JP3758030B2 - Image data interpolation program, image data interpolation method, and image data interpolation apparatus - Google Patents

Description

このように生成される補間画素の階調値は元画素から生成している点は特徴的な点である。すなわち、多階調値の画素に対して二値化を行い、その上でパターンマッチングを試みる。パターンマッチングするとマッチしたパターンに応じた計算式を利用して周囲の画素の階調値から補間画素の階調値を生成する。このように、パターンマッチングの判断を行いつつ、生成される画素はマッチしたパターンに基づいて用意されたルールに基づいて周囲の元画素から生成するので、単なるパターンマッチングに基づく補間処理だけでは得られない自然な補間画素を生成することができる。なお、先の計算でＲＧＢの各値から輝度を求める計算を行ったが、生成される補間画素の階調値は実際にはＲＧＢの各値を持つものであり、それぞれの値について元画素のＲＧＢの値を使用して所定の重み付けを行って算出する。従って、計算上便利な場合においては、輝度を使用した状態で重み付けを導出してから、ＲＧＢの値に対して同じ重み付けを適用して算出するようにしてもよい。
【００７１】
本例に示す３×３画素のパターンは上述のように水平方向（ｉ方向）には輝度値差があまりない、すなわち水平方向のエッジであるので、本式Ｐｎには水平に並んでいる３画素のもののみが代入されるようになっており、原則として元画素Ｌ，Ｍ，Ｎの階調値が代入される。また、本実施形態においては、さらに上記水平方向のエッジと垂直な方向の輝度値変化をも反映させた補間を行うようになっており、当該水平方向のエッジと垂直な方向に輝度値が変化している場合にはＰｎに元画素Ｑ，Ｒ，Ｓの階調値を代入し、エッジ変化の様子を補間画素ｇ，ｈ，ｉの階調値に反映させる。すなわち、注目画素Ｍの方がエッジ側の画素Ｒよりも値が大きい（小さい）ならば、画素Ｒでうめても強調にはならないため、注目画素をそのまま使用することとする。
【００７２】
このため、補間計算を行う前に画素Ｍと画素Ｒとの輝度値を比較し、画素Ｒの輝度値が画素Ｍの輝度値より大きい場合には上記元画素Ｑ，Ｒ，Ｓの階調値を上式に代入して補間画素ｇ，ｈ，ｉを生成している。この結果、補間画素ｄ，ｅ，ｆに比べて補間画素ｇ，ｈ，ｉの輝度値がより大きくなって補間された画像データにおいてエッジがより強調される。
この場合の具体的な判断条件は、
（（Ｍ＝エッジ）＆（Ｒ＞Ｍ））ｏｒ（（Ｍ≠エッジ）＆（Ｒ＜Ｍ））
であり、その際の計算式を示すと、
a=d=（Lx2＋Mx2+N）/5
b=e=（L＋Mx2+N）/4
c=f=（L＋Mx2+Nx2）/5
g=(Qx4+Rx6+Sx3)/13
h=(Q+Rx2+S)/4
i=(Qx3+Rx6+Sx4)/13
となる。一方、画素Ｒの輝度値が画素Ｍの輝度値より小さい場合は強調処理は行わず元画素Ｌ，Ｍ，Ｎの階調値のみが使用される。
【００７３】
この場合の具体的な判断条件は、
ｎｏｔ（（（Ｍ＝エッジ）＆（Ｒ＞Ｍ））ｏｒ（（Ｍ≠エッジ）＆（Ｒ＜Ｍ）））であり、その際の計算式を示すと、
a=d=g=（Lx2＋Mx2+N）/5
b=e=h=（L＋Mx2+N）/4
c=f=i=（L＋Mx2+Nx2）/5
となる。尚、上式で補間画素の階調値を計算するにあたり、元画素ＭをＰｎに代入する場合には重みとして距離の逆比を使用するとその値が大きくなりすぎるため、その場合のみ重みを半分にするようにしてある。
【００７４】
本実施形態においては、以上の規則によって水平エッジに対してその垂直方向への輝度値変化を考慮しつつ補間処理を実行しているが、当該水平エッジ以外にも種々のエッジパターンについての処理が可能であり図１３はその例を示している。上記ステップＳ２０６において同図（ａ）に示すパターンが発見されたときには、さらに３×３画素の領域外の所定画素を参照して同図（ｂ）に示す補間画素を生成する。同パターンにおける規則では、補間画素ａ，ｄ，ｅ，ｇ，ｈ，ｉの階調値計算においては、元画素Ｌ，Ｍ，Ｒを使用する。
【００７５】
補間画素ｂ，ｃ，ｆの階調値計算は５×５画素の領域に含まれる所定画素の輝度値に基づいて、いずれの画素を元画素にするかを変えている。すなわち、当該パターンが直角エッジであれば元画素Ｌ，Ｍ，Ｒを使用して補間画素ｂ，ｃ，ｆを生成するし、直角エッジでなければ元画素Ｇ，Ｈ，Ｎ，Ｓを使用して補間画素ｂ，ｃ，ｆを生成『する。
まず、本パターンが直角であるとする判断条件は、
（（（Ｍ＝Ｋ＝エッジ）＆（Ｆ＝エッジでない））ｏｒ（（Ｍ＝Ｗ＝エッジ）＆（Ｘ＝エッジでない）））
ｏｒ
（（（Ｍ＝Ｋ＝エッジでない）＆（Ｆ＝エッジ））ｏｒ（（Ｍ＝Ｗ＝エッジでない）＆（Ｘ＝エッジ）））
であり、その際の計算式を示すと
c=(L+Mx2+R)/4
b=(Lx4+Mx6+Rx3)/13
f=(Lx3+Mx6+Rx4)/13
a=(Lx2+Mx2+R)/5
d=(Lx3+Mx3+Rx2)/8
e=(L+Mx2+R)/4
g=(L+M+R)/3
h=(Lx2+Mx3+Rx3)/8
i=(L+Mx2+Rx2)/5
となる。次に、この直角以外でも直角に近い第一のパターンとしての条件判断は、
（（（Ｍ＝Ｆ＝エッジ）＆（Ａ＝エッジでない））ｏｒ（（Ｍ＝Ｘ＝エッジ）＆（Ｙ＝エッジでない）））
ｏｒ
（（（Ｍ＝Ｆ＝エッジでない）＆（Ａ＝エッジ））ｏｒ（（Ｍ＝Ｘ＝エッジでない）＆（Ｙ＝エッジ）））
であり、その際の計算式を示すと
c=(G+Hx2+Nx2+S)/6
b=(Lx4+Mx6+Rx3)/13
f=(Lx3+Mx6+Rx4)/13
となり、a,d,e,g,h,iについては同様の計算式とする。そして、上記二つの条件判断がいずれも成立しない場合の計算式を示すと、
c=(G+Hx2+Nx2+S)/6
b=(Gx15+Hx30+Nx20+Sx12)/77
f=(Gx12+Hx20+Nx30+Sx15)/77
となり、この場合もa,d,e,g,h,iについては同様の計算式とする。
【００７６】
さらに、図１３（ｃ）に示すパターンが発見されたときには、同図（ｄ）に示す補間画素を生成する。すなわち、同図（ｃ）に示すパターンは３×３画素領域の対角線を結んだ直線の一部（４５度のエッジ）らしいので、これらの直線を構成する元画素Ｇ、Ｍ、Ｓを使用して補間画素ａ，ｅ，ｉを生成する。また、５×５画素の領域に含まれる所定画素の輝度値に基づいて元画素のエッジ角度や線の太さを判別し、当該判別に応じて補間画素ｂ，ｃ，ｆ，ｄ，ｇ，ｈの生成に使用する元画素を「Ｇ，Ｍ，Ｓ」と「Ｈ，Ｎ」とで変更したり、「Ｇ，Ｍ，Ｓ」と「Ｌ，Ｒ」とで変更したりする。
【００７７】
検出した４５度のエッジか、４５度エッジの一部なのか３０度エッジの一部なのかの判断によって処理は３つに分岐する。この判断条件を図３６に示す。なお、図中で破線で囲った説明は各判断の意図する内容である。
処理１が選択された際の計算式は、
b=(Gx5+Mx10+Sx4)/19
c=(G+Mx2+S)/4
f=(Gx4+Mx10+Sx5)/19
となり、処理２が選択された際の計算式は、
b=(Gx5+Mx10+Sx4)/19
c=(H+N)/2
f=(Gx4+Mx10+Sx5)/19
となり、処理３が選択された際の計算式は、
b=(Hx3+Nx2)/5
c=(H+N)/2
f=(Hx2+Nx3)/5
となる。これら以外の画素についての計算式は、
a=(Gx3+Mx6+Sx2)/11
d=(Gx5+Mx10+Sx4)/19
e=(G+Mx2+S)/4
g=(G+Mx2+R)/4
h=(Gx4+Mx10+Sx5)/19
i=(Gx2+Mx6+Sx3)/11
となる。
【００７８】
図１３（ｅ）に示すパターンが発見されたときには、同図（ｆ）に示す補間画素を生成する。すなわち、上記図（ｃ）に示す直線よりさらに角度の浅い直線（３０度のエッジ）らしいので、これらの直線を構成する元画素Ｇ，Ｍ，Ｎを使用して補間画素ａ，ｂ，ｃ，ｅ，ｆを生成する。さらに、５×５画素の領域に含まれる所定画素の輝度値に基づいて元画素のエッジ角度や突出画素等を判別し、当該判別に応じて補間画素ｄ，ｇ，ｈ，ｉの生成に使用する元画素を「Ｇ，Ｍ，Ｎ」と「Ｌ，Ｒ，Ｓ」とで変更する。
【００７９】
より具体的に説明すると、検出した３０度のエッジについて処理は４つに分岐する。この判断条件を図３７に示す。なお、図中で破線で囲った説明は各判断の意図する内容である。
処理１が選択された際の計算式は、
d=(G+Mx2+N)/4
g=(G+Mx2+S)/4
h=(Gx3+Mx8+Nx5)/16
i=(G+Mx3+Nx3)/7
となり、処理２が選択された際の計算式は、
d=(G+Mx2+N)/4
g=(Lx2+Rx2+S)/5
h=(Lx4+Rx6+Sx3)/13
i=(Lx3+Rx6+Sx4)/13
となり、処理３が選択された際の計算式は、
d=(G+Mx2+N)/4
g=(Lx2+Rx2+S)/5
h=(Lx4+Rx6+Sx3)/13
i=(G+Mx3+Nx3)/7
となり、処理４が選択された際の計算式は、
d=(Lx15+Rx10+Sx6)/31
g=(Lx2+Rx2+S)/5
h=(Lx4+Rx6+Sx3)/13
i=(Lx3+Rx6+Sx4)/13
となる。これら以外の画素についての計算式は、
a=(Gx4+Mx6+Nx3)/13
b=(Gx3+Mx6+Nx4)/13
c=(G+Mx2+Nx2)/5
e=(Gx4+Mx6+Nx3)/13
f=(Gx2+Mx5+Nx5)/12
となる。
【００８０】
以上示したエッジパターンや５×５画素領域の参照手法，使用する元画素の選択手法等はその例示であり、種々の態様が採用可能であるし、補間画素を生成する際の倍率も上述のように縦横にそれぞれ３倍の画素を生成する態様の他、２倍，４倍等種々の倍率を採用することもできる。
まず、２倍補間の例について説明する。図３８〜図４２は元画素Ｍから補間画素ａ〜補間画素ｄが生成される場合の対応を示しており、図３８はパターンマッチングで水平エッジが発見された場合を示している。この場合も、注目画素Ｍの方がエッジ側の画素Ｒよりも値が大きい（小さい）ならば、画素Ｒでうめても強調にはならないため、注目画素をそのまま使用することとする。より具体的な判断条件は、
（（Ｍ＝エッジ）＆（Ｒ＞Ｍ））ｏｒ（（Ｍ＝エッジでない）＆（Ｒ＜Ｍ））
であり、これが成立すれば強調処理を行う。その場合の計算式は、
a=b=M
c=d=R
となり、これ以外の場合は強調処理を行わないので、
a=b=c=d=M
となる。
【００８１】
次に図３９は、パターンマッチングで角エッジが発見された場合であり、直角のエッジか否かの判断条件は、
（（（Ｍ＝Ｋ＝エッジ）＆（Ｆ＝エッジでない））ｏｒ（（Ｍ＝Ｗ＝エッジ）＆（Ｘ＝エッジでない）））
ｏｒ（（（Ｍ＝Ｋ＝エッジでない）＆（Ｆ＝エッジ））ｏｒ（（Ｍ＝Ｗ＝エッジでない）＆（Ｘ＝エッジ）））
であり、これが成立すれば直角であることを強調するために、
b=(L+R)/2
という式で計算し、そうでない場合には、
b=(Gx2+Hx3+Nx3+Sx2)/10
という式で計算する。その他の画素の計算式は、
a=(Lx3+Rx2)/5
c=(L+R)/2
d=(Lx2+Rx3)/5
となる。
【００８２】
図４０は、パターンマッチングで４５度のエッジが発見された場合であり、４５度のエッジの一部なのか３０度のエッジの一部なのかの判断を行うため、図４１に示す条件判断で２つの処理に分岐する。ここで、処理１となった場合の計算式は、
b=(G+S)/2
となり、処理２となった場合の計算式は、
b=(H+N)/2
となる。これ以外の画素の計算式は、
a=(Gx2+S)/3
c=(G+S)/2
d=(G+Sx2)/3
となる。
【００８３】
図４２は、パターンマッチングで３０度のエッジが発見された場合であり、図４３に示す条件判断で３つの処理に分岐する。ここで、処理１となった場合の計算式は、
c=(G+N)/2
d=(G+Nx2)/3
となり、処理２となった場合の計算式は、
c=(Lx3+Rx3+Sx2)/8
d=(Lx2+Rx3+Sx3)/8
となり、処理３となった場合の計算式は、
c=(Lx3+Rx3+Sx2)/8
d=(G+Nx2)/3
となる。これ以外の画素の計算式は、
a=(Gx3+Nx2)/5
b=(Gx2+Nx3)/5
となる。
【００８４】
上記３×３画素の領域において予め用意したエッジパターンと上記２値パターンとが一致したと判別されないときには、第一の補間処理としてこのようなパターンマッチング法が使用せず、ニアリスト法が実行される。ニアリスト法は図１４に示すように、周囲の四つの格子点Ｐｉｊ，Ｐｉ＋１ｊ，Ｐｉｊ＋１，Ｐｉ＋１ｊ＋１と内挿したい点Ｐｕｖとの距離を求め、もっとも近い格子点のデータをそのまま移行させる。これを一般式で表すと、
Ｐｕｖ＝Ｐｉｊ
ここで、ｉ＝［ｕ＋０．５］、ｊ＝［ｖ＋０．５］である。なお、［］はガウス記号で整数部分を取ることを示している。
