閾値プレーン間のシーケンスと組み合わせることで、濃度ムラやスジといった印字誤差を大きく緩和する作用を持てるようになる。
【０１１２】
上記では基準閾値配列について述べたが、次に上記で得られた基準閾値配列を各マルチレベルプレーン方向に展開する手法について述べる。
先に説明したように、多値ディザ処理のシーケンスにより、面積変調で階調を再現する出力装置ではそのドット出力特性は大きく異なってくる。
上記２つの基本構成（基準閾値配列と閾値プレーン間のシーケンス）において、印字ムラ、スジ等の印字精度の補償に関して言えば、複数の多値ディザ閾値プレーン間のシーケンスを変更することによる画質改善効果の方が大きい。
【０１１３】
プリンタのアーキテクチャにより閾値シーケンスが限られてくる場合とは異なり、本実施の形態のようなプリンタでは比較的容易に閾値プレーン間のシーケンスを変更できる。但し、基本的に閾値プレーン間のシーケンスを変更することにより、印字精度からくる印字ムラやスジ等を比較的容易に抑制する効果を持つ反面、解像度と階調再現性に対して大きく影響を及ぼすため、注意深く設計されていなければならない。
【０１１４】
また、ここでは閾値シーケンスに関して大きく２つの最適化のための工夫を取り入れている。
先ず、１つ目の最適化として、図１３を用いて説明する。図１３は先に説明したように、面積変調で画像を再現する出力装置の一般的な出力濃度特性であり、中間階調部から高階調部に比べて低階調部の再現分解能が低い。これは低階調部において、基本１階調の同一サイズのドットを用紙の白地部に階調が１ステップあがる毎に基準閾値配列に従って順番に配置していくよりも、第１基本階調のみでなく、第２基本階調以降のドットを適度に織り交ぜて印字した場合の方が階調再現上滑らかな隣接階調間の濃度変化を得られやすい場合があると言うことを意味することになる。
【０１１５】
この様子を図２０に示す。図２０の(a)は、図７の(a)と同じ、解像度を最も高くする場合のシーケンスによるドット成長行程である。一方、図２０の(b)に示すように、最小ドットのみの構成でなく、別のサイズのドットを織り交ぜてやることでも理論上は同じ濃度の出力を得ることができる。ここで図１３の出力特性から考えると、濃度的な変化は、図２０の(b)の成長行程の方が各隣接階調間では滑らかな濃度変化が得られる場合があると言うことである。但し、この場合、基本１階調ドットより大きなドットを低階調部に出力させるわけであるから、視覚上その大小の不快なパターンが見えないことが前提である。
【０１１６】
ここで最適なドット出力パターンを得るための好適な方法としては、図２１の(a)に示すような成長を示す基本１階調のドットのみを埋めていく出力パターンの代わりに、図２１の(b)に示すように、周期的な組織的ディザの出力パターンを示す階調までの部分のみ、つまり第１基本階調のドットを全体に配置する前に、数基本階調分、先にドットを成長させることである。このドットが視覚に大きく認識できない程度の基本階調であれば、図２１の(a)に示す出力パターンより遥かに滑らかな階調となって見えるわけである。
【０１１７】
この図２１の(b)の出力パターンに従った閾値シーケンスを図２２に示す。この例では、周期的な組織的ディザの出力を示す基準閾値は４までであり、この対応する位置に対して基本３階調までのドットを優先的に出力させている。なお、この処理は、出力パターン自体を視覚に対して好適なものにする効果の他、隣接ドット間が離れているため印字精度等にも強い出力パターンが得られると言う点においても有効である。この有効範囲は全入力画像データが取りうる階調範囲のうち、その値がおよそ０から１０％の範囲であると、効果がより発揮される。
【０１１８】
ここで入力画像データが取り得る階調範囲上限の１０％とは、周期的な組織的ディザ配列を実現するためのディザ基準閾値配列の規定の閾値範囲内における上限２０％のパターンを全閾値プレーン数の半分まで繰り返して優先的に出力させる範囲である。
【０１１９】
上記においては、周期的な出力パターンに対して、基本３階調までのドットを優先的に出力させる場合について述べたが、実際、このシーケンスの設定は、基本階調特性のドット径によって大きく左右される。つまりプリンタのもつ純解像度によって基本ドットサイズは決定されるが、例えば、同じ出力階調数でも３００ｄｐｉ／６００ｄｐｉでは基本となる最小ドットのサイズ、及びピッチは異なる。また、解像度が高くなればその実質的な設計の難易度により、理想ドット径に対して実測されるドットの特性は大きくずれた非線形的なものになってしまう。
【０１２０】
従って、これら様々な要因による異なる基本階調特性（特にドット径）によって、当然シーケンスの設定は、上記した規則に従って図２３の(a)、(b)に示すように、任意最適化される。また、このシーケンスの設定は、当然同様に印字精度により上記した規則に従って、図２３の(a)、(b)に同様に示すように、任意最適化される。
【０１２１】
一方、シーケンスに関するもう一つの好適な例、例えば、第１基本階調において均一にドットを割り振った場合において、第１基本階調のドットサイズが極めて微小で良好な特性をもつもので、周期的な組織的ディザ配列よりも視覚に満足する出力が得られる場合について述べると、経験的に入力画像の低階調側の０〜２０％で再現される画像に対しては、隣接の画素ピッチ間隔に対して、構成される画素のサイズが小さいため、濃度ムラや縦スジ等が目立たないことを利用して、この範囲にある入力画像に対しては空間周波数を上げるようにディザ閾値配列を与える。図２４に一例を示す。
【０１２２】
これにより入力画像データが低階調部のときに変換した多値画像データにより出現するドットサイズの種類は実質的により少なく、入力画像データが中間階調部から高階調部にかけての範囲のデータのときには、変換した多値画像データにより出現するドットパターンの種類が低階調部に比べて実質的に多くなる。
【０１２３】
これによりプリンタの階調再現では非常に重要な要素である低階調部での画素を目立たなくし、階調再現性を向上し、濃度ムラやスジが目立ちやすい部分は、ドットの種類を分散して濃度ムラやスジを目立たなくさせることができる。また、ランダムに閾値を配置させる場合とは異なり、各閾値プレーン間に相関があるため、基本ディザマトリクスから各プレーンの閾値を自動的に求めることができ、ハードウェアの簡素化も期待できる。
【０１２４】
さらに、様々な要因による異なる基本階調特性（特にドット径）によって、当然シーケンスの設定は、上記した規則に従って、図２５の(a)、(b)に示すように、任意最適化される。また、このシーケンスの設定は、当然同様に印字精度により上記した規則に従って、図２５の(a)、(b)に示すように、任意最適化される。なお、閾値シーケンスに関する例は、これもまた先に説明した基準閾値配列の最適化と組み合わせた場合、さらに効果を発揮する。
【０１２５】
図２６の(a)は、通常の局所的に非周期的で均一に分散化された基準閾値配列だけに着目した場合についてのおよそ中間階調部での閾値のオン／オフ特性を示す。また、図２６の(b)は、この実施の形態における、相対的に印字精度が低い走査方向にドットが優先的に連なる非等方的な出力特性を示すパターンとなるように生成した基準閾値配列によるおよそ中間階調部での閾値のオン／オフ特性を示す。
【０１２６】
一方、図２７の各パターンは、模式的にさまざまな多値ディザ処理で実際に用紙上に印字した場合の出力パターンの様子を示す図であるが、図中点線Ｃ−Ｃで示された部分に相当する画素が、例えばインクヘッドのミスディレクション等の影響で右方向にずれている。なお、図２７の(a)〜(c)はそれぞれ図７の(a)〜(c)のシーケンスにそれぞれ対応している。また、図２７の(c)は、基準閾値配列的には、図２６の(a)のように等方的規則に沿ったものである。
【０１２７】
一方、図２７の(d)は、同様にこの実施の形態による出力パターンの様子を示す図であり、基準閾値配列は、図２６の(b)の非等方的にしたものに相当する。図２７の(d)の場合においては、基準閾値配列を横方向に優先的に連結させている。つまり、横方向の隣接画素間の閾値が、相対的に近傍の値を取り易くなっており、横方向に優先的にドットが成長しやすい状態をとる。
【０１２８】
これにより図２７の(c)においてもそれなりに補正効果は期待できるが、さらに、図２７の(d)のような出力パターンを得ることができる、基準閾値配列及び閾値プレーン間シーケンスの構成にすることにより、印字位置精度が低い場合においても、より濃度ムラやスジをより目立たなくさせる効果を発揮する。
【０１２９】
なお、ここでは複数のディザ閾値プレーンに跨る閾値配列を、入力画像データが低階調となる領域にあるときには局所的に周期的で、かつ規則的な出力パターンが優先的に出力される閾値配列に設定する場合、及び複数のディザ閾値プレーンに跨る閾値配列を、入力画像データが中間階調から高階調となる領域にあるときには出現するドットパターンの種類が、入力画像データが低階調となる領域にあるときに出現するドットパターンの種類よりも多くなるように設定する場合の個々について述べたがこれのみに限定するものではなく、この両方の閾値配列に設定した複数のディザ閾値プレーンに跨る閾値配列を予め用意し、プリンタの出力精度に応じていずれの閾値配列を使用するかを選択するようにしてもよい。
【０１３０】
次に、多値ディザ閾値配列にガンマ補正を組み込む場合について説明する。 一般的に、擬似中間調処理で再現できる理論階調数は、単位マトリクス内の異なる閾値の総数で決定される。スクリーン系のディザは、パターンのサイズ、角度等の組み合わせ方によって決まる固有数の階調再現が可能であり、ストカスティック系ディザ及び誤差拡散では通常入力階調数と同じ２５６階調の再現が可能である。
【０１３１】
しかし、この理論階調数は擬似中間調処理部のみを想定した場合である。実際は擬似中間調処理前段の全ての画像処理部で階調損失が起こり得るので、最終的に擬似中間調処理部に入力される画像データは、限られた階調数でしかなく、擬似中間調処理部では全く使用されない出力パターンが存在するようになる。
【０１３２】
通常の画像処理の流れでは、色変換→ＢＧ／ＵＣＲ→ガンマ補正→擬似中間調の順となり、各画像処理部では、デジタル演算処理による丸め誤差、あるいは色域圧縮等による階調損失が発生する。
ここで色変換部、ＵＣＲ部での階調損失は、基本的に復元不可能であるので、ガンマ補正部と擬似中間調部における階調再現性について述べる。
【０１３３】
ガンマ補正処理は、エンジンの基本階調特性を、例えば、輝度リニアや濃度リニアなどのターゲット特性に補正するための処理であり、図２８に示す関係がある。つまり測定されたエンジンの基本階調特性からターゲット特性に対して対象となるガンマ補正曲線を用いて入力画像データを変換することにより、最終的に出力される階調特性をターゲット特性に合わせ込む処理である。なお、図ではターゲットカーブは直線であるが、任意の曲線で置き換えることもできる。
【０１３４】
一般的に、ドットを円で表現する面積変調の出力装置の基本特性は、ターゲット特性よりガンマが立った、図２８のグラフｇの直線より上側の特性となる。γ補正の実処理としては、各色デジタル１ＬＵＴ演算で行われることが多い。
【０１３５】
図２８において、デジタル１ＬＵＴによる演算では、具体的に図２９に示すような変換が行われる。低階調部では、デジタル丸め誤差により複数個繰り返し同じデータに変換され、高階調部においては飛び飛びの値に変換される。つまり、階調再現上重要な低階調部の再現においては、異なる入力画像であっても出力されるハーフトーンパターンは全く同じパターンとなり易く、高階調部では使用されないハーフトーンパターンが存在し、全体として階調再現数が減少し、画像処理上非常に効率が悪いものとなってしまう。この現象は、エンジンの基本特性がターゲット特性から離れているほどデジタル変換精度が落ち、階調再現数が大幅に減少するようになる。但し、インクジェットプリンタに関して言えば、比較的理想に近い特性を持っている。
【０１３６】
そこで、ここでは、ガンマ補正をハーフトーン処理内部に組み込み、階調損失を理論的に抑制するマトリクスを生成する。ここで２値のディザ処理に関してはガンマ補正を組み込んだディザ閾値生成方法については周知であるが、多値ディザ処理の場合は、各プレーン間の基本階調特性が線形的ではないことから、様々な閾値プレーン間のシーケンスにすべて対応するようにした場合、従来の手法は適用困難である。
【０１３７】
例えば、1画素８値、マトリクスサイズが３２×３２の時、ある任意の階調に対して、同時にＯＮ／ＯＦＦが切り替わるドット数は３２×３２×（８−１）／２５５≒２８個であり、通常の擬似中間調処理では、この個数は各階調均等に割り付けされている。
【０１３８】
本処理の基本原理は、このハーフトーン処理において、画素のＯＮ／ＯＦＦを決定する全閾値プレーン間の閾値にガンマ変換特性を組込み、全閾値プレーンの各閾値処理において画素をＯＮさせる個数を、ガンマ特性に合わせて制御する。つまり、階調損失を引き起こすデジタル変換のガンマ補正部をスルーして、擬似中間調部でのＯＮドット総数の調整だけで処理を実現することにより、実質階調数を復元する。この場合のＯＮ数というのは、多値の場合、画素レベル方向、すなわち、７閾値レベルのうちの何番目の閾値までＯＮしたかにより１画素につき最大７個のＯＮ数があり、この数を示している。
【０１３９】
このガンマ補正が組み込まれた多値ディザ閾値配列の算出の仕方を、図３０の流れ図に示す。先ず、ステップＳ１１にて、予めドット数が均等に割り振られた多値ディザマトリクスを用いた多値ディザ処理により実質的なエンジンの階調特性を得る。次に、ステップＳ１２にて、γターケッドの決定を行い、ステップＳ１３にて階調を０にセットする。
【０１４０】
続いて、ステップＳ１４にて、決定されたターゲット特性に合わせて通常のガンマ変換と同様に、各入力階調値に対し、このターゲット特性にあわせて変換するガンマ補正階調値を算出する。このときガンマ補正階調値は整数ではなく、実数として計算させると、より精度を向上できる。
【０１４１】