【００８５】
図１５は、ニアリスト法で画素数を縦横３倍ずつに補間する状況を示している。補間する前には四隅の画素（□△○●）があるとして、補間して生成する画素にはこれらの画素のうちもっとも近い画素のデータをそのまま移行させている。すなわち、この例で言えば四隅の画素に隣接する画素についてそれぞれ複写することになる。また、かかる処理を行うと、図１６に示すように白い画素を背景として黒い画素が斜めに配置される元画像は、図１７に示すように黒の画素が縦横に３倍に拡大されつつ斜め方向に配置されることになる。ニアリスト法においては、画像のエッジがそのまま保持される特徴を有する。それ故に拡大すればジャギーが目立つもののエッジはエッジとして保持される。
【００８６】
このようにして第一の補間処理による補間画素を生成した後、上記ｒａｔｅが「１」でなければさらに上記第二の補間処理を実行する。本実施形態においてはこの第二の補間処理としていわゆるキュービック法による補間処理を実行するようになっており、以下かかる処理を説明する。キュービック法は図１８に示すように、内挿したい点Ｐｕｖを取り囲む四つの格子点のみならず、その一周り外周の格子点を含む計１６の格子点のデータを利用する。
【００８７】
内挿点Ｐｕｖを取り囲む計１６の格子点がそれぞれに値を備えている場合に、内挿点Ｐｕｖはそれらの影響を受けて決定される。例えば、一次式で補間しようとすれば、内挿点を挟む二つの格子点からの距離に反比例させて重みづけ加算すればよい。Ｘ軸方向に注目すると、内挿点Ｐｕｖから上記１６の格子点との距離は、図面上、左外側の格子点までの距離をｘ１、左内側の格子点までの距離をｘ２、右内側の格子点までの距離ｘ３、右外側の格子点までの距離ｘ４と表しつつ、このような距離に対応した影響度合いを関数ｆ（ｘ）で表すことにする。また、Ｙ軸方向に注目すると、内挿点Ｐｕｖから上記１６の格子点との距離は、上方外側の格子点までの距離をｙ１、上方内側の格子点までの距離をｙ２、下方内側の格子点までの距離ｙ３、下方外側の格子点までの距離ｙ４と表しつつ、同様に影響度合いは関数ｆ（ｙ）で表せる。
【００８８】
１６の格子点は以上のような距離に応じた影響度合いで内挿点Ｐｕｖに寄与するので、全ての格子点にデータに対してＸ軸方向とＹ軸方向のそれぞれの影響度合いを累積させる一般式は次式のようになる。
【００８９】
【数２】

また、ここで距離に応じた影響度合いを３次たたみ込み関数で表すとすると、
f(t) = {sin(πt)}/πt
となる。なお、上述した各距離ｘ１〜ｘ４，ｙ１〜ｙ４は格子点Ｐｕｖの座標値（ｕ，ｖ）について絶対値を利用して次のように算出することになる。
x1 = 1+(u-|u|) y1 = 1+(v-|v|)
x2 = (u-|u|) y2 = (v-|v|)
x3 = 1-(u-|u|) y3 = 1-(v-|v|)
x4 = 2-(u-|u|) y4 = 2-(v-|v|)
以上の前提のもとでＰについて展開すると、
【００９０】
【数３】

となる。なお、３次たたみ込み関数と呼ばれるように距離に応じた影響度合いｆ（ｔ）は次のような三次式で近似される。
【００９１】
【数４】

このキュービック法では一方の格子点から他方の格子点へと近づくにつれて徐々に変化していき、その変化具合がいわゆる３次関数的になるという特徴を有している。
【００９２】
図１９と図２０はキュービック法にて補間される際の具体例を示している。理解を容易にするため、垂直方向についてのデータの変化はなく、水平方向についてエッジが生じているモデルについて説明する。また、補間する画素を３点とする。
まず、図２０の具体的数値について説明する。補間前の画素の階調値を左列に「Ｏｒｉｇｉｎａｌ」として示しており、階調値「６４」の画素（Ｐ０、Ｐ１、Ｐ２、Ｐ３）が４点並び、階調値「１２８」の画素（Ｐ４）を１点挟み、階調値「１９２」の画素（Ｐ５、Ｐ６、Ｐ７、Ｐ８、Ｐ９）が５点並んでいる。この場合、エッジは階調値「１２８」の画素の部分である。
【００９３】
ここで各画素間に３点の画素（Ｐｎ１、Ｐｎ２、Ｐｎ３）を内挿することになると、内挿される画素間の距離は「０．２５」となり、上述したｘ１〜ｘ４は内挿点毎に表の中程の列の数値となる。ｘ１〜ｘ４に対応してｆ（ｘ１）〜ｆ（ｘ４）も一義的に計算されることになり、例えば、ｘ１，ｘ２，ｘ３，ｘ４が、それぞれ「１．２５」、「０．２５」、「０．７５」、「１．７５」となる場合、それに対するｆ（ｔ）については、概略「−０．１４」、「０．８９」、「０．３０」、「−０．０５」となる。また、ｘ１，ｘ２，ｘ３，ｘ４が、それぞれ「１．５０」、「０．５０」、「０．５０」、「１．５０」となる場合、それに対するｆ（ｔ）については、「−０．１２５」、「０．６２５」、「０．６２５」、「−０．１２５」となる。また、ｘ１，ｘ２，ｘ３，ｘ４が、それぞれ「１．７５」、「０．７５」、「０．２５」、「１．２５」となる場合、それに対するｆ（ｔ）については、概略「−０．０５」、「０．３０」、「０．８９」、「−０．１４」となる。以上の結果を用いて内挿点の階調値を演算した結果を表の右列に示しているとともに、図１９においてグラフで示している。なお、このグラフの意味するところについて後に詳述する。
【００９４】
垂直方向についてのデータの変化がないものとみなすと、演算は簡略化され、水平方向に並ぶ四つの格子点のデータ（P1,P2,P3,P4 ）だけを参照しつつ、内挿点から各格子点までの距離に応じた影響度合いｆ（ｔ）を利用して次のように算出できる。
P=P1・f(x1)+P21f(x2)+P3・f(x3)+P4・f(x4)
従って、内挿点Ｐ２１について算出する場合には、
P21=64*f(1.25)+64*f(0.25)+64*f(0.75)+128*f(1.75)
=64*(-0.14063)+64*(0.890625)+64*(0.296875)+128*(-0.04688)
=61
となる。
【００９５】
キュービック法によれば３次関数的に表せる以上、そのカーブの形状を調整することによって補間結果の品質を左右することができる。
その調整の一例として、
0＜t＜0.5 f(t) = -(8/7)t**3-(4/7)t**2+1
0.5＜t＜1 f(t) = (1-t)(10/7)
1＜t＜1.5 f(t) = (8/7)(t-1)**3+(4/7)(t-1)**2-(t-1)
1.5＜t＜2 f(t) = (3/7)(t-2)
としたものをハイブリッドバイキュービック法あるいはＭ（モディファイド）キュービック法と呼ぶことにする。
【００９６】
図２１はＭキュービック法にて補間される際の具体例を示しており、キュービック法の場合と同じ仮定のモデルについて補間した結果を示している。また、図１９にもＭキュービック法による補間処理結果を示しており、この例では３次関数的なカーブがわずかに急峻となり、画像全体のイメージがシャープとなる。Ｍキュービック法，通常のキュービック法のいずれを使用するにしても、上記第二の補間処理においてはキュービック法によって補間画素を生成すれば、上記重畳データ計算式の変数、すなわち第一の補間処理データ，ｒａｔｅ，第二の補間処理データの全てが算出されていることになるので、同式に値を代入することによって補間画素の重畳データを算出する。
【００９７】
以下、図２２（ａ）に示すような画像データに対して重畳処理が行われる様子を説明する。ここで、当該説明中における注目画素の周り５×５の領域では輝度値Ｙの出現回数が１５より大きく、上記ｒａｔｅが「０」，「１」でないものとする。従って、ステップＳ１０８にてｒａｔｅが計算された後、ステップＳ１１４にて第一の補間処理を実行する。ステップＳ２０２では図２２（ａ）に示す注目画素周りの３×３画素の領域（実線領域）を抽出し、ステップＳ２０４で当該領域データから図２２（ｂ）に示す２値パターンを作成する。
【００９８】
この２値パターンは予め用意された所定のエッジパターンと一致し、ステップＳ２０８にてパターンマッチング法による処理が行われ、図２２（ｃ）に示す補間画素が生成される。このとき、図２２（ａ）に示す３×３の画素は対角線方向のエッジであって左下に向けて輝度勾配がついている。従って、図２２（ｃ）に示す補間画素においては、かかる３×３画素の性質を反映し対角線方向のエッジを保持しつつ左下に向けて輝度勾配がついてエッジを強調している。
【００９９】
一方、ステップＳ１１８では図２２（ａ）に示す画素データを使用して図２２（ｄ）に示す補間画素を生成する。ここで、上述のように、補間画素生成にはその周りの１６個の画素データが使用され、例えば、図２２（ｄ）に黒点として示す右上の画素を生成するには図２２（ａ）に示す実線および点線領域からなる１６画素を使用する。かかる補間画素においては、キュービック法の性質から上記実線および点線領域の画素相互間の微妙な変化を反映しつつも上記パターンマッチング法に比べてエッジが曖昧になっている。すなわち、補間画素の対角線方向にエッジがあるが左下方向に向けての輝度勾配が緩やかになってエッジが曖昧になっている。
【０１００】
このように第一の補間処理および第二の補間処理にて補間画素を生成した後には第一の補間処理による各補間画素にｒａｔｅが乗じられ、第二の補間処理による各補間画素に（１−ｒａｔｅ）が乗じられて図２２（ｅ）に示すように重畳画素が生成される。同図（ｅ）に示すように、注目画素に対する重畳補間画素を生成すると、ステップＳ１２２の判別を経て新たな座標値の注目画素について全て補間処理し、ステップＳ１２４にて補間画像データを次段の処理へ引き渡す。ただし、補間倍率によっては補間画像データのデータ量が極めて多大になることもあるし、そもそもプリンタドライバ１２ｃが利用可能なメモリ領域がさほど多くない場合もある。このような場合には一定のデータ量ごとに分けて出力するようにしても構わない。
【０１０１】
最後に、本実施形態の総括を図３４に示している。スキャナ１１ａが写真をスキャンして電子化したものは元画像データとなり、以後の過程で処理対象となる。この元画像データを取得する手段は画像データ取得手段Ｃ１１である。
ステップＳ１０４では、元画像データを読み込み、注目画素を中心とした周辺の５×５画素の領域で輝度値のヒストグラムを作成する。ステップＳ１０６では、得られたヒストグラムにて異なる輝度値が出現する回数を取得し、所定のしきい値よりも少なければ、画像の性質が単純な非自然画であると判断して第一の補間処理の重畳比率（ｒａｔｅ）を１としてしまうが、多ければ画像の性質は自然画あるいは非自然画の混合画像である判断して評価関数Ｆによってｒａｔｅを決定する。この評価関数Ｆは上記領域での最大輝度値Ｙｍａｘと最小輝度値Ｙｍｉｎとの差の関数であり、エッジらしさを評価したものとなる。このようにして画像の性質から重畳比率を得る手段は第一重畳比率決定手段Ｃ１４である。
【０１０２】
ステップＳ１１４では、注目画素を中心とした所定の領域でパターンマッチング法かニアリスト法によってエッジを保存しやすい傾向にある補間処理を行なうので、これが第一の補間処理手段Ｃ１２となる。
一方、ステップＳ１１８では、画素相互間の微妙な変化を反映しつつもパターンマッチング法に比べてエッジが曖昧になるキュービック法で補間処理を行うので、これが第二の補間処理手段Ｃ１３となる。
【０１０３】
そして、ステップＳ１２０では、上記のようにして決定された重畳比率（ｒａｔｅ）を使用する計算式（１）により、それぞれ別個に補間処理された画素データに重み付けをつけて加算し、補間画素を生成する。この重畳比率の総和は当然「１」となる。むろん、この手段が画像データ重畳手段Ｃ１５である。
以後、注目画素を逐次移動させていって全画素について処理を実施したら、補間画像が作成される。従って、ステップＳ１２４では完成した補間画像データを出力する。この過程が画像データ出力手段Ｃ１６であるが、一例としてプリンタ１７ｂに出力すれば印刷物が得られることになる。
【０１０４】
次に、本発明の他の実施形態を説明する。この実施形態では、画像データ補間プログラムを印刷システムに使用する。なお、上述した実施形態と共通する構成部品などについてはそのまま参照する。
上述したカラープリンタ１７ｂには、プリンタドライバ１２ｃを介してアプリケーション１２ｄの処理結果が印刷データとして出力され、同カラープリンタ１７ｂは色インクを用いて印刷用紙上にドットを付すことにより対応する画像を印刷する。
【０１０５】
図２３〜図２５にはこのようなカラープリンタの一例としてカラーインクジェットプリンタ２１の概略構成を示している。本カラーインクジェットプリンタ２１は、三つの印字ヘッドユニットからなる印字ヘッド２１ａと、この印字ヘッド２１ａを制御する印字ヘッドコントローラ２１ｂと、当該印字ヘッド２１ａを桁方向に移動させる印字ヘッド桁移動モータ２１ｃと、印字用紙を行方向に送る紙送りモータ２１ｄと、これらの印字ヘッドコントローラ２１ｂと印字ヘッド桁移動モータ２１ｃと紙送りモータ２１ｄにおける外部機器とのインターフェイスにあたるプリンタコントローラ２１ｅとからなるドット印刷機構を備え、印刷データに応じて印刷用紙である記録媒体上で印字ヘッド２１ａを走査しながら画像印刷可能となっている。
【０１０６】
また、図２４は印字ヘッド２１ａの具体的な構成を示しており、図２５はインク吐出時の動作を示している。印字ヘッド２１ａには色インクタンク２１ａ１からノズル２１ａ２へと至る微細な管路２１ａ３が形成されており、同管路２１ａ３の終端部分にはインク室２１ａ４が形成されている。このインク室２１ａ４の壁面は可撓性を有する素材で形成され、この壁面に電歪素子であるピエゾ素子２１ａ５が備えられている。このピエゾ素子２１ａ５は電圧を印加することによって結晶構造が歪み、高速な電気−機械エネルギー変換を行うものであるが、かかる結晶構造の歪み動作によって上記インク室２１ａ４の壁面を押し、当該インク室２１ａ４の容積を減少させる。すると、このインク室２１ａ４に連通するノズル２１ａ２からは所定量の色インク粒が勢いよく吐出することになる。このポンプ構造をマイクロポンプ機構と呼ぶことにする。
【０１０７】
なお、一つの印字ヘッドユニットには独立した二列のノズル２１ａ２が形成されており、各列のノズル２１ａ２には独立して色インクが供給されるようになっている。従って、三つの印字ヘッドユニットでそれぞれ二列のノズルを備えることになり、最大限に利用して六色の色インクを使用することも可能である。図２３に示す例では、左列の印字ヘッドユニットにおける二列を黒インクに利用し、中程の印字ヘッドユニットにおける一列だけを使用してシアン色インクに利用し、右列の印字ヘッドユニットにおける左右の二列をそれぞれマゼンタ色インクとイエロー色インクに利用している。
【０１０８】
一方、印字ヘッド２１ａに形成されているノズル２１ａ２の鉛直方向の間隔は印刷解像度とは一致せず、一般的にはこのノズル２１ａ２は印刷解像度よりも大きな間隔で形成されている。これにもかかわらずより高解像度の印刷を可能とするのは、紙送り方向について紙送りモータ２１ｄを制御するからである。例えば、ノズル２１ａ２の間隔の間で紙を８段階で送り、各段階毎に印字ヘッド２１ａを桁送り方向に操作して印刷すれば解像度は向上する。むろん、桁送り方向については任意の間隔で色インクを吐出すればそれが解像度と言えるから、タイミングの制御次第と言える。なお、厳密な意味では色インクのドット径も解像度の要素となりえるが、ここでは理解の簡便のため無視することにする。
【０１０９】
本実施形態においては、上述したようなハードウェアシステムを前提とし、コンピュータシステム１０の画像入力デバイスで取得した画像データに基づいて印刷を実行する。その際、元の画像データの解像度とカラープリンタ１７ｂの解像度とに差がある場合には補間処理を実行することになる。ここで、アプリケーション１２ｄが印刷処理を実行した際にカラープリンタ１７ｂに対して印刷データが出力される際の解像度と階調度の変化について説明する。
【０１１０】
図２６は画像データの流れを示している。アプリケーション１２ｄはオペレーティングシステム１２ａに対して印刷要求を発生し、その際に出力サイズと印刷用紙と印刷速度とインクの種類とＲＧＢ２５６階調の画像データを受け渡す。すると、オペレーティングシステム１２ａはプリンタドライバ１２ｃに対してこの出力サイズと印刷用紙と印刷速度とインクの種類と画像データを受け渡す。ここで、オペレーティングシステム１２ａはディスプレイドライバ１２ｂを介してディスプレイ１７ａに表示を行いつつ、キーボード１５ａやマウス１５ｂの操作結果をプリンタドライバ１２ｃに出力し、プリンタドライバ１２ｃは上記指定された出力サイズになるよう画像補間処理を実行して印刷データを生成する。通常、この印刷データはＣＭＹＫ２階調であり、オペレーティングシステム１２ａを介してハードウェアポートよりカラープリンタ１７ｂに出力されることになる。
【０１１１】
このように、本実施形態においては、印刷制御プログラムをコンピュータシスム１０にて実行してカラープリンタ１７ｂに印刷データを出力しているが、対象となる印刷装置は上述したインクジェット方式のカラープリンタ２１に限定されるものではない。例えば、同カラープリンタ２１はマイクロポンプ機構を採用するインクジェット方式のものであるがマイクロポンプ機構以外のものを採用することも可能である。図２７に示すようにノズル２１ａ６近傍の管路２１ａ７の壁面にヒータ２１ａ８を設けておくとともに、このヒータ２１ａ８に加熱して気泡を発生させ、その圧力で色インクを吐出するようなバブルジェット（Ｒ）方式のポンプ機構も実用化されている。
【０１１２】
また、他の機構として図２８にはいわゆる電子写真方式のカラープリンタ２２の主要部概略構成を示している。感光体としての回転ドラム２２ａの周縁には回転方向に対応して帯電装置２２ｂと露光装置２２ｃと現像装置２２ｄと転写装置２２ｅとが配置され、帯電装置２２ｂにて回転ドラム２２ａの周面を均一に帯電させた後、露光装置２２ｃによって画像部分の帯電を除去し、現像装置２２ｄで帯電していない部分にトナーを付着させ、転写装置２２ｅによって同トナーを記録媒体としての紙上に転写させる。その後、ヒータ２２ｆとローラ２２ｇとの間を通過させて同トナーを溶融して紙に定着させている。そして、これらが一組となって一色のトナーによる印刷を行わせることになるので、合計四色分が個別に備えられている。
【０１１３】
すなわち、その印刷装置の具体的な構成は特に限定されるものではないし、このような個別的な印刷手法の適用範囲のみならずその適用態様についても各種の変更が可能である。
次に、上述した印刷システムを利用して出力解像度に応じた最適な画像処理を実行する処理について説明する。
図２９は、この印刷システムの概略構成を示すブロック図である。印刷品質取得手段Ｃ２１は、実行された印刷ジョブにおいて印刷させようとする印刷品質を取得し、第一の補間処理手段Ｃ２２および第二の補間処理手段Ｃ２３はこの画像データにおける構成画素数を増やす補間処理を行う。ここで、第一の補間処理手段Ｃ２２は補間処理としてパターンマッチング法とニアリスト法とを実行可能になっており、第二の補間処理手段Ｃ２３は補間処理としてキュービック法を実行可能になっている。
【０１１４】
第二重畳比率決定手段Ｃ２４は上記印刷品質取得手段Ｃ２１が取得した印刷品質に基づいて上記第一の補間処理手段Ｃ２２および第二の補間処理手段Ｃ２３による補間画素の重畳比率を決定する。同第二重畳比率決定手段Ｃ２４が重畳比率を決定すると、画像データ重畳手段Ｃ２５が当該重畳比率にて上記第一の補間処理手段Ｃ２２および第二の補間処理手段Ｃ２３による補間画素データを重畳する。そして、印刷制御処理手段Ｃ２６は画像データ重畳手段Ｃ２５が重畳した画素データに基づいて印刷用データを生成するとともに印刷を実行する。
【０１１５】
図３０は上述したプリンタドライバ１２ｃが実行する印刷処理に関連するソフトウェアフローを示している。図３０においてステップＳ３０２では元画像データを取得する。例えば、アプリケーション１２ｄにてスキャナ１１ａから画像を読み込み、所定の画像処理を行った後で印刷処理すると、所定の解像度の画像データがオペレーティングシステム１２ａを介してプリンタドライバ１２ｃに引き渡されるため、この引渡の段階が該当する。
【０１１６】
ステップＳ３０３では印刷設定を入力する。アプリケーション１２ｄにて印刷処理を実行する際には、オペレーティングシステム１２ａがＧＵＩ環境を提供するものとすると図３１に示すように印刷操作用のウィンドウ表示が行われる。ここで入力されるパラメータなどは各種のものを採用可能であるが、本実施形態においては、「（印刷の）部数」、「開始ページ」、「終了ページ」などがある。また、操作指示ボタンとしては「ＯＫ」ボタンと「キャンセル」ボタンとともに、「プリンタの設定」ボタンも用意されている。
【０１１７】
「プリンタの設定」を指示すると、図３２に示すようなウィンドウ表示が行われる。このウィンドウ表示はプリンタ毎の機能に応じた各種の設定を行うために用意されており、後述するようにこのウィンドウにおける設定内容に応じて重畳比率が変更される。本例では「（印刷）解像度」として「３６０ｄｐｉ」と「７２０ｄｐｉ」の一方を選択できる。また、「用紙」として「Ａ４」か「Ｂ５」のサイズおよび「普通紙」か「光沢紙」の品質、「印刷速度」として「速い」か「遅い」か、「インク」として「顔料」か「染料」かを選択できる。むろん、この印刷設定は一例であり、これらの全てをユーザに選択させるように構成する必要はなく、上記カラープリンタ１７ｂと双方向通信を行ってインク種類等を行う等種々の態様を採用可能である。
【０１１８】
本実施形態においては、上述の設定内容はオペレーティングシステム１２ａの管理下にある設定ファイルに格納されるようになっており、既に設定がなされているのであれば当該設定ファイルを参照して読み出すし、操作者が印刷操作に伴って設定内容を変更する場合には変更後の設定内容を印刷設定として読み出す。このようなステップＳ３０３の処理はソフトウェアとしてみるときに上記印刷品質取得工程あるいは印刷品質取得機能ということになるが、当該印刷品質取得工程を含めてコンピュータに実行させる各種のステップは、オペレーティングシステム１２ａ自体やハードウェアを直接に含まないものとして理解することができる。これに対して、ＣＰＵなどのハードウェアと有機一体的に結合したものと考えると印刷品質取得手段Ｃ２１に該当する。
【０１１９】
ステップＳ３０４では読み込んだ画像データにおいて、補間される画素を注目画素とし、当該注目画素を中心として周辺の５×５画素の領域の画素データを参照画素としつつ、同参照画素の輝度値のヒストグラムを作成する。ステップＳ３０６では得られたヒストグラムにて異なる輝度値が出現する回数を取得し、同出現回数が１５より小さいか否かを判別する。異なる輝度値が多いものほど参照画素中に色数が多いことから、ここでは輝度値出現回数が１５より小さいものを非自然画であるとしており、上記ステップＳ３０６にて出現回数が１５より小さいと判別されたときにステップＳ３１０にて上記第一の補間処理の重畳比率（ｒａｔｅ）を１にする。
【０１２０】
ステップＳ３０６にて輝度値出現回数が１５より小さくないと判別されたときにはステップＳ３０７にて上記ステップＳ３０３で入力した印刷設定内容に基づいて評価関数Ｆを決定する。同ステップＳ３０７における評価関数Ｆの決定では印刷品質が高いときに上記第二の補間処理の重畳比率を大きくなるように評価関数Ｆ決定する。詳細は後述するが、本実施形態においては印刷品質が高い場合と低い場合との２種の評価関数が用意されており、印刷品質が高い場合の評価関数値は印刷品質が低い場合の評価関数値より小さくなっている。また、この評価関数Ｆは上記参照画素の輝度値幅すなわち参照画素中の最大輝度値Ｙｍａｘと最小輝度値Ｙｍｉｎとの差の関数である。
【０１２１】
ステップＳ３０８では上記ステップＳ３０７にて決定された評価関数Ｆに対して上記ヒストグラムに基づく最大輝度値Ｙｍａｘと最小輝度値Ｙｍｉｎとの差を代入し、ｒａｔｅを決定する。従って、ステップＳ３０４，Ｓ３０６，Ｓ３０７，Ｓ３０８，Ｓ３１０に示す一連の処理が第二重畳比率決定工程あるいは第二重畳比率決定機能に相当するし、これらがＣＰＵなどのハードウェアと有機一体的に結合したものと考えると第二重畳比率決定手段Ｃ２４を構成することになる。
【０１２２】
ステップＳ３０８あるいはステップＳ３１０にてｒａｔｅを決定すると、ステップＳ３１２にて同ｒａｔｅが「０」であるか否かを判別し、ｒａｔｅが「０」である時には第一の補間処理を行わないようになっている。ステップＳ３１２にてｒａｔｅが「０」であると判別されないときにはステップＳ３１４にて第一の補間処理を実行する。従って、この処理が上記第一の補間処理工程あるいは第一の補間処理機能に相当するし、これらがＣＰＵなどのハードウェアと有機一体的に結合したものと考えると第一の補間処理手段Ｃ２２を構成することになる。
【０１２３】
さらに、ステップＳ３１６ではｒａｔｅが「１」であるか否かを判別し、ｒａｔｅが「１」である時には第二の補間処理を行わないようになっている。ステップＳ３１６にてｒａｔｅが「１」であると判別されないときにはステップＳ３１８にて第二の補間処理を実行する。従って、この処理が上記第二の補間処理工程あるいは第二の補間処理機能に相当するし、これらがＣＰＵなどのハードウェアと有機一体的に結合したものと考えると第二の補間処理手段Ｃ２３を構成することになる。このようにしてｒａｔｅを決定し、第一の補間処理あるいは第二の補間処理を実行した後にはステップＳ３２０にて当該ｒａｔｅに基づいて上述した式（１）により画素データを重畳して補間画素を生成する。
重畳データ＝
（第一の補間処理）×ｒａｔｅ＋（第二の補間処理）×（１−ｒａｔｅ）
…（１）