続いて、ステップＳ１５にて、算出した出力ガンマ補正階調値から、この値に最も近い値を、変化するドット数が各階調均等に割り振られた多値ディザマトリクスの出力特性の曲線上から得てこのときのＯＮドット数に換算する。なお、出力曲線は実際に測定した点を用いて任意補間したものである。この時、出力ガンマ補正階調値を実数で計算させると、１ドット単位までの分解能が得られる。
【０１４２】
続いて、ステップＳ１６にて、閾値優先順位表からＯＮ画素分の抽出を行い、ステップＳ１７にて、オンさせるドット数に応じて、全閾値プレーン間の優先順位の小さい順に多値ディザの閾値を割り当て、ステップＳ１８にて、階調を１つインクリメントする。そして、ステップＳ１４からＳ１８の処理を全階調にわたって繰返し行うことで、全閾値プレーンにおいてすべての閾値を埋めることが出来る。
【０１４３】
この時、多値ディザの基準閾値配列は、既に、全ての優先順位が計算されているため、この優先順位を図３１の例に示すように、各シーケンスに沿って、全閾値プレーン間において予め展開し、最終的な全閾値プレーン間の優先順位を求めておく。なお、低階調部において周期的なパターンを出力させる場合は、基準閾値配列の優先順位は組織的ディザの規則に沿った優先順位とし、中間階調から高階調にかけては基準閾値配列を求める計算過程で既に基準閾値配列内の優先順位が決定されている。
【０１４４】
図３１の例においては、説明を簡単にするため基準閾値は４まで、閾値プレーンは３プレーンとしている。なお、実際には基準閾値の優先順位は１から１６間である。ここで、図３１の閾値プレーン間のシーケンスを見てみると、まず基準閾値１及び第１閾値プレーンの部分が対象となり、この部分に相当する部分に優先順位が割り振られていく。次に基準閾値１及び第２閾値プレーンの部分が対象となり、この部分に相当する部分に次の優先順位が割り振られていく。これをシーケンスの全順番の１〜１２まで行うことにより、全閾値プレーン間に１から４８の優先順位が割り振られる。
【０１４５】
そしてこの優先順位に沿って対応する階調の画素のオンする数だけ、それに対応する出力を示す閾値を設定していくことにより、どのような複雑な閾値プレーン間のシーケンスであっても、全閾値プレーンにおいて閾値が一意に決定される。
【０１４６】
以上により、本実施の形態においては、基準閾値配列と閾値プレーン間のシーケンスを最適に組み合わせることで、１つのディザ閾値プレーンの組で、印字精度や、実際のドットの出力特性に応じて、各階調間で最適な出力特性となる多値ディザ処理を行うことが可能となり、さらに、この多値ディザ閾値自体にガンマ補正処理を組み込むことで、より階調再現性の高い画像を得ることが可能となった。
【０１４７】
なお、上記した実施の形態では、基本的に、例えば、ブラックの場合など単色での構成について述べたが、これをこのままカラー画像に拡張することは容易に実現できる。但し、注意を要するのは色間の出力パターンの関係に関して若干の考察が必要となる。
【０１４８】
カラー画像の場合、通常各色毎に多値ディザ処理を行う。ここでは少なくとも２色以上の色に対して上記した実施の形態の基準閾値配列を持つ多値ディザ処理の構成を適用する。
【０１４９】
いくつかの色は上記した実施の形態の基準閾値配列の構成を使用しなくても良い。例えば、Ｙｅｌｌｏｗのように視覚に極めてドット粒子が目立ちにくい色に関しては、上記した実施の形態の基準閾値配列の構成を用いず、単純な従来型のディザ閾値配列を適用しても良いし、Ｂｌａｃｋのようにエッジをより強調させたいような色の場合は、多値誤差拡散処理を適用して、よりエッジ効果を強める処理を行うこともできる。
【０１５０】
一方、上記した実施の形態の処理を適用する色の場合は、例えば、全く同じ閾値パターンを各色に適用すると、Ｄｏｔ−Ｏｎ−Ｄｏｔの出力パターンとなり、出力特性の変動等によりドット印字位置がずれた場合、色ムラ等に弱くなってしまうという問題があるため、各色毎に基準閾値配列を異ならせる必要がある。この場合に各色毎に基準閾値配列を個別に作成しても良いが、一度作成した基準閾値配列を反転や回転、あるいはシフトといった操作により作成した方がより容易に実現である。
【０１５１】
これは一般的に、２値のプリンタの場合は、各色のパターンの相関性により、より色モアレに関してはシビアな設計が要求されるが、多値のプリンタの場合は、各色のパターン自体の組み合わせによる色モアレは発生しにくいため、比較的簡単な閾値の変更操作により、高精細な画像が得られることが期待できるからである。また、もともと分散性の強い閾値配列でもあるので上記閾値操作でも十分である。但し、低階調部を周期的な規則的組織的ディザで画像を再現する場合は、それに対応する階調部分に対しては周期性が非常に強いパターンとなるために、色間の干渉を考慮して適度な閾値設計が必要となる。
【０１５２】
この好適な一例としては、最小ディザ単位における閾値の割り当てを、例えば、図３２に示すように反転や回転あるいはシフトといった操作により再配置するか、あるいはまったく別のパターンを新規に作成し、この新規作成した基準閾値パターンをもとに全マトリクス内の閾値を再生成すれば良い。この時、図３２に示すように各色間での低階調部においては、同じ位置にドットが重ならず、なるべく並置されるようにドットが配置されることが好ましい。これは低階調部であるにもかかわらず、ドットが重なってしまう部分は実質２次色となり、ドットの存在自体がより視覚に目立ってしまうからである。さらに、この階調部分は周期性を持たせてあるため、この周期性も目立ちやすくなってしまうからである。
【０１５３】
一方、カラーに関する閾値プレーン間のシーケンスに関しては、各色成分の実質上の印字精度によりそのシーケンスを最適に設定することが可能である。これは色毎の各多値ディザ処理が独立に処理されるためにであり、この構成は容易に実現できる。
【０１５４】
また、統計的な印字精度が同じでも、一般的に各色により濃度ムラや縦スジの視覚への影響は大きく異なることが知られている。例えば、同じ印字精度の時は、Ｙ→Ｃ→Ｍ→Ｋの順に、より視覚にノイズとして目立つとされている。そこで、ここでは、この各色による多値ディザ処理において、各色毎に閾値プレーン間のシーケンスを適宜変更して擬似階調処理を行うことにより、より最適な出力画像を得ることができる。
【０１５５】
以上、ここでは、ＣＭＹＫの４色のカラーについて述べたが、これは４色に限らず、ＣＭＹの３色、あるいは他の色の組み合わせでも容易に実現できることは言うまでもない。また、この実施の形態では全般にわたって多値ディザ処理について説明したが、閾値プレーン間シーケンスの設定等、多値に限定された処理の部分を除けば、大部分は２値のディザ処理にも容易に適用できるものである。
【０１５６】
なお、この実施の形態では多値ディザ処理について説明をしたが、これに限る必要はなく当業者であれば容易に濃度パターン法等にも応用できる。つまり、多値ディザ処理においては入力画像データとディザ閾値が１：１の対応で入力画像データと最終的に出力される画像データは１：１の関係となるのに対して、濃度パターン法においては入力画像データと変換閾値が１：Ｋ（Ｋ≧２）の対応で入力画像データと最終的に出力される画像データは１：Ｋ（Ｋ≧２）関係となるだけである。当然、Ｋは主走査／副走査方向の両方、あるいはどちらか一方にどのように拡張してもよい。
【０１５７】
また、この実施の形態では２次元平面画像出力手段としてプリンタを使用した場合について述べたが必ずしもこれに限定するものではなく、ＣＲＴディスプレイや液晶ディスプレイなどのディスプレイを２次元平面画像出力手段として使用してもよい。
【０１５８】
【発明の効果】
本発明によれば、粒状性を向上させ、かつ写真画像に適し、階調再現性に優れた擬似階調処理ができる。
【図面の簡単な説明】
【図１】本発明の実施の形態における全体のハードウェア構成を示すブロック図。
【図２】同実施の形態における画像処理部の構成を示すブロック図。
【図３】同実施の形態におけるプリンタエンジンの構成を示すブロック図。
【図４】同実施の形態における各階調の画素サイズを示す図。
【図５】同実施の形態における擬似階調処理部の構成を示すブロック図。
【図６】同実施の形態におけるディザ基準閾値配列を示す図。
【図７】同実施の形態における閾値プレーンにおけるシーケンスの一例を示す図。
【図８】図７の多値ディザ処理による各出力例を示す図。
【図９】図７の(a)のシーケンスにおける画素成長例を示す図。
【図１０】低階調部においてランダムに分散されたドットパターンと規則的に分散された組織的ディザによるドットパターンを示す図。
【図１１】分散系の組織的ディザの出力パターンにおける角度変化の例を示す図。
【図１２】分散系の組織的ディザの出力パターンにおける特異模様の出現例を示す図。
【図１３】多値ディザ処理の基本階調特性を示すグラフ。
【図１４】同実施の形態における基準閾値配列を示す図。
【図１５】同実施の形態における閾値生成処理を示す流れ図。
【図１６】同実施の形態における各基本ディザ閾値におけるドット再現を示す図。
【図１７】周期的な組織的ディザで構成する範囲を基準閾値範囲の２０％にした場合のパターンを示す図。
【図１８】同実施の形態における他の基準閾値配列を示す図。
【図１９】同実施の形態における他の基準閾値配列を示す図。
【図２０】同実施の形態における各シーケンスの画素成長例を示す図。
【図２１】同実施の形態における各シーケンスの他の画素成長例を示す図。
【図２２】同実施の形態におけるシーケンスの一例を示す図。
【図２３】図２２に示すシーケンスの変更例を示す図。
【図２４】同実施の形態におけるシーケンスの他の例を示す図。
【図２５】同実施の形態におけるシーケンスの他の例を示す図。
【図２６】同実施の形態における基本ディザ閾値におけるドット再現を示す図。
【図２７】印字ムラを含んだ各シーケンスにおける出力パターンを示す図。
【図２８】ガンマ特性及びその補正を説明するための図。
【図２９】通常のテーブル変換によるガンマ変換を示す図。
【図３０】同実施の形態におけるガンマ補正を組込んだ全閾値プレーン間のディザ閾値を決定する処理を示す流れ図。
【図３１】同実施の形態において全閾値プレーン間の優先順位を求める操作を説明するための図。
【図３２】同実施の形態における各色間の低階調部のドットの配置関係を示す図。
【図３３】ライン記録ヘッド及びその印字例を示す図。
【図３４】ディザマトリクスを使用して行う２値のディザ処理を説明するための模式図。
【図３５】図３４の２値のディザ処理による印字出力例を示す図。
【図３６】ディザマトリクスを使用して行う多値のディザ処理を説明するための模式図。
【図３７】図３６の多値のディザ処理による印字出力例を示す図。
【図３８】多値のディザ処理におけるシーケンスを示す図。
【符号の説明】
２４…擬似階調処理部
３２〜３５…インクジェットヘッド
５１…主カウンタ
５２…副カウンタ
５３…エンコード部
５４…ＬＵＴ部[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing apparatus used for a printer, a copying machine, a facsimile, an MFP (Multi-Function Peripheral), etc., which dithers multi-value input image data and converts it into image data having a smaller number of gradations. The present invention relates to a color image processing apparatus.
[0002]
[Prior art]
Conventionally, in an image forming apparatus such as a printer using a line head such as a line LED (light emitting diode) head, a line thermal head, or a line inkjet head, dots of the same size are printed on a recording sheet while maintaining the resolution of the head. A binary image was formed. That is, in the case of a line LED head, dots having the same size are printed on the recording paper while maintaining the spacing in the raster direction of a plurality of LEDs that are a plurality of recording elements arranged in a line, and in the case of a line thermal head, A plurality of heating elements, which are a plurality of recording elements arranged in a line, are printed on the recording paper with the same spacing in the raster direction, and in the case of a line inkjet head, a plurality of lines arranged in a line A binary image is formed by printing dots of the same size on the recording paper while maintaining the intervals in the raster direction of the plurality of ink ejection openings as the recording elements. In addition, it is generally performed by scanning these heads a plurality of times to cope with a resolution higher than an element interval.