従って、この処理が上記画像データ重畳工程あるいは画像データ重畳機能に相当し、これらがＣＰＵなどのハードウェアと有機一体的に結合したものと考えると画像データ重畳手段Ｃ２５を構成することになる。このようにして上記注目画素に対する補間データの生成が終了すると、ステップＳ３２２にて上記入力した元画像データの全注目画素についての重畳処理が終了したか否かを判別し、全注目画素についての処理が終了するまで上記ステップＳ３０４以降の処理を繰り返す。
【０１２４】
全注目画素についての重畳処理が終了したら、ステップＳ３２４にて補間された画像データに基づく印刷を行う。プリンタドライバ１２ｃの場合、解像度変換だけで印刷データが得られるわけではなく、色変換であるとか、ハーフトーン処理が必要になる。従って、ここでは、かかる変換処理等を実行した上で印刷データを上記カラープリンタ１７ｂに出力しており、この処理が上記印刷制御処理工程あるいは印刷制御処理機能に相当し、これらがＣＰＵなどのハードウェアと有機一体的に結合したものと考えると印刷制御処理手段Ｃ２６を構成することになる。
【０１２５】
次に、以上のようなフローに対するより具体的な処理について説明する。本実施形態においては、ステップＳ３０２で取得する元画像がコンピュータグラフィックス（非自然画）であるか写真（自然画）であるかを判定し、印刷品質のみならず当該判定結果を重畳比率に反映させるようになっている。むろん、補間手法決定の誤りを防止するために、上記ｒａｔｅを「１」もしくは「０」に固定しない構成も可能であるが、本実施形態においては輝度値出現回数が１５より小さければまず非自然画であるとし、非自然画に対して第二の補間処理を実行して画像の輪郭を曖昧にさせないようにすることを重視してある。
【０１２６】
この判別は具体的には図１４に示す輝度値のヒストグラムを利用しており、ステップ１０４においては、５×５画素の領域の各参照画素について輝度値を求め、輝度が取りうる範囲において画素数のヒストグラムを集計する。そして当該輝度値出現回数すなわち、分布数が「０」でない輝度値がいくつ表れているかカウントし、ステップＳ３０６における判別を行う。
このように、輝度値のヒストグラムにおいて分布数が「０」でない輝度値が１５より多ければ、上記第一の補間処理の重畳比率を評価関数によって決定する。本実施形態においては、評価関数は予め高印刷品質用と低印刷品質用との２種類用意されており、上記図３２に示すウィンドウで設定した設定内容のうち、一つでも高品質を示すものがあればステップＳ３０６にて高印刷品質用評価関数を使用するよう決定する。図３３は評価関数Ｆ（ｙ）の一例を示しており、同図Ｆ１が低印刷品質用評価関数であり、同図Ｆ２が高印刷品質用評価関数である。
【０１２７】
評価関数Ｆ１（ｙ）は「ｙ」が「０≦ｙ≦２５５」の範囲で値を有し、「０≦Ｆ（ｙ）≦１」の範囲で変動する。また、「０≦ｙ≦６４」の範囲では「Ｆ（ｙ）＝０」であり、「１９２≦ｙ≦２５５」の範囲では「Ｆ（ｙ）＝１」であり、「６４≦ｙ≦１９２」の範囲ではＦ（ｙ）は「０」から「１」まで直線的に増加する。また、評価関数Ｆ２（ｙ）は「ｙ」が「０≦ｙ≦２５５」の範囲で値を有し、「０≦Ｆ（ｙ）≦０．７」の範囲で変動する。また、「０≦ｙ≦６４」の範囲では「Ｆ（ｙ）＝０」であり、「１９２≦ｙ≦２５５」の範囲では「Ｆ（ｙ）＝０．７」であり、「６４≦ｙ≦１９２」の範囲ではＦ（ｙ）は「０」から「０．７」まで直線的に増加する。
【０１２８】
ここで、「ｙ」には上記図１４における輝度値幅「Ｙｍａｘ−Ｙｍｉｎ」が代入される。従って、輝度値幅が大きいほど、すなわち参照画素相互の関係がよりエッジらしいものほどｒａｔｅが大きくなる。さらに、同じ輝度値幅であっても評価関数Ｆ１の値は常に評価関数Ｆ２の値以上である。すなわち、同じ輝度値幅でも高品質印刷の場合は画像の階調性を損なわないで補間処理を行う第二の補間処理の重畳比率が大きくなる。
【０１２９】
「高印刷解像度」，「高印刷用紙品質」，「低速印刷」のいずれかまたは組み合わせを選択した状態、すなわち高い印刷品質で印刷を行うときに、元画像の階調性を損なわない補間を行えば、微妙な階調変化が印刷結果において利用者の目に見える効果として再現される。一方、低い印刷品質で印刷を行うときには微妙な階調変化を再現可能な印刷データを使用して印刷を行っても印刷結果において利用者の目に見える効果として再現されない。従って、高印刷品質時に第二の補間処理を行う価値があると言え、本実施形態では評価関数Ｆ２を使用することで同第二の補間処理の重畳比率を高くする。
【０１３０】
また、本実施形態ではインクの種類も上記評価関数の選択に影響を与えている。ここで、インクの種類は「顔料」と「染料」が選択可能であり、いずれのものが高品質であるとは一概に決定できないが、「顔料」は「染料」に比べてにじまない傾向にあり、両者を比べると印刷結果において「顔料」の方がエッジが目立つ傾向にある。従って、インクの種類は印刷品質を反映していると言え、本実施形態において両インクで同程度の印刷結果を得ようとすると「顔料」の場合に「染料」に比べて第二の補間処理の比率が高くなるようにすればよい。そこで、本実施形態においては、インクが「顔料」である時には上記評価関数Ｆ２を使用するようになっている。
【０１３１】
むろん、本実施形態における評価関数および上記取得した印刷設定の反映のさせ方は一例であって、種々の態様を採用することができる。例えば、評価関数は上述のように２種類である必要はなく、複数種類のものを用意することができる。上記実施形態のように印刷品質に影響を与える印刷設定項目が４種類ある場合に４つの評価関数を用意し、高品質に設定された項目が一つ増える毎により最大値の小さな評価関数を選択するようなことが考えられる。
【０１３２】
また、このように予め複数の評価関数を用意しておく必要もなく、上述の評価関数Ｆ１を一つ用意しておいて高品質印刷時に当該評価関数を「０．７倍」したものを使用してｒａｔｅを決定するように構成することもできるし、高品質に設定された項目が一つ増えるごとに評価関数を「０．９倍」する等の構成も可能である。さらに、関数の形状も上記図３３に示す例に限る必要はなく、輝度値幅を変数としたときに輝度値幅とともに単調に増加する関数であればよいし、輝度値幅「０」〜「２５５」にわたってなめらかに変化する評価関数を採用すること等もできる。
【０１３３】
上述のようにしてステップＳ３０７にて使用する評価関数が決定されると、ステップＳ３０８にて当該決定された評価関数によって重畳比率ｒａｔｅを決定する。図１４のようなヒストグラムであれば、分布数が「０」でない輝度値が１９個あるのでステップＳ３０６の判別を経てステップＳ３０７の処理を実行する。また、図３２に示すように解像度にて「７２０ｄｐｉ」，用紙にて「光沢紙」、印刷速度にて「速い」、インクにて「顔料」が選択されている状態では、解像度と用紙とインクにて高品質印刷用の評価関数を選択するべき印刷設定がなされているので、ステップＳ３０７にて評価関数Ｆ２を選択する。そして、ステップＳ３０８にて評価関数Ｆ２によってｒａｔｅが決定され、ステップＳ３１４とステップＳ３１８にてそれぞれ第一の補間処理と第二の補間処理が実行される。
【０１３４】
第一の補間処理による補間画素を生成した後、上記ｒａｔｅが「１」でなければさらに上記第二の補間処理を実行する。
Ｍキュービック法，通常のキュービック法のいずれを使用するにしても、上記第二の補間処理においてはキュービック法によって補間画素を生成すれば、上記重畳データ計算式（１）の変数、すなわち第一の補間処理データ，ｒａｔｅ，第二の補間処理データの全てが算出されていることになる。式（１）に値を代入することによって補間画素の重畳データを算出する。
【０１３５】
ここで、上記図３２に示すようにこの状態における印刷品質は高品質であって、微妙な階調変化を印刷結果に反映できる状態である。このとき評価関数はＦ２であるのでキュービック法の重畳比率が高くなっている。
一方、上記図３２に示すウィンドウにおいて、解像度で「３６０ｄｐｉ」，用紙にて「普通紙」、印刷速度にて「速い」、インクにて「染料」が選択されている状態では印刷品質は低く、第二の補間処理による効果は目立たない。この状態では、上述のように評価関数Ｆ１が使用され、上記図３２に示す設定状態と比較して第二の補間処理の重畳比率が低くなっている。このときには階調値を再現するようなキュービック法よりエッジを強調するパターンマッチング法あるいはニアリスト法の重畳比率の方が高くなり、低印刷品質でありながらもシャープな印刷結果を得ることができる。
【０１３６】
最後に、本発明の総括を図３５に示している。
「プリンタの設定」のウィンドウ表示ではプリンタ毎の機能に応じた各種の設定が可能であり、その設定内容は設定ファイルに格納される。既に設定がなされているのであれば当該設定ファイルを参照して読み出すし印刷操作に伴って設定内容を変更する場合には変更後の設定内容を読み出す。ステップＳ３０３ではこのようにして印刷設定を入力し、印刷品質取得手段Ｃ２１に該当する。
【０１３７】
一方、ステップＳ３０４では、元画像データを読み込み、注目画素を中心とした周辺の５×５画素の領域で輝度値のヒストグラムを作成する。ステップＳ３０６では、得られたヒストグラムにて異なる輝度値が出現する回数を取得し、所定のしきい値よりも少なければ、画像の性質が単純な非自然画であると判断して第一の補間処理の重畳比率（ｒａｔｅ）を１としてしまうが、多ければ画像の性質は自然画あるいは非自然画の混合画像である判断して評価関数Ｆによってｒａｔｅを決定するために評価関数を決定する。評価関数Ｆは印刷品質に応じて決定され、低画質用の評価関数Ｆ１と高画質用の評価関数Ｆ２とがある。評価関数Ｆは、上記領域での最大輝度値Ｙｍａｘと最小輝度値Ｙｍｉｎとの差の関数であり、エッジらしさを評価したものとなる。エッジらしさが高い場合には評価関数Ｆ１でエッジが強調されやすい比較的演算負荷の低い補間処理の割合を高め、そうでなければ評価関数Ｆ２で滑らかさを強調する演算負荷の高い補間処理の割合を高める。このようにして印刷品質の高低から重畳比率を得る手段は第二重畳比率決定手段Ｃ２４である。
【０１３８】
ステップＳ３１４では、注目画素を中心とした所定の領域でパターンマッチング法かニアリスト法によってエッジを保存しやすい傾向にある補間処理を行なうので、これが第一の補間処理手段Ｃ２２となる。
一方、ステップＳ３１８では、画素相互間の微妙な変化を反映しつつもパターンマッチング法に比べてエッジが曖昧になるキュービック法で補間処理を行うので、これが第二の補間処理手段Ｃ２３となる。
【０１３９】
そして、ステップＳ３２０では、上記のようにして決定された重畳比率（ｒａｔｅ）を使用する計算式（１）により、それぞれ別個に補間処理された画素データに重み付けをつけて加算し、補間画素を生成する。この重畳比率の総和は当然「１」となる。むろん、この手段が画像データ重畳手段Ｃ２５である。
以後、注目画素を逐次移動させていって全画素について処理を実施したら、補間画像が作成され、ステップＳ３２４では完成した補間画像データを使用してプリンタ１７ｂに印刷させることになる。この過程が印刷制御処理手段Ｃ２６である。
【０１４０】
このように、本発明においては、所定の評価関数に基づいて第一の補間処理および第二の補間処理を重畳する。従って、重畳された画素はパターンマッチング法のみの場合に比べてエッジが曖昧になっているが、キュービック法のみの場合に比べるとエッジがシャープになっている。また、キュービック法のみの場合に比べて微妙な階調変化が低減しているが、パターンマッチング法のみの場合に比べると階調変化が豊かになっている。また、上記評価関数が輝度値幅の関数であることから、画像の性質に応じた補間処理比率を決定可能である。さらに、補間処理の効果に対して直接的に影響を与える印刷品質に対してより適した補間処理の重畳比率を高くする。この結果、個々の補間処理の長所がより目立つようになる。また、両者の欠点を際だたせてしまうようなこともない。従って、補間対象画像の性質判別に基づく補間手法決定の誤りを防止しつつ、印刷品質に応じた的確な補間処理を行うことができる。
【図面の簡単な説明】
【図１】 本発明の一実施形態にかかる画像データ補間装置のブロック図である。
【図２】 同画像データ補間装置の具体的ハードウェアのブロック図である。
【図３】 本発明の画像データ補間装置の他の適用例を示す概略図である。
【図４】 本発明の画像データ補間装置の他の適用例を示す概略図である。
【図５】 本発明の画像データ補間装置の他の適用例を示す概略図である。
【図６】 本発明の画像データ補間装置の他の適用例を示す概略図である。
【図７】 プリンタドライバが実行する解像度変換に関連する処理のフローチャートである。
【図８】 第一の補間処理フローチャートである。
【図９】 輝度値のヒストグラムを示す図である。
【図１０】 評価関数Ｆ（ｙ）の具体例を示す図である。
【図１１】 ３×３画素の輝度パターンの一例を示す図である。
【図１２】 参照画素である５×５画素の領域を示す図である。
【図１３】 エッジパターンの具体例を示す図である。
【図１４】 ニアリスト法の概念図である。
【図１５】 ニアリスト法で各格子点のデータが移行される状況を示す図である。
【図１６】 ニアリスト法の補間前の状況を示す概略図である。
【図１７】 ニアリスト法の補間後の状況を示す概略図である。
【図１８】 キュービック法の概念図である。
【図１９】 キュービック法の具体的適用時におけるデータの変化状況を示す図である。
【図２０】 キュービック法の具体的適用例を示す図である。
【図２１】 Ｍキュービック法の具体的適用例を示す図である。
【図２２】 画像データに対して重畳処理が行われる具体的な様子を示す図である。
【図２３】 インクジェット方式のカラープリンタの概略ブロック図である。
【図２４】 同カラープリンタにおける印字ヘッドユニットの概略説明図である。
【図２５】 同印字ヘッドユニットで色インクを吐出させる状況を示す概略説明図である。
【図２６】 本印刷システムにおける画像データの流れを示すフロー図である。
【図２７】 バブルジェット（Ｒ）方式の印字ヘッドで色インクを吐出させる状況を示す概略説明図である。
【図２８】 電子写真方式のプリンタの概略説明図である。
【図２９】 印刷システムの概略構成を示すブロック図である。
【図３０】 プリンタドライバが実行する印刷処理に関連する処理のフローチャートである。
【図３１】 印刷処理の操作ウィンドウを示す図である。
【図３２】 プリンタの設定の操作ウィンドウを示す図である。
【図３３】 評価関数Ｆ（ｙ）の具体例を示す図である。
【図３４】本発明の概略構成を示す図である。
【図３５】本発明の概略構成を示す図である。
【図３６】 ４５度のエッジに対する条件判断を示す図である。
【図３７】 ３０度のエッジに対する条件判断を示す図である。
【図３８】 ２倍補間の場合に水平エッジが発見された場合の元画素と補間画素との対応を示す図である。
【図３９】 ２倍補間の場合に直角エッジが発見された場合の元画素と補間画素との対応を示す図である。
【図４０】 ２倍補間の場合に４５度エッジが発見された場合の元画素と補間画素との対応を示す図である。
【図４１】 ４５度のエッジに対する条件判断を示す図である。
【図４２】 ２倍補間の場合に３０度エッジが発見された場合の元画素と補間画素との対応を示す図である。
【図４３】 ３０度のエッジに対する条件判断を示す図である。
【符号の説明】
１０…コンピュータシステム
１１ａ…スキャナ
１１ｂ…デジタルスチルカメラ
１１ｃ…ビデオカメラ
１２…コンピュータ本体
１２ａ…オペレーティングシステム
１２ｂ…ディスプレイドライバ
１２ｃ…プリンタドライバ
１２ｄ…アプリケーション
１３ａ…フロッピー（Ｒ）ディスクドライブ
１３ａ１…フロッピー（Ｒ）ディスク
１３ｂ…ハードディスク
１３ｃ…ＣＤ−ＲＯＭドライブ
１３ｃ１…ＣＤ−ＲＯＭ
１４ａ…モデム
１５ａ…キーボード
１５ｂ…マウス
１７ａ…ディスプレイ
１７ｂ…カラープリンタ
１８ａ…カラーファクシミリ装置
１８ｂ…カラーコピー装置
２１…カラーインクジェットプリンタ
２２…カラープリンタ[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to an image data interpolation program, an image data interpolation method, and an image data interpolation apparatus.
[0002]
[Prior art]
  When an image is handled by a computer or the like, the image is expressed by a dot matrix pixel, and each pixel is expressed by a gradation value. For example, photographs and computer graphics are often displayed with pixels of 640 dots in the horizontal direction and 480 dots in the vertical direction on a computer screen.
  On the other hand, the performance of color printers is remarkably improved, and the dot density is extremely high, such as 720 dpi (dot / inch). Then, if an image of 640 × 480 dots is printed in correspondence with each dot, the image becomes extremely small. In this case, since the gradation values are different and the meaning of the resolution itself is different, it is necessary to interpolate between dots and convert the data into printing data. That is, if the image is printed small with a one-to-one correspondence, a process for increasing the pixel of the image data (this is called high resolution or enlargement) is performed, and in the opposite case, the pixel of the image data is decreased. Processing (this is called resolution reduction or reduction) is performed.
[0003]
  Conventionally, as a method for interpolating dots in such a case, nearest neighbor interpolation (nearlist neighbor interpolation: hereinafter referred to as the nearlist method) or cubic convolution interpolation (cubic convolution interpolation: hereinafter) A technique such as a cubic method) and a pattern matching method are known. Each of these interpolation methods has its own characteristics. For example, in the cubic method, although the amount of calculation is large, image interpolation can be performed without impairing the gradation of the original image. Therefore, it is suitable for use in image interpolation of natural images generally composed of multi-gradation pixels. In the pattern matching method, although the gradation of the original image tends to be impaired, the outline of the image can be sharpened. Therefore, it is suitable for use in image interpolation such as logos and illustrations, which generally have a small number of gradations.
[0004]
[Problems to be solved by the invention]
  In order to print on the printing device based on the image data, the pattern matching method is conventionally used when the image to be interpolated is considered to be a logo or illustration, and the cubic method is used when the image is considered to be a natural image. is doing. However, it is generally not easy to automatically and accurately determine the nature of an image, that is, a logo or illustration image and a natural image, and if the determination is incorrect, an inappropriate interpolation process may be performed. . Further, the distinction between a logo or illustration and a natural image is not absolute, and there may be an object that is logo-like and has a clear outline in a natural image.
[0005]
  Furthermore, even if processing according to the nature of the image is performed, it is useless to perform interpolation processing that reproduces subtle gradations for natural images, even though low-quality printing is performed during actual printing. It is.
The present invention has been made in view of the above problems, and is an image data interpolation capable of preventing an error in determining an interpolation method based on property determination of an image to be interpolated and performing an accurate interpolation process according to print quality. An object is to provide a program, an image data interpolation method, and an image data interpolation apparatus.
[0006]
[Means for Solving the Problems]