[0003]
In an image forming apparatus equipped with such a recording head, a character / line image is simply reproduced as a binary image corresponding to the resolution of the head or the scanning interval, and the graphic / photo image is represented by a systematic dither method or The image was reproduced by pseudo gradation processing such as an error diffusion method. In the pseudo gradation processing in this case, it is very difficult to maintain both high resolution and high gradation reproduction. In particular, in the systematic dither processing, the resolution and the gradation are in conflict. Note that the pseudo gradation processing is also used for color characters and light and shade characters.
[0004]
On the other hand, in an image forming apparatus equipped with such a recording head, multi-value image data generated by multi-value dither processing is used for input image data, and the output area within one pixel is modulated to modulate the output area within one pixel. An image forming apparatus capable of expressing the image with several levels of gradation has also appeared. FIG. 33 shows an example of the state of a recording head composed of a plurality of recording elements used in these apparatuses and the output dots. In the figure, 1 is a recording head, and 2 is an ink discharge port. Reference numeral 3 denotes an output dot. FIG. 33 shows a dot output example of an image forming apparatus that can represent one pixel in three values including white for simplicity. Further, by arranging four or three of these linear recording elements in parallel, a color image of C (cyan), M (magenta), Y (yellow), K (black), or a color image of CMY. Can be recorded.
[0005]
In such an image forming apparatus capable of printing multi-value image data, after performing various image processing such as color conversion processing, UCR (under color removal) processing, or gamma correction, the specified number of gradations specific to the printer engine In order to reproduce the above, multi-value pseudo gradation processing such as multi-value dither processing using a screen angle for each color or multi-value error diffusion processing is performed to obtain multi-value image data of 1 pixel number of bits. Then, a larger amount of information is concentrated on one pixel to improve image reproducibility.
[0006]
In general, systematic dither processing is light in processing, has a high degree of freedom in configuration, has high speed, and can reduce costs. However, it is said that the error diffusion process is superior in terms of image quality. The systematic dither process cuts out the quantization error caused by the threshold process as it is, whereas the error diffusion process stores the quantization error in the peripheral pixels. As a result, in the error diffusion process optimized from the viewpoint of the output characteristics, the output pattern is the output pattern with high frequency characteristics that is most inconspicuous from the viewpoint of human visual characteristics, and the edge preservation effect is also large compared to the systematic dither processing. This is an advantage in terms of image quality.
[0007]
On the other hand, in the case of multi-value pseudo halftone processing, it has been found that the difference in image quality does not occur as much as in the case of binary. This is because, as an effect of multi-leveling, the quantization error that is rounded down becomes much smaller as the level of multi-leveling is increased than in the case of binarization. In particular, as the number of gradations that can be expressed by one pixel at high resolution increases, the difference in image quality disappears.
[0008]
Furthermore, recently, a method has been developed that achieves output characteristics equivalent to error diffusion processing at the same high-speed processing as systematic dither processing by using a fixed mask dither with improved stochastic dither or clusters.
[0009]
[Problems to be solved by the invention]
By the way, the general binary dither processing basically only needs to consider the threshold arrangement of a single dither matrix of one single plane, and 2 by pixel-to-pixel comparison between the input pixel and the threshold of the dither matrix at the corresponding position. A value output image is obtained. This is shown in FIG. FIG. 34 is a schematic diagram when a known 4 × 4 Bayer type dither matrix is used. Here, for simplification of explanation, the threshold value of the dither matrix corresponding to input 4 bits is compared with the input image. For example, if the input pixel value is larger than the corresponding threshold value of the dither matrix, 1 (black), In this example, 0 (white) is output and a binarized output state having a combination of 1 or 0 as a whole is obtained.
[0010]
Here, as shown in FIG. 34, the dither matrix is repeatedly used on the tile at the basic dither matrix (reference threshold array) size period, and the above-described processing is similarly performed on all input pixels. Also, in an output device such as a general printer, pixels cannot be configured with a digital square lattice, and the output is often close to a circle due to process limitations of each output device. The state of output in this case is shown in FIG. When a solid image is generally printed as shown in the dot size of FIG. 35, the shape of the print pixel completely covers the ideal square pixel so that no gap is generated, that is, a resolution pitch of √2 times or more. Designed to be a circle with a diameter.
[0011]
On the other hand, in the multi-value dither processing, in addition to the basic dither matrix arrangement described above, consideration in the depth (pixel level) direction is also necessary. For example, when D-value multi-value dither processing is performed, (D-1) threshold planes are required, and the dither threshold value of each threshold plane is compared with the input image to obtain an output image of D value. FIG. 36 shows a schematic schematic diagram of the multi-value dither processing in this case, and FIG. 37 shows the output state. FIG. 36 is a schematic diagram showing 8-level multi-value output including 0 (white).
[0012]
At this time, generally, in the dither processing, it is better in terms of image quality to have some correlation between the threshold planes. Therefore, (D-1) dither matrix threshold values are automatically set based on this reference threshold array. Often calculated to
[0013]
As the multi-value dither processing in consideration of the correlation between the planes, there are two sequences shown in (a) and (b) of FIG. For the sake of simplicity, FIG. 38 shows multi-value dither processing for converting input 8-bit image data into a one-pixel four-value (2-bit) image using a 2 × 2 basic threshold array.
[0014]
The sequence method of FIG. 38 (a) is a method of filling the thresholds in units of planes in ascending order, and is basically hardly influenced by the appearance state of dots of adjacent pixels, such as an ink jet printer, so This is a dither process used in a printer that can stably reproduce image formation. The resolution is almost equal to the resolution performance of the engine, is very high, the dot density is high, and is an ideal method for reproducing an image by area modulation. However, since the screen is easily filled with pixels of the same size and proximity size, it is easily affected by printing accuracy.
[0015]
The sequence method of FIG. 38 (b) is a method in which threshold values are sequentially filled in any one pixel to be processed in ascending order, such as a laser printer or a thermal printer. This is a dithering process that is often used in printers that are easily affected and are difficult to form with a single pixel and are unstable. This is a case where the resolution is low and the dot density is coarse. When this dither threshold array is a dot-concentrated type, an image called a halftone dot is formed. Since the resolution is low, minute print accuracy unevenness in pixel units is absorbed.
[0016]
In both cases, all threshold values are automatically derived by defining one reference threshold plane and the pixel growth order in the depth direction.
[0017]
Further, regarding color image formation, in recent printer image quality design, the importance of gradation reproducibility equivalent to that of photographic image quality, particularly including the highlight portion, is increasing. In particular, a gradation reproduction method for further improving the graininess is one important technical problem. As a technique for satisfying this graininess, a light ink is used in addition to standard four inks of C (cyan), M (magenta), Y (yellow), and K (black). There is a method of improving the graininess of the highlight portion by combining inks such as light magenta. However, the number of recording heads and driving mechanisms is increased by the number of added inks. Further, when the recording head is a head having the same number of nozzles as the line head for each color, it becomes a heavy burden in terms of cost.
[0018]
Considering the case of four colors C, M, Y, and K, multi-value dither processing includes halftone dot dither using a screen angle, distributed dither represented by Bayer, or an intermediate cluster. Various methods such as dithering have already been developed.
[0019]
However, these dithering processes have many problems. For example, when a halftone dot using a screen angle is applied to dither processing, moire such as rosetta occurs due to interference between colors. In addition, when a distributed dither matrix such as the conventional Bayer type is used, the degree of freedom of dot arrangement is small, and a visually noticeable texture is generated at a specific gradation portion. Thus, there are still many problems to be solved in order to obtain optimum output characteristics over all colors and all gradations.
[0020]
These are not limited to binary values, and the same phenomenon occurs even when applied to multiple values. In particular, it occurs remarkably in the dither processing of the sequence shown in FIG. 38B, but does not disappear completely even in the dither processing of the sequence in FIG.
[0021]
Furthermore, what can be said in general for these systematic dither processes including the cluster type is that the periodicity is easily noticeable over the entire input gradation range. In particular, in the case of an output device having a relatively low resolution such as a printer, there is a problem that its periodicity is very conspicuous visually.
[0022]
As described above, the conventional fixed period type dither still has various problems, and the design of the reference threshold array considering each output characteristic with different characteristics for various output devices has been improved. There is room to do.
[0023]
Recently, a method of realizing output characteristics equivalent to error diffusion processing at the same high speed processing as systematic dither processing by using a fixed mask dither with improved stochastic dither or cluster has been developed. A suitable example is "The Void-and-Cluster Method for Dither Array Generation" (SPIE / IS & T Symposium on Electronic Imaging Science and Technology, San Jose, CA, February, 1993) by Robert Unichney. However, in these processes, only theoretical output characteristics in an ideal system are considered, and the output characteristics in a dot overlap model of a binary printer are considered at most.
[0024]
Further, depending on an output device capable of multi-value reproduction or the like, an optimal output result may not be obtained through all input gradation images. (Even in classic dither processing, if it is limited to a specific gradation, the result is more visually pleasing than any improved stochastic dither pattern.)
Accordingly, the actual output characteristics and the like of each unique output device are not taken into consideration, and the actual characteristics are hardly taken into consideration for multi-level output devices.
Therefore, the present invention provides an image processing apparatus and a color image processing apparatus that can improve gradation reproducibility of an output device such as a printer.
[0025]
[Means for Solving the Problems]
In the present invention, input gradation image data of 1 pixel M gradation is dithered by a dither processing means using a reference threshold value array, and converted to output image data of 1 pixel N (M> N ≧ 2) gradation. In the image processing apparatus for outputting an image from the two-dimensional planar image output means, the dither processing means is configured such that the reference threshold value array is locally periodic in a region having a relatively low gradation within a prescribed threshold value range, and The image processing apparatus has a regular threshold arrangement characteristic and is set to have a locally non-periodic threshold arrangement characteristic in a region that is relatively intermediate to high gradation within a prescribed threshold range.
[0026]
Further, according to the present invention, multi-level dither processing is performed on the input grayscale image data of 1 pixel M grayscale by using a threshold value array straddling a plurality of dither threshold planes by the dither processing means, and 1 pixel N (M>N> 2). In the image processing apparatus that converts the output image data to gradation and outputs the image by the two-dimensional plane image output means, the dither processing means sets the threshold value array across a plurality of dither threshold planes, and the input image data has a low gradation. The image processing apparatus is set to a threshold value array in which a local periodic and regular output pattern is preferentially output by the two-dimensional planar image output means.
[0027]
In the present invention, input gradation image data of 1 pixel M gradation is dithered by a dither processing means using a reference threshold value array and converted to output image data of 1 pixel N (M> N ≧ 2) gradation. Then, in the image processing apparatus that outputs an image by the two-dimensional planar image output unit, the dither processing unit divides the region within the specified threshold range in the reference threshold value array into two, and the reference threshold value array of each of the divided regions Are set in such an image processing apparatus that the output image by the two-dimensional planar image output means has a pattern showing different output characteristics in each region.
[0028]
Further, the input gradation image data of 1 pixel M gradation is converted into output image data of K (K ≧ 2) pixel N (N ≧ 2) gradation using the reference threshold value array by the image conversion means, and then two-dimensionally. In the image processing apparatus that outputs an image by the planar image output means, the image conversion means is configured such that the reference threshold value array is locally periodic and regular in a region where the gradation is relatively low within a prescribed threshold value range. The image processing apparatus has threshold arrangement characteristics and is set so as to have locally aperiodic threshold arrangement characteristics in a region in which a middle gradation and a high gradation are relatively within a prescribed threshold range.
[0029]
In the present invention, color input gradation image data of 1 pixel M gradation is subjected to dither processing using a reference threshold array by a dither processing means, and output image data of 1 pixel N (M> N ≧ 2) gradation is obtained. In the color image processing apparatus that outputs the image by the two-dimensional planar image output means after the conversion, the dither processing means sets the reference threshold value array for at least two kinds of color components to a relatively low gradation within a specified threshold range. A threshold arrangement characteristic that is locally periodic and regular in a certain area, and a non-periodic threshold arrangement characteristic that is relatively non-periodic in a region that is relatively intermediate to high gradation within a prescribed threshold range. It is in a color image processing apparatus set to have.
[0030]
In the present invention, color input gradation image data of 1 pixel M gradation is subjected to dither processing using a reference threshold array by a dither processing means, and output image data of 1 pixel N (M> N ≧ 2) gradation is obtained. In the color image processing apparatus that outputs an image by the two-dimensional planar image output means after conversion, the dither processing means is an area that has a relatively low gradation within a prescribed threshold range for at least two types of color components. Criteria that have locally periodic and regular threshold arrangement characteristics and a local non-periodic threshold arrangement characteristics in a region that is relatively intermediate to high gradation within a specified threshold range The first dither processing function for performing dither processing using the threshold array, and the dither processing using the reference threshold array having a periodic threshold array characteristic that is locally periodic with respect to the remaining color components. Second dithering to do In the color image processing apparatus having a function.