  To achieve the above purposeTheAn image data interpolation program for performing pixel interpolation on image data obtained by expressing a multi-tone image with dot matrix pixels on a computer, the image data acquisition function for acquiring the image data, and the image data A first interpolation processing function for performing interpolation without reducing the degree of pixel change, and a second interpolation processing function for performing interpolation on the image data without impairing the gradation of the image. A first superimposition ratio determination function that determines the property of the image based on reference pixels around the pixel and determines the superimposition ratio of the first and second interpolation processes based on the same property, and the determined superimposition ratio The image data superimposing function for superimposing the image data by the first interpolation processing function and the interpolation image data by the second interpolation processing function, and outputting the superimposed data as data after the interpolation processing. And an image data output function as a to be executed by a computer forAlso good.
[0007]
  the aboveConstitutionIn image processing, when performing pixel interpolation on image data in which an image is expressed in multi-tone with dot matrix pixels by a computer, image data is acquired by an image data acquisition function. Interpolation processing can be performed on the acquired image data by the first interpolation processing function and the second interpolation processing function, and the first interpolation processing function performs pixel change with respect to the image data. Interpolation is performed without reducing the degree, and the second interpolation processing function performs interpolation without impairing the gradation of the image.
[0008]
  The first superimposition ratio determining function determines the characteristics of the image based on reference pixels around the pixel to be interpolated, determines the superimposition ratio of the first and second interpolation processes based on the same characteristics, The superimposing function superimposes the image data by the first interpolation processing function and the interpolated image data by the second interpolation processing function at the superimposition ratio determined by the first superimposition ratio determining function. Then, the data superimposed by the image data output function is output as data after interpolation processing. IeThe secondTwo types of interpolation processing are possible, one interpolation processing function and the second interpolation processing function, and both processed image data are superimposed at a ratio that reflects the nature of the image. Therefore, assuming that the image to be interpolated is either a logo or a natural image, the interpolation method that is not suitable for performing the interpolation of the target image is not executed, and the interpolation method selection is decisively wrong. There is no.
[0009]
  However, the above first superimposition ratio can naturally take not only the ratio in the case of mixing the first and second interpolation processes but also a value of “0” or “1” indicating only one of them.
  Explained aboveConstitutionAccording to the above, since the two interpolation processes are superimposed based on the evaluation function depending on the reference pixel data, it is possible to prevent an error in determining the interpolation method based on the property determination of the interpolation target image.
  Here, in the first interpolation processing function, it is only necessary to be able to execute the interpolation processing without reducing the degree of change of the pixel, and the edge portion is maintained or emphasized without blurring the so-called edge. Applicable interpolation processing is applicable. Further, the second interpolation processing function only needs to be able to execute the interpolation processing without losing the gradation of the image. Interpolation processing that can reproduce the tone value change is applicable. Further, by referring to a plurality of pixels around the pixel to be interpolated in the first superposition ratio determination function, it becomes possible to grasp the change tendency of the pixel in the referred pixel, and the interpolation target image is “natural image”, The image properties such as “Illustration” are revealed. Therefore, it can be said that the reference pixel data reflects the property of the interpolation target image, and the superimposition ratio of the appropriate one of the first interpolation processing and the second interpolation processing is increased according to the property of the interpolation target image. be able to.
[0010]
  There are various interpolation processing methods as described above, and even if two types of interpolation processing are executed as described above, it is not necessary to limit the superposition to only two types of interpolation processing. As an example of the configuration for that,UpThe first interpolation processing function and the second interpolation processing function can select an interpolation processing to be executed from a plurality of types of interpolation processing. In other words, there are multiple types of interpolation that do not reduce the degree of pixel change executed by the first interpolation processing function, and interpolation that does not impair the gradation of the image executed by the second interpolation processing function. There are several types. Therefore, an interpolation process more suitable for the type of the interpolation target image can be performed by selecting a suitable one to be used for the interpolation target image from these or selecting a usable one.
  in this way, ComplementThe range of options for the inter-process is widened, and a more appropriate interpolation process can be executed.