[0031]
Further, according to the present invention, dither processing is performed on the color input gradation image data of 1 pixel M gradation by using the reference threshold value array by the dither processing means to output image data of 1 pixel N (M> N ≧ 2) gradation. In the color image processing apparatus that outputs an image by the two-dimensional planar image output means after conversion, the dither processing means is an area that has a relatively low gradation within a prescribed threshold range for at least two types of color components. Criteria that have locally periodic and regular threshold arrangement characteristics and a local non-periodic threshold arrangement characteristics in a region that is relatively intermediate to high gradation within a specified threshold range A color image processing apparatus having a first dither processing function for performing dither processing using a threshold array and a second dither processing function for performing dither processing by error diffusion processing for the remaining color components.
[0032]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described with reference to the drawings. In this embodiment, the present invention is applied to a color ink jet printer.
FIG. 1 is a block diagram showing the overall hardware configuration, in which color image data of 1 pixel M gradation is transferred from a host computer 11 to a printer 12 which is a two-dimensional planar image output means. That is, the host computer 11 transfers code or raster data from the driver 111 to the printer controller 121 of the printer 12 in accordance with the interface characteristics with the printer 12.
[0033]
The printer 12 controls the printer engine 122 by the printer controller 121. The printer controller 121 incorporates the coded image data sent from the host computer 11, for example, a page description language such as PDL into a bitmap and performs each image processing. Stored in image memory.
[0034]
The printer engine 122 converts bitmap image data from the printer controller 121 into a drive signal, and performs a printing operation by transporting paper, driving a color inkjet head, and the like.
[0035]
The host computer 11 and the printer 12 do not necessarily have a one-to-one relationship, and may be used as a network printer in a recently popular network. In this case, the relationship is a multiple-to-one relationship. The interface between the printer controller 121 and the printer engine 122 basically depends on the printer architecture and is not specified.
[0036]
FIG. 2 is a block diagram showing an example of the configuration of the image processing unit in the printer controller 121. The color conversion processing unit 21, the BG / UCR processing unit 22, the gamma (γ) correction unit 23, and the pseudo gradation processing unit 24 are shown. For example, a standard RGB color signal inputted to an 8-bit monitor for each color is first converted into CMY colors of color reproduction colors in a printer by a color conversion processing unit 21, and a BG / UCR processing unit 22. To supply. Note that R, G, and B indicate red, green, and blue colors, and C, M, and Y indicate cyan, magenta, and yellow colors.
[0037]
The BG / UCR processing unit 22 extracts the black component from the CMY color, determines the subsequent CMY color, and finally converts it to the CMYK color and supplies it to the gamma correction unit 23. K represents black.
[0038]
The gamma correction unit 23 performs density correction on the CMYK colors according to the actual output characteristics of the printer, and supplies it to the pseudo gradation processing unit 24. The pseudo gradation processing unit 24 converts the data of one pixel into multi-valued image data having a smaller number of gradations of about 2 to 4 bits for each color according to the printing capability of the printer 12 by multi-value dither processing for each color. It is supposed to convert.
[0039]
FIG. 3 is a block diagram showing a hardware configuration of the printer engine 122. The printer engine 122 includes a control unit 31. The control unit 31 uses the multi-valued image data of each color bit and the cyan inkjet head 32 uses the image data from the printer controller 121. The magenta ink jet head 33, the yellow ink jet head 34, and the black ink jet head 35 are respectively driven and controlled, and the head moving device 36 for reciprocating the heads 32 to 35 in the direction of the rotation axis of the rotary drum, and the printing paper are rotated. A sheet conveying motor 37 that conveys the drum, a drum motor 38 that rotates the rotating drum, and a sheet fixing device 39 that includes a charging roller that charges and fixes a printing sheet wound around the rotating drum are driven and controlled.
[0040]
The printer engine 122 is provided with a reciprocating mechanism in which the heads 32 to 35 are mounted side by side along the rotation axis direction of the rotary drum, and the print paper conveyed by the paper conveyance motor 37 is wound around the rotary drum. The wound printing paper is charged and fixed by the paper fixing device 39, and then the drum motor 38 is used to rotate the rotating drum and the ink jet heads 32 to 35 are driven based on the print data. When the rotary drum rotates once, the inkjet heads 32 to 35 are moved by a half of the interval between the ink discharge ports, and the inkjet heads 32 to 35 are continuously used as print data. Driven on the basis of a sheet of paper when the rotating drum rotates twice. Printing is completed, thereby, adapted to be printed at twice the resolution of the ink discharge interval of the ink jet heads 32 to 35 with respect to the print paper.
[0041]
The pseudo gradation processing unit 24 constitutes a main part of the present invention. For the function of this processing unit, for example, 8-bit, 256 gradation (0: white, 255: black) input gradation image data is simulated. An example will be described in which halftone processing is performed to convert each color to 3 bits and 8 gradations (0: white, 7: black). Note that the input and output are not limited to the above-mentioned number of gradations, and it can be easily understood from the following embodiments that the number of gradations can be changed.
[0042]
When a 3-bit image of each color can be handled as the capability of the printer, for example, multi-value image data of each color 3-bit can be obtained by pseudo gradation processing. As shown in FIG. 4, a total of 8 gradations including white can be reproduced in one pixel by using seven types of variable dot sizes for each pixel as shown in FIG. This is called the basic 8-gradation characteristic.
[0043]
In general, it is desirable that the size of each dot of each gradation is adjusted in advance for each color so that it has a linear characteristic if possible, but it is perfectly adjusted due to process limitations. Is almost impossible. For example, with regard to this ink jet printer, it is relatively easy to realize that the ink discharge volume has a linear characteristic for each drop, rather than linearly bringing the luminance and density.
[0044]
Further, it is not impossible to adjust the target characteristics by adjusting the number of ink drops and the drive waveform of each gradation, but in this case, the drive waveform becomes complicated and it is easy to perform redundant processing. Further, even if this basic 8-gradation characteristic is matched with the target characteristic, deviation from the target ideal gradation curve is inevitably caused when reproducing all 256 gradations by the pseudo halftone process. These characteristics are greatly affected even if the characteristics of the paper used are slightly different.
[0045]
Therefore, in terms of design, the simplest method is to make the configuration as simple as possible so that the gradation characteristics are not greatly distorted, and to correct the print characteristics of the engine by processing such as gamma correction. However, at least the dot size of the maximum gradation value is generally formed as a circle with a diameter larger than that completely covering the square pixel of the pure resolution possessed by the engine.
[0046]
The pseudo gradation processing unit 24 performs multi-value dither processing, and its basic hardware configuration will be described with reference to FIG. Note that the realization configuration of the multi-value dither processing may basically adopt any realization method, and FIG. 5 is an example.
[0047]
Reference numeral 51 denotes a main counter which periodically counts with an arbitrary constant in the main scanning direction. Here, the size corresponds to a period up to 128 pixel count in the main scanning direction. A sub-counter 52 periodically counts with an arbitrary constant in the sub-scanning direction. Here, the size corresponds to a period up to 128 pixel count in the sub-scanning direction.
[0048]
Reference numeral 53 denotes an encoding unit, which encodes an encoded MAX 6-bit code from the counter values input from the main counter 51 and the sub-counter 52 based on a multi-plane dither threshold array corresponding to the position. Output. Here, the maximum number of reproduced gradations that does not exceed 256 gradations by multi-value dither processing when the input image data is 8 bits, 256 gradations, 3 bits, 8 gradations after pseudo gradation processing, is set to MAX 6 bits. The maximum number x of thresholds that can be realized is
256 / {x * (8-1) +1} ≧ 1 Therefore, x ≦ 36
Therefore, it is based on the fact that the reproduction of pseudo gradation processing up to 256 gradations necessary and sufficient by multi-value dither processing can be covered with MAX 6 bits. The basic hardware configuration can be easily realized with a RAM or the like.
[0049]
Reference numeral 54 denotes a LUT (Look Up Table) section, which is also composed of a RAM or the like. The conversion result by the actual multi-value dither processing based on the encoded MAX 6-bit data and 8-bit, 256-gradation input image data is 3 bits. , Output in 8 gradations.
[0050]
The pseudo gradation processing unit 24 having such a configuration can represent input image data of 8 bits and 256 gradations per pixel, and can represent pseudo gradations of 3 bits per pixel, 8 gradations and 256 gradations by multi-value dither processing. It becomes. Further, when the encoding unit 53 and the LUT unit 54 are constituted by a RAM or the like, the dither reference threshold array shown in FIG. 6 or the plane shown in FIG. Multi-value dither processing that can be changed to any sequence is possible by calculating based on the sequence of multi-value threshold arrays that are in-between and initially loading the tabulated data via each selector 55, 56, 57 Become.
[0051]
Next, a specific configuration of an algorithm for multi-value dither processing will be described.
6 and 7 show configuration examples of sequence algorithms for multi-value dither processing. In order to simplify the description, the dither threshold value array having a very small size will be described.
[0052]
FIG. 6 shows a reference threshold array, which is a screw-type dither matrix having a screen angle of 45 degrees. In this case, the pseudo gradation reproduction number of 8 values per pixel is 8 × (8-1) + 1 = 57 gradations, and the number of gradations is small from the original. This will be described in the key configuration. Even if the number of gradations increases, the basic processing configuration does not change.
[0053]
In the case of the reference threshold array of FIG. 6, the number of bits of the main counter 51 and the sub-counter 52 in FIG. 5 are both 2 bits, and the LUT unit 54 uses the 3 bit data encoded by the encoding unit 53 and the input image data. Value dither processing is performed and output as 3-bit image data.
[0054]
FIG. 7A, FIG. 7B, and FIG. 7C show the threshold array sequence in the depth direction, that is, the pixel level direction when FIG. 6 is used as the reference threshold array. This threshold value is not normalized by 0 to 255 and is represented by a simple serial number of large and small. In FIG. 7, the horizontal axis item represents the reference threshold value, and the vertical axis item represents the level number of the multi-value plane.
[0055]
First, the threshold arrangement sequence in FIG. 7A has the same threshold arrangement as in FIG. 38A and is an ideal threshold arrangement, but it is easily affected by printing accuracy and has density unevenness and vertical stripes. Problems occur. Further, the threshold sequence in FIG. 7B has the same threshold array configuration as in FIG. 38B, and the problem of density unevenness and vertical stripes resulting from engine accuracy is less noticeable, but the resolution is lowered. Occurs. In addition, the threshold sequence in FIG. 7C is a threshold sequence configuration example showing the intermediate characteristics. FIG. 8 shows a print example by the above-described three types of multi-value dither processing when the reference threshold value array of FIG. 8A shows the printing result according to FIG. 7A, FIG. 8B shows the printing result according to FIG. 7B, and FIG. 8C shows the printing result shown in FIG. It is a printing result by (c).
[0056]
In the configuration of the multi-value dither processing as described above, a technique for realizing the optimum image reproducibility by combining these with respect to both sides of the above-mentioned dither reference threshold value array and the sequence between the multi-value dither threshold value planes will be described below. . These two basic configurations have some correlation with each other as shown in FIG. 7 from the viewpoint of image quality.
[0057]
As another feature of the multi-value dither processing, the sequence of FIG. 7A will be described as an example for simplification of description.
As shown in FIG. 9, in the output of the multi-value dither processing, as in the case of the output of the binary dither processing, the pixel to be printed is in the on or off (solid white) state in terms of the first threshold plane. Only the low gradation part that is the object, that is, the highlight part. All threshold comparisons in the first threshold plane are turned on. In a higher gradation part, all pixels are filled with dots of some size. If anything, it is an AM modulation type that has a very high spatial frequency. Output characteristics. In this case, each pixel corresponds to one halftone dot, and the image is such that each halftone dot gradually grows. Note that the number of reproduction levels of the halftone dots themselves is small.
[0058]
It has been found that such a dot formation process can provide a much higher quality image, especially in terms of graininess, than a dot reproduction method that takes a binary on or off state at the same resolution. . Further, in a low gradation part which is processed in the first threshold plane reproduced in the same FM modulation as the binary dither output, the paper with respect to the resolution pitch of the output device compared to a normal binary printer. Since the dots formed on the top are very small, an image with good graininess can be obtained.
[0059]
In the case of multilevel dither processing, the number of gradations that must be reproduced for each threshold plane can be simply divided into the number of planes by taking the above sequence as an example. In the case of processing, 256 / (8-1) ≈36 gradations are sufficient, and this gradation reproduction is simply repeated for 7 threshold planes.
[0060]
Therefore, since it is sufficient to optimize the pattern design for only 36 gradations, there is no periodicity for all 256 gradations as in the case of binary values, and furthermore, compared with the threshold design that must prevent textures from being generated. Optimization is relatively easy.
[0061]
As in this example, in the case of a printer capable of forming multi-valued pixels, the minimum drop, that is, the dot size from the first basic gradation dot to the number of basic gradation dots, is not shown in the resolution pitch of the printer. As shown also in FIG. 4, since it is smaller, adjacent dots do not contact each other. Note that the minimum drop dot size varies depending on the paper and printing accuracy.
[0062]
In such a case, it is possible to design a pattern in which each dot is dispersed as much as possible, which is visually preferable. Further, at this time, it is preferable to design the dither threshold value array so that the repetition period is not visually visible.