[0011]
  As a specific example of such a configuration,UpThe first interpolation processing function can also execute pattern matching interpolation that performs interpolation according to a predetermined rule when a predetermined pattern exists in the reference pixel and interpolation by the near list method. That is, in pattern matching interpolation, a predetermined pattern is often detected based on the degree of change in pixels. Therefore, interpolation that maintains or emphasizes the degree of change is performed on the detected pattern. Image interpolation can be executed without reducing the degree of change by determining an interpolation rule in advance.
[0012]
  In the near list method, the value of the original pixel closest to the interpolation pixel is used as the value of the interpolation pixel, so that the degree of change is maintained. Therefore, both are systems of the first interpolation processing function, and either one of them can be selected according to the case. Specifically, in the above pattern matching interpolation, the interpolation process itself cannot be executed when a predetermined pattern does not exist in the reference pixel. In this case, the execution of the near list method may be selected.
  in this wayYongInterpolation can be easily performed without reducing the degree of pixel change.
[0013]
  Furthermore, as a specific example of the second interpolation processing function,UpThe second interpolation processing function is configured to be able to execute interpolation by the cubic method. That is, in the interpolation by the cubic method, generally 16 pixels around the interpolation pixel are set as reference pixels, and the interpolation pixel data is reflected while reflecting the gradation value of the reference pixel with the degree of influence depending on the distance between the reference pixel and the interpolation pixel. Is generated. Therefore, according to the cubic method, it is possible to reflect a subtle gradation value change of the reference pixel with respect to the interpolation pixel, and the cubic method is a system of the second interpolation processing function. By superimposing such a cubic method, it is possible to cope with image interpolation such as a multi-gradation natural image.
  in this wayYongInterpolation can be easily performed without impairing the gradation of the image.
[0014]
  One characteristic invention can be extracted in a configuration in which two systems of interpolation processing are superimposed on an interpolation target image. That is, the present invention is an image data interpolation program for performing pixel interpolation on image data in which an image is expressed in multiple gradations by a plurality of pixels by a computer. The image data is acquired and interpolated in the image data. The reference pixel around the pixel to be interpolated is obtained on the condition that the feature amount representing the degree of non-natural image quality of the image is acquired based on the reference pixel around the target pixel and the acquired feature amount exceeds a predetermined value. If it is determined that an edge pattern is present, the above-mentioned rules exist for selecting neighboring pixels used for weighted averaging for generating pixels to be interpolated, which are defined in advance by different rules for each edge pattern. A configuration is adopted in which a peripheral pixel is selected according to a selection rule corresponding to an edge pattern, and pixel interpolation using the selected pixel is performed. .
  In the invention configured as described above, image data is acquired when performing pixel interpolation on image data in which an image is expressed in multiple gradations by a plurality of pixels by a computer. Next, based on reference pixels around the pixel to be interpolated in the image data, a feature amount representing the degree of non-natural image quality of the image is acquired. And when it is determined that an edge pattern exists in the reference pixels around the pixel to be interpolated on the condition that the acquired feature amount exceeds a predetermined value, the rule is determined in advance for each edge pattern, Among the selection rules of the surrounding pixels used for the weighted average for generating the pixel to be interpolated, the surrounding pixels are selected according to the selection rule corresponding to the existing edge pattern, and pixel interpolation using the selected pixel is performed. Do. That is, the superposition ratio determined based on the quality of image quality is also regarded as a feature quantity indicating the degree of the quality of the image quality, and therefore whether or not to execute a predetermined interpolation process is determined according to the value of the feature quantity. If interpolation is performed according to a rule determined in advance for each edge pattern existing in reference pixels around the pixel to be interpolated, interpolation processing suitable for each edge pattern can be performed.
  In another aspect of the present invention, the feature amount may be determined by an evaluation function that depends on the reference pixel data. That is, the evaluation function is a function depending on the data of the reference pixel, and the reference pixel reflects the property of the interpolation target image as described above. Therefore, the feature amount can be easily determined by using an evaluation function that depends on reference pixel data.
  As described above, according to the present invention, the feature amount can be easily calculated.
[0015]
  Above featuresAn example of a specific method for determining,UpBased on the tone value of the reference pixelFeature valueIt is good also as a structure which determines. That is, the gradation value of the reference pixel reflects the property of the interpolation target image, and the function depending on the gradation value is used according to the property of the interpolation target image.Above featuresCan be determined.
  As described above, according to the present invention, it is possible to easily depend on the nature of the interpolation target image.Above featuresCan be determined.
[0016]
  Furthermore, as an example of the configuration in this case, in another aspect of the present invention,,UpBased on the number of appearances of different gradation values in the reference pixelAbove featuresIt is good also as a structure which determines. That is, since images such as logos and illustrations usually have a small number of colors, it is considered that several pixels in the reference pixels have the same gradation value, and the number of pixels having different gradation values is small. On the other hand, since natural images usually have a large number of colors, the number of pixels having different gradation values in the reference pixels is large. In this way, the number of appearances of different gradation values in the reference pixel reflects the nature of the interpolation target image and depends on the nature of the image.Above featuresCan be determined. As a specific example, the number of appearances of different gradation values in the reference pixel is examined.Above featuresIs increased.
  Thus, according to the present invention, it is possible to easily respond to the non-natural image quality.Above featuresCan be determined.
[0017]
  Furthermore, as an example of such a configuration, in another aspect of the present invention,,UpWhen the number of appearances of different gradation values in the reference pixel is smaller than a predetermined threshold valueThe above feature value is the maximum valueIt is good also as a structure. That is, as described above, the number of appearances of the gradation value of the multi-tone image tends to be large and the number of appearances of the gradation value of the low-gradation image tends to be small. Consider it an image such as an illustration or logo,It is always determined whether an edge pattern exists in the reference pixels around the pixel to be interpolated.As a result, images such as illustrations and logosLeaveIt is possible to prevent the outline from being blurred.
  Thus, according to the present invention, it is possible to easily respond to the non-natural image quality.Above featuresCan be determined.
[0018]
  Furthermore, as another example of the configuration, in another aspect of the present invention,,UpThe larger the gradation value width of the reference pixel,Above featuresIt is good also as a structure which enlarges. Here, the gradation value width is the difference between the maximum value and the minimum value of the gradation value, and this gradation value width becomes large when the reference pixel forms a so-called edge. Therefore, the larger the gradation value width,Above featuresThe reference pixel has an edge-like part by increasingIn this case, it is determined whether an edge pattern exists in the reference pixels around the pixel to be interpolated.Specifically, the value of the above evaluation functionAbove featuresAnd the evaluation function can be realized as a monotonically increasing function with respect to the gradation value width.
  As described above, according to the present invention, it is possible to perform image interpolation so as not to reduce the edge as the reference pixel has an edge-like portion.
[0019]
  Furthermore, as a specific example of the gradation value, in another aspect of the present invention, the gradation value of the reference pixel may be a luminance value of the reference pixel. In other words, when an image is expressed with multiple gradations using pixels in a dot matrix, the gradation value data of each color is usually stored for one pixel, and the luminance value of the pixel is determined from the gradation value of each color. By using the value accurately capture the characteristics of one pixel,Above featuresIt can be reflected in the calculation.
  Thus, according to the present invention, the characteristics of one pixel can be accurately reflected in the evaluation function.
[0020]
  By the way, such an image interpolation program can also be applied to a print control program that executes a print control process while executing an interpolation process in a computer in order to input such image data and cause the printing apparatus to print the image data. .
  For this reason,UpA print quality acquisition function for acquiring print quality when printing is performed by the printing apparatus using the recorded image data, and a superposition ratio of the first and second interpolation processes is determined based on the acquired print quality The second superposition ratio determining function and the print control processing function for executing the print control process based on the superposed data may be executed.
[0021]
  the aboveConstitutionIn order to execute the print control process while executing the interpolation process in the computer in order to input the image data expressing the image by the dot matrix pixel and print it by the printing apparatus, the print quality acquisition function is The printing quality when printing is performed by the printing apparatus using the image data is acquired. The second superimposition ratio determination function determines the superimposition ratio of the first and second interpolation processes based on the acquired print quality, and the image data superimposition function is determined by the second superimposition ratio determination function. The image data obtained by the first interpolation processing function and the interpolation image data obtained by the second interpolation processing function are superimposed at a superposition ratio. The print control processing function executes a print control process based on the superimposed data. IeThe secondTwo types of interpolation processing are possible, one interpolation processing function and the second interpolation processing function, and both processed image data are superimposed and printed at a ratio based on print quality. Therefore, a print result can be obtained in a state where an effective interpolation process is performed for each print quality. In addition, assuming that the nature of the interpolation target image is either a logo or a natural image, the interpolation method selection is decisive, without executing only the interpolation processing that is not suitable for interpolation of the target image. Incorrect print results are not obtained.
[0022]
  Of course, in this case as well, the second superposition ratio can take not only the ratio in the case of mixing the first and second interpolation processes, but also a value of “0” or “1” indicating only one of them.
  in this way,twoSince the system interpolation processing is superimposed based on the evaluation function that depends on the reference pixel data, it is possible to prevent an error in determining the interpolation method based on the property determination of the interpolation target image.
  The second superposition ratio determining function determines the superposition ratio based on the print quality. In other words, there is a causal relationship between the print quality and the effect of the interpolation process, such that if the print quality is low, the effect does not appear in the print result even if the interpolation process is executed so as not to impair the delicate gradation. Therefore, by determining the superimposition ratio based on the print quality, the superimposition ratio of the appropriate one of the first interpolation process and the second interpolation process can be increased according to the effect of the interpolation process on the print result. it can. As a specific example, it is conceivable to use an evaluation function depending on the print quality, and the superposition ratio can be easily determined by using the evaluation function.
  in this way, ReviewThe superposition ratio can be easily calculated using the valence function.
[0023]
  In this way, as an example of a configuration for increasing the ratio of those that are likely to have an effect of interpolation processing based on print quality,UpThe second superposition ratio determining function is configured to increase the superposition ratio of the second interpolation processing as the acquired print quality increases. In other words, the second interpolation process interpolates pixels without impairing the gradation, and the subtle gradation change reproduced in the interpolation process is also reproduced in the print result. It is meaningless if the quality is not high. Therefore, the higher the print quality is, the larger the superposition ratio of the second interpolation process is, so that the interpolation process can be superimposed according to the print quality.
[0024]
  in this wayYongInterpolation processing according to print quality can be easily performed.
  Furthermore, there are various modes for determining such a second superposition ratio.,UpThe second superposition ratio determination function is configured to prevent only the first interpolation process from being executed when the acquired print quality is high. That is, in the first interpolation process, the degree of change in pixels is not reduced, but there is a tendency to deteriorate the gradation. The gradation characteristic of the original image is obtained only by performing the first interpolation process during high-quality printing. May be printed with interpolation that does not reproduce the image at all. Thus, by making sure that the superposition ratio is not such that only the first interpolation processing is executed during high-quality printing, the interpolation method selection will not be erroneously determined in relation to the printing result.
  in this way,markThere is no decisive mistake in selecting an interpolation method in relation to the printing result.