[0063]
In the case of a printer capable of forming such a multi-value dither image, a dither threshold array in which dots are dispersed as much as possible can be designed. On the other hand, in an actual printer, on a two-dimensional plane in the main / sub-scanning direction. It is rare that the physical accuracy is the same in both scanning directions, and usually the accuracy of either one is lowered depending on the architecture of the printer. In the case of an ink jet printer, the accuracy decreases in the main scanning direction due to variations in the volume and direction of ink ejected from an ink ejection port that is a recording element.
[0064]
At this time, as described above, in the dither processing that reproduces dots that are isotropically distributed in all directions as much as possible, the equivalent processing is performed even if the printing accuracy is biased. Is not substantially done. Unnecessary noise frequency components due to density unevenness and streaks on actual printing are not successfully canceled out. However, in the case of a low gradation part where adjacent dots are separated from the basic resolution pitch, this density unevenness and streaks are relatively inconspicuous to the eye, and just from the middle to the upper floor where the adjacent dots touch or do not touch each other. It becomes most noticeable when the dot size is the key.
[0065]
Further, in such a multi-value printer, although it depends on the basic gradation characteristics, in most cases, in the low gradation part, the fine dots are more locally than the aperiodic randomly distributed dot pattern. A dot pattern with a regularly and regularly distributed systematic dither produces a visually pleasing smooth output. However, since human visual characteristics show strong sensitivity in the horizontal and vertical directions, higher image quality can be obtained when adjacent dots are arranged in an oblique direction.
[0066]
This is shown in FIGS. 10 (a) and 10 (b). (b) is a pattern that is regularly and uniformly arranged, and this pattern looks visually smoother than the randomly dispersed pattern as shown in (a). However, the output pattern of the systematic dither of the distributed system is different when the angle changes in switching between adjacent gradations as shown in FIGS. 11A and 11B, or as shown in FIG. Since there are gradations that give rise to a pattern of dots, dots should be small enough with respect to the pixel pitch, and the systematic dither pattern should be kept only in low gradation areas where such gradations are not visually noticeable. is important.
[0067]
Further, when the maximum seventh basic gradation dot formed on the paper is set to a size that completely covers at least the square pixel of the basic resolution, the other basic gradation dot characteristics are generally as shown in FIG. . Note that FIG. 13 shows the measurement of each density when the same size dots are printed on one side for each basic gradation on an appropriate sheet. From this figure, it can be seen that the difference between the 0th basic gradation density, that is, the first basic gradation density that is completely filled with the first drop from the solid white density of the paper is the density between the other adjacent basic gradations. That is greater than the difference. Therefore, the simple multi-value dither threshold sequence in the reproduction of the low gradation part, which is very important for gradation reproduction, has a low gradation resolution in the low gradation part, a large density change between gradations, and a gradation jump. There is also the possibility that it will be easily noticeable.
[0068]
In consideration of the above points, using FIG. 14, input gradation image data of 8 bits, 256 gradations (0: white, 255: black) is subjected to pseudo halftone processing, and each color has 3 bits, 8 gradations (0: white, 7: Black) will be specifically described. A2 in FIG. 14 indicates a reference threshold value array, and the matrix size is 30 × 30.
[0069]
Here, the maximum number x of different thresholds capable of realizing the maximum number of reproduction gradations not exceeding 256 gradations by multi-value dither processing is:
256 / {x * (8-1) +1} ≧ 1 Therefore, x ≦ 36
It is. In other words, the number of gradations handled by each threshold plane in the threshold array having 7 planes is 36 gradations. That is, there are 36 gradation output patterns in one plane. Incidentally, at this time, the gradation of 36 * 7 + 1 = 253 gradations can be reproduced with all the threshold planes.
[0070]
When the 36 gradations are formed by a matrix of the minimum unit, the gray scale is 6 × 6, and the gradations for 36 gradations can be reproduced by turning on any one pixel in each 6 × 6 threshold matrix one by one. Here, as shown in FIG. 14, if the 6 × 6 threshold array A1 is the minimum dither unit, the 30 × 30 all threshold array A2 has five minimum threshold arrays A1 in the main scanning direction and in the sub scanning direction. This is the size that can fit a total of 25 25 pieces. If the matrix size is set to an integer multiple of the minimum dither unit in this way, it is convenient that the joint of the iterative processing by the systematic dither can be smoothly shifted.
[0071]
Next, the arrangement of reference threshold values will be described with reference to the numerical values shown in each threshold value matrix of FIG. Note that the blank portion of the threshold matrix means that a numerical value of 5 or more is filled. In the dithering process, the output pixel is turned on as the input gradation increases in the order from the smallest threshold.
[0072]
In the low gradation part, as shown in FIG. 14, each threshold value is sequentially turned on from 1 to 4. However, as can be seen from this threshold arrangement, the low gradation part is locally periodic ( 6 × 6 unit matrix) Dither threshold value array. Further, the threshold arrangement of the periodic systematic dither is devised so as not to be arranged horizontally or vertically between adjacent pixels. This realizes smooth gradation reproduction that is periodic and inconspicuous in the low gradation part. In addition, the interval between adjacent pixels in this periodic dither arrangement is experimentally larger than two pixels. That is, every other horizontal or vertical direction is not arranged. In this way, the graininess of the low gradation part is not lowered.
[0073]
Next, the process of filling the threshold value in the blank portion is performed. Basically, if one pixel of each minimum dither unit is turned on for each input gradation, all 36 gradations are reproduced. It can be done.
From the threshold arrangement of the periodic systematic dither in the low gradation part to the intermediate gradation part to the high gradation part (in this example, the range of the threshold value 5 to 36) between the adjacent minimum dither blocks. The threshold value configuration is realized so that the output pattern is locally aperiodic.
[0074]
The simplest method for realizing this is a method of determining a range of thresholds 5 to 36 which is a remaining unthresholded portion by using a random number. That is, the next gradation value portion is selected at random for each minimum dither unit, and a threshold value is assigned to the selected portion in ascending order. As a result, a threshold value of 1 to 36 can be allocated over the entire matrix size.
[0075]
In general, it has been found that the threshold pattern determined by random numbers is noisy due to the obstacles caused by forming a continuous surface that is visually uncomfortable when a uniform gradation surface is processed. The system uses a threshold arrangement that is the most uniformly distributed in the tones, and in addition, in a multi-level printer, adjacent dots do not touch each other with a low basic tone dot, which is more visually apparent than in the binary case. Unpleasant black blocks are difficult to recognize. Therefore, even if the remaining threshold value is generated by random numbers, no visually unpleasant texture is generated.
[0076]
On the other hand, a more preferable method for obtaining the threshold value is a method of calculating a portion having the best dispersibility in each basic dither unit by convolution filter processing while referring to the peripheral minimum dither unit. This process is shown in the flowchart of FIG.
[0077]
First, in step S1, it is assumed that the threshold portion indicated by the numerical values 1 to 4 in FIG. 14 is turned on, and the size of 30 × 30, where 1 is the on portion and 0 is the remaining portion. The pattern is the initial pattern. Next, in step S2, a convolution filter process is performed on the initial pattern, and the least sparse part in the pattern at the position where the value is 0, that is, the minimum value is obtained as a result of the filter operation. The part to take is detected. At this time, it was found that an excellent output pattern can be obtained by using a filter having the following shape as an example of a suitable filter.
[0078]
[Expression 1]

[0079]
Here, i is a convolution variable in the main scanning direction, j is a convolution variable in the sub-scanning direction, and ki, kj, and n are arbitrary constants.
[0080]
The optimum values of ki and kj are determined by the dot diameter (minimum dot diameter) and the pitch interval to be actually printed, and the exponent n is determined by the dot shape, particularly the dot edge shape. Yes. This equation is an approximate calculation model in which optical characteristics of ink dots printed on paper are patterned.
[0081]
At this time, it is assumed that there are a plurality of positions having the same value. This is likely to occur because the initial pattern is a periodic systematic dither, but a suitable result can be obtained even if the position selected at random is selected at random at this time.
[0082]
Subsequently, in step S3, the rank is stored for the pixel at the detected position, and a pattern in which the bit at the position is changed from 0 to 1 is generated. In this case, the target is not the minimum dither cycle unit, but a series of ranks in all matrix sizes is determined.
[0083]
This is repeated until there is no 0 bit pattern, and the priority order of all 30 × 30 pixels is determined. Note that priorities are assigned in the order of 101 to 900 in the calculation process if the first order is assigned in advance to the portion (25 × 4 = 100) that is already reproduced in a periodic pattern. Is assigned to a threshold of 5 to 36.
[0084]
A threshold is assigned according to this priority. In the case of a matrix of size 30 × 30, the allocation is such that the threshold value is increased by 25 pixels by one. As a result, the positions of the remaining pixels having threshold values of 5 to 36 are determined, and finally all threshold values in the 30 × 30 matrix are filled. This is a 30 × 30 size dither reference threshold array.
[0085]
At this time, the weight of the filter operation is relatively changed between the main scanning direction and the sub-scanning direction so that the threshold value is generated so that dots are continued in a direction with relatively low printing accuracy. That is, weights are given to the values of ki and kj in the above equation (1). Specifically, by setting ki <kj, it is possible to generate a pattern that can be easily connected in the main scanning direction with low printing accuracy. It should be noted that the strength of this connection is desirably set optimally according to the printing accuracy as shown in FIG. 16 by changing the ratio of ki and kj.
[0086]
As a result, in the case of a multi-value printer, the reference threshold value array generated anisotropically does not have a large correction effect on the printing accuracy by itself. It is possible to greatly reduce printing errors such as unevenness and streaks. Note that when binary dither processing is performed with a binary printer, dots that are larger than the resolution pitch from the beginning are connected between adjacent pixels, and this is strong against density unevenness and streaks. Therefore, only this reference threshold value array has an effect.
[0087]
In addition, it turned out that about 1/10 of the number of pixels in a basic dither unit is good for switching between the threshold arrangement by the periodic systematic dither and the local aperiodic dither threshold arrangement. . This is because if too many pixels are fixed at the threshold value of systematic dither, the degree of freedom in the vacant area will be significantly reduced, and an unnatural texture or the like will occur at a specific gradation. This is because there is a possibility of end. Experimentally, the range constituted by the periodic systematic dither has a good result of about 0 to 20% of the reference threshold range.
[0088]
Here, the case of the upper limit of 20% is a pattern that is configured by the pattern shown in FIG. 17 (ON320 / all pixels 1600 = 20%) and has a periodicity. This 20% is the limit for giving the remaining dot arrangements freedom. If more than this part is specified periodically, the texture will be remarkably generated.
[0089]
Moreover, although it is a matrix size, if this size is too small, periodicity or unnecessary textures can be seen, so an appropriate size that is not redundant is necessary. This optimum size varies depending on the basic characteristics of each dot and compatibility with the paper. If the dot design of each basic gradation pixel does not show an extremely large non-linearity, the input gradation image data of 1 pixel M gradation is subjected to multi-value dither processing to 1 pixel N (M >N> 2) When converting to image data having a smaller number of gradations, the matrix size of the multi-value dither processing is K × L, and the number of output gradations after the multi-value conversion is N gradations. If you leave
[0090]
[Expression 2]

[0091]
K and L are set to be an integer in the range of, and especially in the case of a square matrix (L × L),
[0092]
[Equation 3]

[0093]
If it is set so as to be an integer in the range, the occurrence of periodicity and texture can be suppressed. The minimum limit value indicates a visual tolerance limit, and the maximum limit value indicates a limit that does not result in a redundant size. The matrix size derived from this is a small size that can no longer be referred to as a large-scale matrix size with respect to a preferred size of 128 × 128 or more, which is generally referred to as binary stochastic dither, and has a smaller hardware configuration. Can be realized.
[0094]
Furthermore, in consideration of taking an integer value that is divisible by the basic period of systematic dither, a non-square matrix such as 30 × 24 as shown on the right side of FIG. .
[0095]
Although an example in which the minimum dither unit is a square matrix has been described here, the minimum dither unit may be a square matrix as shown in FIG. In FIG. 18, 8-bit, 256-gradation (0: white, 255: black) input gradation image data is subjected to pseudo halftone processing to convert each color into 2-bit, 4-gradation (0: white, 3: black). A case will be described as an example.
[0096]
In this case, the maximum number x of different thresholds capable of realizing the maximum number of reproduction gradations that does not exceed 256 gradations by multi-value dither processing is:
256 / {x * (4-1) +1} ≧ 1 Therefore, x ≦ 85
It is. In other words, the number of gradations assigned to each threshold plane in the threshold array having three planes is 85 gradations. It should be noted that 85 gradations are not necessarily required, and each threshold plane is in charge of output patterns for 80 gradations this time in order to make the explanation easy to understand. Incidentally, at this time, 80 * 3 + 1 = 241 gradations can be reproduced in all threshold planes.
[0097]
When the 80 gradations are constituted by a matrix of the minimum unit, the size is 10 × 8, and the gradations for 80 gradations can be reproduced by turning on arbitrary pixels of the 10 × 8 threshold matrix one by one. Here, as shown in FIG. 18, if the 10 × 8 threshold array B1 is the minimum dither unit, the 40 × 40 total threshold array B2 has four minimum threshold arrays B1 in the main scanning direction and in the sub scanning direction. It is a size that can fit a total of 20 of five.
[0098]
If the matrix size is set to an integer multiple of the minimum dither unit in this way, it is convenient that the joint of the iterative processing by the systematic dither can be smoothly shifted. In FIG. 18, the minimum dither unit is configured with 10 pixels in the main scanning direction and 8 pixels in the sub-scanning direction. However, this allocation method may be changed in the main / sub-scanning direction.
[0099]
If the dither matrix size also meets the following conditions:
[0100]
[Expression 4]

[0101]
Thus, for example, a non-square matrix size such as 40 × 48 may be used.
[0102]
In the examples of FIGS. 14 and 18, the dither matrix having the minimum number of pixels is configured to have an oblique component compulsorily. If the above-described method is used, for example, as shown in FIG. In addition, with respect to a screen dither matrix having an appropriate number of gradations, an appropriate angle, and an appropriate matrix size, the threshold values in the entire matrix can be similarly generated using only the pattern of the low gradation part. In this case as well, the conditions of the two matrix sizes are naturally satisfied.
[0103]
In addition, the total number of gradations to be reproduced need not always be 256 gradations, and even when a screen dither matrix is used, a unique gradation reproduction number is determined. It only needs to be reproduced.
Thus, in the low gradation part, the reference threshold value array is calculated using a pattern in accordance with the periodic systematic dither rule.
[0104]
Next, another preferred example will be described. In the low gradation portion, a dither reference threshold value array is set so that a pattern having a random output characteristic that is locally aperiodic is set, and the intermediate gradation portion From the high gradation portion to the high gradation portion, the dither reference threshold value array is set so as to obtain an anisotropic output pattern in which dots are preferentially connected in the scanning direction with relatively low printing accuracy.
[0105]
This is true in the case where the output is more beautiful in terms of visual characteristics when the dots are reproduced locally and non-periodically than when the dots are reproduced periodically in the low gradation part. Also in this case, the range of 0 to 20% of the reference threshold value array is a range showing characteristics different from the output characteristics from the intermediate gradation portion to the high gradation portion depending on the printing accuracy. In other words, density unevenness and streaks are always easy to see from the intermediate to high gradation parts regardless of the printing accuracy, but in the low gradation parts, it is likely that density unevenness and streaks are relatively difficult to see due to the printing accuracy. There is a characteristic.
[0106]
Here, even in the low gradation part where density unevenness and streaks due to fluctuations in printing accuracy are relatively inconspicuous, the threshold value of the low gradation part is set as shown below depending on the appearance of the output printed by the printer. You may switch.
[0107]
That is, the threshold setting in the case of a printer characteristic where printing unevenness and streaks are not conspicuous in the low density portion is completely isotropic instead of the pattern shown in FIG. An output pattern that generates a uniform number of output dots in each row / column in the sub-scanning direction may be obtained and used as an initial pattern, or any uniform gradation in a low gradation area may be optimized. A pattern obtained by processing with a diffusion algorithm may be used as the initial pattern.
[0108]
Then, using this initial pattern, a threshold value for a higher gradation part is generated at random, or a convolution filter is used to anisotropically connect dots preferentially in a scanning direction with relatively low printing accuracy. A threshold value that is an output pattern is obtained. Similarly, it is desirable that the connection strength from the intermediate gradation portion to the high gradation portion is optimally set according to the printing accuracy. In addition, since the order of the threshold value allocation on the low gradation part side is not determined from the beginning like the pattern of systematic dither, the order of the matrix in each low gradation part is stochastically statistically determined in advance. A threshold value for obtaining a uniform number of output dots in each row / column may be obtained, or may be calculated by applying a threshold calculation method on the high gradation side to the low gradation side. Since the matrix size is relatively small and the number of gradations on the low gradation side is known at most, it is easy to perform manual optimization.
[0109]
Also, in the case of printer characteristics where density unevenness and streaks are actually noticeable even in the low gradation part, the dots are preferentially connected in the scanning direction where the printing accuracy is relatively low, starting from the low gradation part. A dither reference threshold array is set so as to have an anisotropic output pattern. In this case, instead of the pattern shown in FIG. 14, the threshold value setting is a completely anisotropic pattern, and an output pattern that forcibly generates output dots that are continuous in the main scanning direction of the matrix is obtained. Or an output pattern processed by error diffusion in which the coefficient of the error diffusion matrix is optimized so that the output is continuous in an arbitrary uniform gradation in the main scanning direction may be used as the initial pattern. Then, using this initial pattern, a threshold value for a higher gradation part is obtained in the same manner as in the above method. It is also possible to use a threshold arrangement of systematic dither that shows characteristics linked to the main scanning direction in the low gradation part.
[0110]
As a result, optimal output can be obtained in the entire gradation range even when dots are reproduced locally aperiodically rather than periodically reproducing dots in the low gradation part. The reference dither threshold that is to be obtained is determined. Further, it is possible to obtain a reference dither threshold value that becomes an anisotropic pattern in order to make printing unevenness and streaks inconspicuous from the intermediate gradation portion to the high gradation portion.
[0111]
As described above, the reference threshold value array generated by this method does not have a large correction effect on the printing accuracy alone in the multi-value output device. By combining them, it is possible to greatly reduce printing errors such as density unevenness and streaks.
[0112]
Although the reference threshold value array has been described above, a method for developing the reference threshold value array obtained above in the direction of each multilevel plane will be described next.
As described above, the dot output characteristics of an output device that reproduces gradation by area modulation differ greatly depending on the sequence of multi-value dither processing.
In the above two basic configurations (sequence between the reference threshold value array and the threshold plane), the image quality improvement effect by changing the sequence between multiple multi-value dither threshold planes in terms of compensation for print accuracy such as print unevenness and streaks. Is bigger.
[0113]
Unlike the case where the threshold sequence is limited depending on the architecture of the printer, in the printer as in the present embodiment, the sequence between the threshold planes can be changed relatively easily. However, basically changing the sequence between the threshold planes has the effect of relatively easily suppressing printing unevenness and streaks resulting from printing accuracy, but it greatly affects the resolution and gradation reproducibility. Therefore, it must be carefully designed.
[0114]
Further, here, two optimization techniques for the threshold sequence are incorporated.
First, the first optimization will be described with reference to FIG. FIG. 13 is a general output density characteristic of an output device that reproduces an image by area modulation, as described above, and the reproduction resolution of the low gradation part is lower than that of the intermediate gradation part to the high gradation part. This is because in the low gradation part, only the first basic gradation is used rather than sequentially arranging dots of the same size of one basic gradation on the white background part of the paper in accordance with the reference threshold value arrangement every time the gradation is raised by one step. Rather, it means that when the dots after the second basic gradation are appropriately interlaced and printed, it is easier to obtain a smooth density change between adjacent gradations in terms of gradation reproduction. become.
[0115]
This is shown in FIG. FIG. 20 (a) shows the same dot growth process as in FIG. 7 (a) in the sequence for the highest resolution. On the other hand, as shown in FIG. 20 (b), it is theoretically possible to obtain an output with the same density not only by the configuration of the minimum dots but also by interweaving dots of different sizes. Here, considering the output characteristics of FIG. 13, the change in density is that the growth process in FIG. 20B may provide a smooth change in density between adjacent gradations. . However, in this case, since a dot larger than the basic one-tone dot is output to the low-gradation portion, it is assumed that the unpleasant pattern of the size is not visually visible.
[0116]
Here, as a preferable method for obtaining the optimum dot output pattern, instead of the output pattern in which only the dot of the basic one gradation showing the growth as shown in FIG. As shown in (b), only the portion up to the gradation indicating the output pattern of the periodic systematic dither, that is, before arranging the dots of the first basic gradation throughout, several basic gradations first. It is to grow dots. If the basic gradation is such that the dots are not visually recognizable, the gradation appears to be much smoother than the output pattern shown in FIG.
[0117]
FIG. 22 shows a threshold sequence according to the output pattern shown in FIG. In this example, the reference threshold value indicating the output of periodic systematic dither is up to 4, and dots of up to three basic gradations are preferentially output at this corresponding position. In addition to the effect of making the output pattern itself suitable for vision, this processing is also effective in that an output pattern that is strong in printing accuracy and the like can be obtained because adjacent dots are separated. . This effective range is more effective when the value is in the range of approximately 0 to 10% of the gradation range that can be taken by all input image data.
[0118]
Here, 10% of the upper limit of the gradation range that can be taken by the input image data refers to a pattern having an upper limit of 20% within the prescribed threshold range of the dither reference threshold array for realizing a periodic systematic dither array. This is a range in which output is preferentially repeated to half of the number.
[0119]
In the above description, the case where dots up to three basic gradations are preferentially output with respect to a periodic output pattern has been described. However, in practice, the setting of this sequence largely depends on the dot diameter of the basic gradation characteristics. Is done. That is, although the basic dot size is determined by the pure resolution of the printer, for example, the basic minimum dot size and pitch are different at 300 dpi / 600 dpi even with the same number of output gradations. In addition, if the resolution is increased, the characteristics of the dots actually measured with respect to the ideal dot diameter become non-linear because of the substantial design difficulty.
[0120]
Accordingly, naturally, the setting of the sequence is arbitrarily optimized according to the above-described rules by the different basic gradation characteristics (particularly the dot diameter) due to these various factors, as shown in FIGS. Also, the setting of this sequence is arbitrarily optimized as shown in FIGS. 23 (a) and 23 (b) in accordance with the rules described above according to the printing accuracy.
[0121]
On the other hand, in another preferred example relating to the sequence, for example, when dots are uniformly allocated in the first basic gradation, the dot size of the first basic gradation is extremely small and has good characteristics. In the case where an output that is more visually pleasing than a systematic dither arrangement can be obtained, it is empirically determined that an image reproduced at 0 to 20% on the low gradation side of the input image has an adjacent pixel pitch interval. On the other hand, a dither threshold value array is given so as to increase the spatial frequency for an input image in this range by using the fact that the size of the configured pixel is small and density unevenness and vertical stripes are not conspicuous. . An example is shown in FIG.
[0122]
As a result, the types of dot sizes that appear in the multivalued image data converted when the input image data is in the low gradation part are substantially smaller, and the input image data is in the range of data from the intermediate gradation part to the high gradation part. In some cases, the number of types of dot patterns that appear by the converted multi-value image data is substantially larger than that of the low gradation portion.
[0123]
This makes the pixels in the low gradation area inconspicuous in the gradation reproduction of the printer inconspicuous, improves the gradation reproduction, and distributes the dot types in areas where density unevenness and streaks are conspicuous. Therefore, uneven density and streaks can be made inconspicuous. Also, unlike the case where thresholds are randomly arranged, there is a correlation between the threshold planes, so that the thresholds of each plane can be automatically obtained from the basic dither matrix, and simplification of hardware can be expected.
[0124]
Furthermore, due to different basic gradation characteristics (particularly dot diameter) due to various factors, naturally the setting of the sequence is arbitrarily optimized according to the rules described above, as shown in FIGS. Further, the setting of this sequence is arbitrarily optimized as shown in (a) and (b) of FIG. It should be noted that the example relating to the threshold sequence is further effective when combined with the optimization of the reference threshold sequence described above.
[0125]
FIG. 26 (a) shows the on / off characteristics of the threshold value in the middle gradation portion when focusing only on the normal locally non-periodic and uniformly distributed reference threshold value array. FIG. 26B shows a reference threshold value generated in this embodiment so as to be a pattern showing an anisotropic output characteristic in which dots are preferentially arranged in a scanning direction with relatively low printing accuracy. The threshold ON / OFF characteristics in the middle gradation portion according to the arrangement are shown.
[0126]
On the other hand, each pattern in FIG. 27 is a diagram schematically showing an output pattern when actually printed on a sheet by various multi-value dither processing, and is indicated by a dotted line CC in the figure. Are shifted to the right due to the influence of misdirection of the ink head, for example. Note that (a) to (c) in FIG. 27 correspond to the sequences (a) to (c) in FIG. 7, respectively. FIG. 27C shows an isotropic rule as shown in FIG. 26A in terms of the reference threshold value array.
[0127]
On the other hand, FIG. 27D is a diagram showing the state of the output pattern according to this embodiment, and the reference threshold value array corresponds to the anisotropic one shown in FIG. In the case of FIG. 27D, the reference threshold value arrays are preferentially connected in the horizontal direction. That is, the threshold value between adjacent pixels in the horizontal direction is relatively easy to take a neighboring value, and the dot is likely to grow preferentially in the horizontal direction.
[0128]
As a result, the correction effect can be expected as it is in FIG. 27C, but the reference threshold value array and the sequence between the threshold planes can be obtained as shown in FIG. 27D. As a result, even when the printing position accuracy is low, the effect of making density unevenness and streaks more inconspicuous is exhibited.
[0129]
It should be noted that here, a threshold array that spans a plurality of dither threshold planes is a threshold array that is locally periodic when the input image data is in a low gradation region, and a regular output pattern is preferentially output. When the input image data is in a region where the input image data has a high gradation from the intermediate gradation, the type of the dot pattern that appears when the input image data has a low gradation is set. The individual cases where the dot pattern is set to be larger than the types of dot patterns that appear when it is in the region have been described, but the present invention is not limited to this, and it extends over a plurality of dither threshold planes set in both threshold arrays. A threshold value array may be prepared in advance, and which threshold value array is used may be selected according to the output accuracy of the printer.
[0130]
Next, a case where gamma correction is incorporated in the multi-value dither threshold array will be described. In general, the number of theoretical gradations that can be reproduced by pseudo halftone processing is determined by the total number of different threshold values in the unit matrix. The screen dither can reproduce the number of gradations determined by the combination of pattern size, angle, etc., and the stochastic dither and error diffusion can reproduce the same 256 gradations as the number of input gradations. It is.
[0131]
However, the number of theoretical gradations is a case where only the pseudo halftone processing unit is assumed. Actually, gradation loss may occur in all the image processing units before the pseudo halftone processing, so the image data finally input to the pseudo halftone processing unit has only a limited number of gradations, and the pseudo halftone processing. There are output patterns that are not used at all in the processing unit.
[0132]
In the flow of normal image processing, the order is color conversion → BG / UCR → gamma correction → pseudo halftone. In each image processing unit, a rounding error due to digital arithmetic processing or gradation loss due to color gamut compression or the like occurs.
Here, since the gradation loss in the color conversion unit and the UCR unit cannot basically be restored, the gradation reproducibility in the gamma correction unit and the pseudo halftone unit will be described.
[0133]
The gamma correction processing is processing for correcting the basic gradation characteristics of the engine to target characteristics such as luminance linearity and density linearity, and has a relationship shown in FIG. In other words, by converting the input image data from the measured engine basic tone characteristics to the target characteristics using the target gamma correction curve, the final output tone characteristics are matched with the target characteristics. It is. In the figure, the target curve is a straight line, but can be replaced with an arbitrary curve.
[0134]
In general, the basic characteristic of an area modulation output device that expresses dots with circles is a characteristic that is higher than the straight line of the graph g in FIG. The actual processing of γ correction is often performed by digital 1LUT calculation for each color.
[0135]
In FIG. 28, in the calculation by the digital 1LUT, a specific conversion as shown in FIG. 29 is performed. In the low gradation part, the data is repeatedly converted into the same data by a digital rounding error, and in the high gradation part, it is converted into a jump value. In other words, in the reproduction of the low gradation part that is important for gradation reproduction, the output halftone pattern is likely to be the same pattern even in different input images, and there are halftone patterns that are not used in the high gradation part, As a whole, the number of gradation reproductions is reduced, and the efficiency of image processing is extremely poor. In this phenomenon, as the basic characteristics of the engine are further away from the target characteristics, the digital conversion accuracy is lowered, and the number of gradation reproductions is greatly reduced. However, when it comes to ink jet printers, it has relatively close characteristics.
[0136]
Therefore, here, a gamma correction is incorporated in the halftone process to generate a matrix that theoretically suppresses gradation loss. Here, with regard to binary dither processing, a dither threshold generation method incorporating gamma correction is well known, but in the case of multi-value dither processing, the basic tone characteristics between the planes are not linear. When all the sequences between the threshold planes are supported, the conventional method is difficult to apply.
[0137]
For example, when one pixel has 8 values and the matrix size is 32 × 32, the number of dots that are simultaneously turned ON / OFF for a given gradation is 32 × 32 × (8-1) / 255≈28. In the normal pseudo halftone processing, this number is assigned equally to each gradation.
[0138]
The basic principle of this process is that, in this halftone process, a gamma conversion characteristic is incorporated in the threshold value between all threshold planes for determining ON / OFF of pixels, and the number of pixels to be turned ON in each threshold process of all threshold planes is determined by gamma. Control according to the characteristics. In other words, the actual number of gradations is restored by implementing the processing only by adjusting the total number of ON dots in the pseudo halftone part through the digital conversion gamma correction part that causes gradation loss. The number of ONs in this case is, in the case of multi-values, there are a maximum of 7 ON numbers per pixel depending on the pixel level direction, that is, the number of thresholds among the 7 threshold levels. Show.
[0139]
The flow chart of FIG. 30 shows how to calculate the multi-value dither threshold value array incorporating this gamma correction. First, in step S11, a substantial engine gradation characteristic is obtained by multi-value dither processing using a multi-value dither matrix in which the number of dots is equally allocated in advance. Next, γ target is determined in step S12, and the gradation is set to 0 in step S13.
[0140]
Subsequently, in step S14, a gamma correction gradation value to be converted in accordance with the target characteristic is calculated for each input gradation value in the same manner as the normal gamma conversion in accordance with the determined target characteristic. At this time, if the gamma correction gradation value is calculated as a real number instead of an integer, the accuracy can be improved.
[0141]
Subsequently, in step S15, from the calculated output gamma correction gradation value, a value closest to this value is obtained from the curve of the output characteristics of the multi-value dither matrix in which the number of changing dots is equally allocated to each gradation. Convert to the number of ON dots at this time. The output curve is arbitrarily interpolated using the points actually measured. At this time, if the output gamma correction gradation value is calculated as a real number, a resolution of one dot unit can be obtained.
[0142]
Subsequently, in step S16, ON pixels are extracted from the threshold priority table, and in step S17, the threshold values of the multi-value dither are set in ascending order of priority among all threshold planes according to the number of dots to be turned on. In step S18, the gradation is incremented by one. Then, by repeating the processing of steps S14 to S18 over all gradations, all threshold values can be filled in all threshold planes.
[0143]
At this time, since all the priorities of the reference threshold value array of the multi-value dither have already been calculated, the priorities are set in advance between all the threshold planes along each sequence as shown in the example of FIG. Expand and obtain priorities among all final threshold planes. When a periodic pattern is output in the low gradation part, the priority order of the reference threshold value array is a priority order according to the rules of systematic dither, and the calculation for obtaining the reference threshold value array from the intermediate gradation to the high gradation is performed. Priorities in the reference threshold array have already been determined in the process.
[0144]
In the example of FIG. 31, the reference threshold is set to 4 and the threshold plane is set to 3 planes for the sake of simplicity. Actually, the priority order of the reference threshold is between 1 and 16. Here, looking at the sequence between the threshold planes in FIG. 31, the reference threshold value 1 and the first threshold plane part are first targeted, and the priority order is assigned to the part corresponding to this part. Next, the reference threshold 1 and the second threshold plane are targeted, and the next priority is assigned to the portion corresponding to this portion. By performing this operation from 1 to 12 in the entire order of the sequence, a priority order of 1 to 48 is allocated between all threshold planes.
[0145]
Then, by setting thresholds indicating the outputs corresponding to the number of corresponding gradation pixels that are turned on in accordance with the priority order, all complex sequences between threshold planes can be performed. A threshold is uniquely determined in the threshold plane.
[0146]
As described above, in the present embodiment, by combining the sequence between the reference threshold value array and the threshold plane optimally, one set of dither threshold value planes can be used for each level according to the print accuracy and the actual dot output characteristics. It is possible to perform multi-value dither processing with optimal output characteristics between the keys, and by incorporating gamma correction processing into the multi-value dither threshold itself, it is possible to obtain an image with higher gradation reproducibility. It became.
[0147]
In the above-described embodiment, the configuration of a single color is basically described, for example, in the case of black. However, it is possible to easily extend this to a color image as it is. However, it is necessary to pay some attention to the relationship between the output patterns between colors.
[0148]
In the case of a color image, multivalue dither processing is usually performed for each color. Here, the configuration of the multi-value dither processing having the reference threshold value array of the above-described embodiment is applied to at least two colors.
[0149]
Some colors may not use the configuration of the reference threshold array of the above-described embodiment. For example, for a color such as Yellow where dot particles are extremely inconspicuous, a simple conventional dither threshold array may be applied without using the configuration of the reference threshold array of the above-described embodiment. In the case of a color that wants to emphasize the edge more like the above, it is possible to apply a multi-value error diffusion process to perform a process that further enhances the edge effect.
[0150]
On the other hand, in the case of a color to which the processing of the above-described embodiment is applied, for example, if the same threshold pattern is applied to each color, a dot-on-dot output pattern is obtained, and the dot printing position is shifted due to a change in output characteristics. In this case, there is a problem that it becomes weak against color unevenness and the like, so it is necessary to make the reference threshold value array different for each color. In this case, the reference threshold value array may be individually created for each color. However, it is easier to create the reference threshold value array once created by an operation such as inversion, rotation, or shift.
[0151]
In general, in the case of a binary printer, a more severe design is required with respect to the color moire due to the correlation between the patterns of each color, but in the case of a multi-value printer, a combination of the patterns of each color itself. This is because the color moiré due to is less likely to occur, and a high-definition image can be expected to be obtained by a relatively simple threshold changing operation. In addition, since the threshold array is originally highly dispersed, the above threshold operation is sufficient. However, when the image is reproduced with periodic regular systematic dither in the low gradation part, the pattern corresponding to the gradation part has a very strong periodicity. Considering this, an appropriate threshold design is required.
[0152]
As a preferable example, the threshold value assignment in the minimum dither unit is rearranged by an operation such as inversion, rotation, or shift as shown in FIG. 32, or a completely different pattern is newly created. The threshold values in the entire matrix may be regenerated based on the created reference threshold pattern. At this time, as shown in FIG. 32, in the low gradation part between the respective colors, it is preferable that the dots are arranged so that the dots are not juxtaposed at the same position but are juxtaposed as much as possible. This is because the portion where the dots overlap is a secondary color in spite of the low gradation portion, and the presence of the dots is more visually noticeable. Furthermore, since this gradation portion has periodicity, this periodicity is also conspicuous.
[0153]
On the other hand, regarding the sequence between threshold planes relating to color, it is possible to optimally set the sequence depending on the actual printing accuracy of each color component. This is because each multi-value dither process for each color is processed independently, and this configuration can be easily realized.
[0154]
Further, it is known that, even if the statistical printing accuracy is the same, in general, the influence of density unevenness and vertical stripes on visual perception differs greatly depending on each color. For example, when the printing accuracy is the same, it is more noticeable as noise in order of Y → C → M → K. Therefore, in this multi-value dither process for each color, a more optimal output image can be obtained by performing pseudo gradation processing by appropriately changing the sequence between the threshold planes for each color.
[0155]
As mentioned above, although four colors of CMYK have been described here, it is needless to say that this is not limited to four colors, and can be easily realized by combining three colors of CMY or other colors. In this embodiment, multi-level dither processing has been described in general. However, most of the processing is also easy for binary dither processing except for processing limited to multi-level processing such as setting of a sequence between threshold planes. Is applicable.
[0156]
In this embodiment, the multi-value dither processing has been described. However, the present invention is not limited to this, and can be easily applied to a density pattern method or the like by those skilled in the art. In other words, in the multi-value dither processing, the input image data and the dither threshold have a 1: 1 correspondence, and the input image data and the finally output image data have a 1: 1 relationship, whereas in the density pattern method, The correspondence between the input image data and the conversion threshold is 1: K (K ≧ 2), and the input image data and the finally output image data only have a 1: K (K ≧ 2) relationship. Naturally, K may be extended in both or either of the main scanning / sub-scanning directions.
[0157]
In this embodiment, the case where a printer is used as the two-dimensional planar image output means has been described. However, the present invention is not limited to this, and a display such as a CRT display or a liquid crystal display is used as the two-dimensional planar image output means. May be.
[0158]
【The invention's effect】
According to the present invention, it is possible to perform pseudo gradation processing that improves graininess, is suitable for a photographic image, and has excellent gradation reproducibility.
[Brief description of the drawings]
FIG. 1 is a block diagram showing an overall hardware configuration according to an embodiment of the present invention.
FIG. 2 is a block diagram showing a configuration of an image processing unit in the embodiment.
FIG. 3 is a block diagram showing a configuration of a printer engine in the embodiment.
FIG. 4 is a diagram showing a pixel size of each gradation in the embodiment.
FIG. 5 is a block diagram showing a configuration of a pseudo gradation processing unit in the embodiment;
FIG. 6 is a diagram showing a dither reference threshold value array in the same embodiment.
FIG. 7 is a diagram showing an example of a sequence in a threshold plane in the embodiment.
FIG. 8 is a diagram showing an example of each output by the multi-value dither process of FIG.
FIG. 9 is a diagram showing an example of pixel growth in the sequence of FIG.
FIG. 10 is a diagram showing a dot pattern with a randomly distributed dot pattern and a regularly distributed systematic dither in a low gradation part.
FIG. 11 is a diagram illustrating an example of an angle change in an output pattern of a systematic dither in a distributed system.
FIG. 12 is a diagram showing an example of the appearance of a unique pattern in the output pattern of a distributed systematic dither.
FIG. 13 is a graph showing basic gradation characteristics of multi-value dither processing.
FIG. 14 is a diagram showing a reference threshold value array in the same embodiment.
FIG. 15 is a flowchart showing threshold generation processing in the embodiment;
FIG. 16 is a diagram showing dot reproduction at each basic dither threshold value in the embodiment.
FIG. 17 is a diagram showing a pattern in a case where the range constituted by periodic systematic dither is 20% of the reference threshold range.
FIG. 18 is a diagram showing another reference threshold arrangement in the same embodiment;
FIG. 19 is a diagram showing another reference threshold arrangement in the embodiment;
20 is a diagram showing an example of pixel growth in each sequence in the embodiment. FIG.
FIG. 21 is a diagram showing another example of pixel growth in each sequence in the embodiment.
FIG. 22 is a diagram showing an example of a sequence in the embodiment.
FIG. 23 is a diagram showing a modification example of the sequence shown in FIG. 22;
FIG. 24 is a diagram showing another example of the sequence in the embodiment.
FIG. 25 is a diagram showing another example of the sequence in the embodiment.
FIG. 26 is a diagram showing dot reproduction at a basic dither threshold value in the embodiment.
FIG. 27 is a diagram showing an output pattern in each sequence including printing unevenness.
FIG. 28 is a diagram for explaining a gamma characteristic and its correction.
FIG. 29 is a diagram showing gamma conversion by normal table conversion.
FIG. 30 is a flowchart showing processing for determining a dither threshold between all threshold planes incorporating gamma correction in the embodiment;
FIG. 31 is a diagram for explaining an operation for obtaining a priority order among all threshold planes in the embodiment;
FIG. 32 is a diagram showing the arrangement relationship of dots in a low gradation portion between colors in the embodiment.
FIG. 33 is a diagram illustrating a line recording head and a print example thereof.
FIG. 34 is a schematic diagram for explaining binary dither processing performed using a dither matrix;
FIG. 35 is a diagram showing an example of print output by the binary dither processing in FIG. 34;
FIG. 36 is a schematic diagram for explaining multi-value dither processing performed using a dither matrix.
FIG. 37 is a diagram showing an example of print output by the multi-value dither processing in FIG. 36;
FIG. 38 is a diagram showing a sequence in multi-value dither processing.
[Explanation of symbols]
24. Pseudo gradation processing unit
32-35 ... Inkjet head
51 ... Main counter
52 ... Sub counter
53 ... Encoding part
54 ... LUT section

Claims (25)

The input gradation image data of 1 pixel M gradation is dithered by the dither processing means using the reference threshold value array and converted to output image data of 1 pixel N (M> N ≧ 2) gradation, and then the two-dimensional plane. In an image processing apparatus that outputs an image by an image output means,
A two-dimensional planar image output means having different output accuracy in the main scanning direction and the sub-scanning direction as the two-dimensional planar image output means;
The dither processing means has a threshold arrangement characteristic that is locally periodic and regular in a region having a relatively low gradation within a prescribed threshold range, and has a regular threshold arrangement characteristic within the prescribed threshold range. It is set so as to have a locally non-periodic threshold arrangement characteristic in a region from a relatively intermediate gradation to a high gradation , and further, the locally non-periodic threshold arrangement characteristic is output to the two-dimensional planar image. An image processing apparatus having an anisotropic threshold arrangement characteristic in which dots are preferentially consecutively grown in a scanning direction where the output accuracy of the means is relatively low .   The dither processing means has a local periodic and regular threshold arrangement characteristic in the reference threshold arrangement, and each pixel of the image output by the two-dimensional plane image output means has at least 3 pixels in the main scanning direction and the sub scanning direction. The image processing apparatus according to claim 1, wherein the image processing apparatus is set so as to be spaced apart from each other.   The dither processing means is arranged such that each pixel of the image output by the two-dimensional planar image output means is spaced apart from a distance corresponding to two pixels by a locally periodic and regular threshold array characteristic in the reference threshold array. The image processing apparatus according to claim 1, wherein the image processing apparatus is set as described above. Dithering means, locally aperiodic threshold array characteristics in the reference threshold array, according to claim 1, wherein it has set derived from the approximate calculation model that mimics the output characteristics of the 2-dimensional plane image output means Image processing device. Dithering means, locally aperiodic threshold array characteristics in the reference threshold array, the image processing apparatus according to claim 1, wherein the set to derive randomly. The anisotropic threshold arrangement characteristic is large when the difference in output accuracy between the main scanning direction and the sub-scanning direction in the two-dimensional planar image output means is large, and is anisotropic when the difference in output accuracy is small. The image processing apparatus according to claim 1 , wherein the image processing apparatus is set so as to reduce the performance. The image processing according to any one of claims 1 to 6 , wherein the dither processing means performs dither processing using a plurality of threshold value arrays generated by correlating the reference threshold value array together with the reference threshold value array. apparatus.   2. The dither processing unit according to claim 1, wherein a region of 20% or less (however, 0 is not included) within a prescribed threshold range in the reference threshold array is a region having a relatively low gradation. Image processing device. The input gradation image data of 1 pixel M gradation is converted into output image data of K (K ≧ 2) pixel N (N ≧ 2) gradation using the reference threshold value array by the image conversion means, and then the two-dimensional planar image. In an image processing apparatus that outputs an image by an output means,
As the two-dimensional planar image output means, a two-dimensional planar image output means having different output accuracy in the main / sub-scanning direction is used.
The image conversion means has a reference threshold value array characteristic that is locally periodic and regular in a region having a relatively low gradation within a specified threshold range, and within the specified threshold range. It is set so as to have a locally non-periodic threshold arrangement characteristic in a region from a relatively intermediate gradation to a high gradation , and the local non-periodic threshold arrangement characteristic is output to the two-dimensional planar image. An image processing apparatus having an anisotropic threshold arrangement characteristic in which dots are sequentially grown so that dots are successively connected in a scanning direction in which the output accuracy of the means is relatively low . The image conversion means has a locally periodic and regular threshold arrangement characteristic in the reference threshold arrangement, and each pixel of the image output by the two-dimensional planar image output means has at least 3 pixels in the main scanning direction and the sub-scanning direction. The image processing apparatus according to claim 9 , wherein the image processing apparatus is set so as to be spaced apart from each other. The image converting means is arranged such that each pixel of the image output by the two-dimensional planar image output means is spaced apart from a distance corresponding to two pixels by the local periodic and regular threshold arrangement characteristics in the reference threshold arrangement. The image processing apparatus according to claim 9 , wherein the image processing apparatus is set to be configured as described above. Image converting means, a locally aperiodic threshold array characteristics in the reference threshold array, according to claim 9, wherein it has set derived from the approximate calculation model that mimics the output characteristics of the 2-dimensional plane image output means Image processing device. The image processing apparatus according to claim 9 , wherein the image conversion means derives and sets locally aperiodic threshold value array characteristics in the reference threshold value array at random. The anisotropic threshold arrangement characteristic is large when the difference in output accuracy between the main scanning direction and the sub-scanning direction in the two-dimensional planar image output means is large, and is anisotropic when the difference in output accuracy is small. The image processing apparatus according to claim 9 , wherein the image processing apparatus is set so as to reduce the performance. 15. The image processing according to claim 9, wherein the image conversion means performs image conversion using a plurality of threshold value arrays generated by correlating the reference threshold value array together with the reference threshold value array. apparatus. 10. The image converting means according to claim 9 , wherein an area of 20% or less (excluding 0) within a prescribed threshold range in the reference threshold array is a relatively low gradation area. Image processing device. Two-dimensional after the color input gradation image data of one pixel M gradation is converted to output image data of one pixel N (M> N ≧ 2) gradation by performing dither processing using the reference threshold value array by the dither processing means. In a color image processing apparatus that outputs an image by a planar image output means,
A two-dimensional planar image output means having different output accuracy in the main scanning direction and the sub-scanning direction as the two-dimensional planar image output means;
The dither processing means has a reference threshold arrangement for at least two types of color components locally periodically and regularly in a region where the gradation is relatively low within a specified threshold range. , And setting a local non-periodic threshold arrangement characteristic in a region that is relatively intermediate to high gradation within a specified threshold range, and further, the local non-periodic threshold arrangement characteristic The output accuracy of the two-dimensional planar image output means is set so as to have an anisotropic threshold arrangement characteristic that sequentially grows so that dots are preferentially connected in the scanning direction. Color image processing device. The dither processing means uses a preset reference threshold array for dither processing for at least one color component, and inverts, rotates or shifts the reference threshold array for the remaining color components. 18. The color image processing apparatus according to claim 17, wherein the array is used for dither processing. The reference threshold value array for each color component used by the dither processing means for the dither processing is such that the pixel dots of the respective color components output by the two-dimensional planar image output means are in the same position in a region having a relatively low gradation within the specified threshold range. The color image processing apparatus according to claim 17 , wherein the color image processing apparatus is set so as not to be disposed on the screen. The dither processing means performs multi-value dither processing using a threshold array that straddles a plurality of dither threshold planes, and the threshold array that straddles each dither threshold plane is in an area where the input image data has a low gradation. 18. The color image processing apparatus according to claim 17, wherein the threshold value array is set so that a local periodic and regular output pattern is preferentially outputted by the two-dimensional planar image output means. The dither processing means performs multi-value dither processing using a threshold array that straddles a plurality of dither threshold planes. The threshold image straddling each dither threshold plane is changed from an intermediate gradation to a high gradation. Dots appearing by the output from the two-dimensional planar image output means when the type of the dot pattern that appears when output by the two-dimensional planar image output means when in the area is in the area where the input image data has a low gradation 18. The color image processing apparatus according to claim 17 , wherein the color image processing apparatus is set so as to be larger than types of patterns. The dither processing means performs multi-value dither processing using a threshold array that straddles a plurality of dither threshold planes. When the input image data is in a region having a low gradation, a threshold array straddling each dither threshold plane is used. A first dither processing function for performing multi-value dither processing by setting a threshold value array in which a local periodic and regular output pattern is preferentially output by the two-dimensional planar image output means; and input image data Is in the region from the intermediate gradation to the high gradation, the type of the dot pattern that appears when the threshold array straddling each dither threshold plane is output by the two-dimensional planar image output means is that the input image data is low gradation Multi-value dither processing is performed by setting the number to be larger than the types of dot patterns that appear by the output from the two-dimensional planar image output means. A second dither processing function, a color image processing apparatus according to claim 17, wherein selecting said respective dither processing function in accordance with the output accuracy of the two-dimensional plane image output unit. The dither processing means performs multi-value dither processing using a threshold array that straddles a plurality of dither threshold planes. When the input image data is in a region having a low gradation, a threshold array straddling each dither threshold plane is used. A first dither processing function for performing multi-value dither processing by setting a threshold value array in which a local periodic and regular output pattern is preferentially output by the two-dimensional planar image output means; and input image data Is in the region from the intermediate gradation to the high gradation, the type of the dot pattern that appears when the threshold array straddling each dither threshold plane is output by the two-dimensional planar image output means is that the input image data is low gradation Multi-value dither processing is performed by setting the number to be larger than the types of dot patterns that appear by the output from the two-dimensional planar image output means. A second dither processing function, a color image processing apparatus according to claim 17, wherein selecting said respective dither processing function in accordance with the color component. Two-dimensional after the color input gradation image data of one pixel M gradation is converted to output image data of one pixel N (M> N ≧ 2) gradation by performing dither processing using the reference threshold value array by the dither processing means. In a color image processing apparatus that outputs an image by a planar image output means,
A two-dimensional planar image output means having different output accuracy in the main scanning direction and the sub-scanning direction as the two-dimensional planar image output means;
The dither processing means has a locally periodic and regular threshold arrangement characteristic in a region having a relatively low gradation within a prescribed threshold range for at least two types of color components, and has a prescribed threshold arrangement characteristic. A first dither processing function for performing dither processing using a reference threshold value array having a locally non-periodic threshold value array characteristic in a region that is relatively intermediate to high gradation within the threshold range of locally periodic for the color components, and using the reference threshold array with regular threshold array characteristics and a second dither processing function for performing dither processing, the local aperiodic The threshold arrangement characteristic is set to have an anisotropic threshold arrangement characteristic that sequentially grows so that dots are preferentially connected in a scanning direction in which the output accuracy of the two-dimensional planar image output means is relatively low. A color image processing apparatus. Two-dimensional after the color input gradation image data of one pixel M gradation is converted to output image data of one pixel N (M> N ≧ 2) gradation by performing dither processing using the reference threshold value array by the dither processing means. In a color image processing apparatus that outputs an image by a planar image output means,
A two-dimensional planar image output means having different output accuracy in the main scanning direction and the sub-scanning direction as the two-dimensional planar image output means;
The dither processing means has a local periodic and regular threshold arrangement characteristic in a region having a relatively low gradation within a predetermined threshold range for at least two kinds of color components, and has a predetermined threshold arrangement characteristic. A first dither processing function for performing dither processing using a reference threshold value array having locally non-periodic threshold value array characteristics in a region that is relatively intermediate to high gradation within the threshold range of A second dither processing function that performs dither processing by error diffusion processing for color components, and the output accuracy of the two-dimensional plane image output means is relatively A color image processing apparatus having an anisotropic threshold arrangement characteristic in which dots are successively grown so that dots are preferentially connected in a low scanning direction .

Images