[0025]
  Here, the print quality is determined by various parameters, and the print quality acquisition function can acquire the print quality by acquiring the various parameters. For example, when the quality of the printing paper is acquired, the printing quality is acquired. There are various types of printing paper, such as glossy paper and plain paper, and even if printing is performed under the same conditions other than printing paper based on the same image data, the printing results will be very different if the printing paper is different. Because it comes. Therefore, if the print quality is acquired by the print quality acquisition function and the superposition ratio is determined based on the quality of the print paper by the second superposition ratio determination function, the interpolation processing according to the print quality is superimposed. It can be carried out.
[0026]
  in this wayYongInterpolation processing according to print quality can be easily performed.
  Another example of print quality is printing speed. That is, recent printing apparatuses may have a high-speed printing mode that prioritizes printing speed even if the printing image quality is lowered to some extent and a low-speed printing mode that prioritizes high image quality even if the printing speed is slow. Therefore, if such a difference in print speed is acquired, a difference in print quality is acquired. Therefore, the print speed is acquired by the print quality acquisition function, and the second overlap ratio determination function is used to superimpose based on the print speed. If the ratio is determined, interpolation processing according to the print quality can be superimposed.
  in this wayYongInterpolation processing according to the printing speed can be easily performed.
[0027]
  Furthermore, another example of print quality is print resolution. That is, as described above, the printing accuracy of the printing apparatus is extremely high, and during normal printing, interpolation is performed between dots of the original image data. At this time, even if the original image data has the same amount of data and the printed area is the same, the print results are very different due to the difference in print resolution (dpi). Therefore, if the print resolution is acquired, the print quality is acquired, the print resolution is acquired by the print quality acquisition function, and the superimposition ratio is determined based on the print resolution by the second superimposition ratio determination function. Interpolation of interpolation processing according to print quality can be performed.
  in this wayYongInterpolation processing according to the printing resolution can be easily performed.
[0028]
  Furthermore, as another example of the print quality, there is a type of ink. That is, the ink used in the printing apparatus includes pigments and dyes, and the pigments are less likely to bleed when compared with each other. Therefore, when printing on a printing paper, the so-called edge tends to be more conspicuous in the pigment, and the print quality differs if the type of ink is different. Therefore, if the type of ink is acquired by the print quality acquisition function and the first interpolation processing ratio is smaller in the second superposition ratio determination function when the pigment is used, the print quality can be reduced. It is possible to superimpose interpolation processing according to the above.
  in this wayYongInterpolation processing according to the type of ink can be easily performed.
[0029]
  Needless to say, the present invention is not only established as a program itself but also as a recording medium thereof. Such a recording medium may be a magnetic recording medium, a magneto-optical recording medium, or any recording medium that will be developed in the future. In addition, the duplication stages such as the primary duplication product and the secondary duplication product are equivalent without any question. Further, even when a part is software and a part is realized by hardware, the idea of the invention is not completely different, and a part is stored on a recording medium and is appropriately changed as necessary. It may be in the form of being read.
[0030]
  in this way,Image data interpolation according to the present inventionWhen the program proceeds according to such control, it is natural that the invention exists in the procedure at the bottom, and it can be easily understood that the program can be applied as a method. For this reason,Claims 7 to 12In the invention according to this embodiment, the same operation is basically performed. In other words, the present invention is not necessarily limited to a tangible medium, and there is no difference that the method is effective.
[0031]
  In addition, it can be easily understood that such a program is realized in an actual computer, and that the present invention can be applied as an apparatus including such a computer. For this reason,Claims 13 to 18In the invention according to this embodiment, the same operation is basically performed. Of course, such an apparatus may be implemented alone, or may be implemented together with other methods in a state of being incorporated in a certain device. Can be changed as appropriate.
[0032]
【The invention's effect】
  As explained aboveSecondSince the system interpolation processing is superimposed based on the evaluation function that depends on the reference pixel data, it is possible to prevent an error in determining the interpolation method based on the property determination of the interpolation target image.
  AlsoYongInterpolation can be easily performed without reducing the degree of pixel change.
  Also,According to the first, seventh, and thirteenth aspects of the invention, it is possible to execute an interpolation process suitable for each edge pattern.
  Also,According to the inventions of Claims 2, 8, and 14, the feature amount is determined based on the evaluation function depending on the data of the reference pixel.It is possible to prevent an error in determining an interpolation method based on property determination of an interpolation target image.Further, the feature amount can be easily calculated.
[0033]
  further,According to the third, ninth, and fifteenth aspects of the present invention, the feature amount can be easily determined according to the non-natural image quality.Can be determined.
  further,According to the invention of claim 4, claim 10, or claim 16, the feature amount can be easily determined according to non-natural image quality.Can be determined.
[0034]
  further,Claims 5, 11, and 17According to the invention, the reference pixel has an edge-like portionIn this case, it can be determined whether or not an edge pattern exists in the reference pixels around the pixel to be interpolated.
  further,Claims 6, 12, and 18According to this invention, the feature of one pixel can be accurately reflected in the evaluation function.
  further,twoSince the system interpolation processing is superimposed based on the evaluation function that depends on the reference pixel data, it is possible to prevent an error in determining the interpolation method based on the property determination of the interpolation target image.
[0035]
  further, ReviewThe superposition ratio can be easily calculated using the valence function.
  furtherYongInterpolation processing according to print quality can be easily performed.
  further,markThere is no decisive mistake in selecting an interpolation method in relation to the printing result.
[0036]
DETAILED DESCRIPTION OF THE INVENTION
  Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing the main configuration of the image data interpolation apparatus of the present invention.
Assuming digital processing, an image is expressed by a dot matrix pixel, and image data is composed of a collection of data representing each pixel. In a system where processing is performed in units of pixels, image enlargement / reduction is performed in units of pixels. The present image data interpolation apparatus performs such a pixel-by-pixel enlargement process, and the image data acquisition means C11 acquires such image data, and the first interpolation processing means C12 and the second interpolation. The processing means C13 performs an interpolation process for increasing the number of constituent pixels in the image data. Here, the first interpolation processing means C12 can execute the pattern matching method and the near list method as the interpolation processing, and the second interpolation processing means C13 can execute the cubic method as the interpolation processing. .
[0037]
  The first superimposition ratio determining unit C14 determines the property of the image based on the reference pixels around the pixel to be interpolated in the image data acquired by the image data acquiring unit C11, and the first interpolation processing based on the property. A superimposition ratio of interpolation pixels by means C12 and second interpolation processing means C13 is determined. When the first superimposition ratio determining means C14 determines the superposition ratio, the image data superimposing means C15 superimposes the interpolated pixel data by the first interpolation processing means C12 and the second interpolation processing means C13 at the superimposition ratio. The image data output means C16 outputs the pixel data superimposed by the image data superimposing means C15.
[0038]
  In the present embodiment, a computer system 10 is employed as an example of hardware that realizes such an image data interpolation apparatus.
FIG. 2 is a block diagram showing the computer system 10.
The computer system 10 includes a scanner 11a, a digital still camera 11b, and a video camera 11c as image input devices, and is connected to a computer main body 12. Each input device can generate image data in which an image is expressed by a dot matrix pixel and output the image data to the computer main body 12. Here, the image data is displayed in 256 gradations for each of the three primary colors of RGB. About 16.7 million colors can be expressed.
[0039]
  A floppy (R) disk drive 13a, a hard disk 13b, and a CD-ROM drive 13c as external auxiliary storage devices are connected to the computer main body 12, and main programs related to the system are recorded on the hard disk 13b. Necessary programs and the like can be read from the floppy (R) disk 13a1, the CD-ROM 13c1, and the like.
  In addition, a modem 14a is connected as a communication device for connecting the computer main body 12 to an external network or the like, and it can be connected to an external network via the public communication line, and can be installed by downloading software and data. It has become. In this example, the modem 14a accesses the outside via a telephone line, but it is also possible to adopt a configuration where the network is accessed via a LAN adapter. In addition, a keyboard 15a and a mouse 15b are also connected for operating the computer main body 12.
[0040]
  Furthermore, a display 17a and a color printer 17b are provided as image output devices. The display 17a has a display area of 800 pixels in the horizontal direction and 600 pixels in the vertical direction, and the above-described display of 16.7 million colors can be performed for each pixel. Of course, this resolution is only an example, and can be changed as appropriate, such as 640 × 480 pixels or 1024 × 768 pixels.
[0041]
  The color printer 17b is an ink jet printer, and can print an image by adding dots on a printing paper as a recording medium using four color inks of CMYK. The image density is capable of high-density printing such as 360 × 360 DPI or 720 × 720 DPI, but the gradation limit is a two-gradation expression such as whether or not color ink is applied.
  On the other hand, in order to display or output to the image output device while inputting an image using such an image input device, a predetermined program is executed in the computer main body 12. Among them, an operating system (OS) 12a is operating as a basic program, and the operating system 12a performs a print output to a display driver (DSP DRV) 12b for displaying on the display 17a and a color printer 17b. A printer driver (PRT DRV) 12c is installed. These types of drivers 12b and 12c depend on the models of the display 17a and the color printer 17b, and can be added to and changed from the operating system 12a according to the respective models. In addition, depending on the model, additional functions beyond standard processing can be realized. That is, various additional processes within an allowable range can be realized while maintaining a common processing system on the standard system called the operating system 12a.
[0042]
  Of course, the computer main body 12 is provided with a CPU 12e, a RAM 12f, a ROM 12g, an I / O 12h, etc. as a premise for executing such a program, and the CPU 12e for executing arithmetic processing sets the RAM 12f in a temporary work area or setting. A basic program written in the ROM 12g is appropriately executed while being used as a storage area or a program area, and external devices and internal devices connected via the I / O 12h are controlled.
[0043]
  Here, the application 12d is executed on the operating system 12a as a basic program. The processing contents of the application 12d are various. The operation of the keyboard 15a and the mouse 15b as operation devices is monitored, and when operated, various external devices are appropriately controlled to execute corresponding arithmetic processing, Furthermore, the processing result is displayed on the display 17a or output to the color printer 17b.
[0044]
  In such a computer system 10, image data is acquired by a scanner 11 a as an image input device, and predetermined image processing is executed by an application 12 d, and then displayed on a display 17 a or a color printer 17 b as an image output device. Is possible. In this case, simply focusing on the correspondence between the pixels, when the pixel density in the color printer 17b and the pixel density of the scanner 11a match, the size of the scanned original image and the size of the printed image match. If there is a deviation, the size of the image will be different. In the case of the scanner 11a, there are many things that approximate the pixel density of the color printer 17b, but the pixel density of the color printer 17b in which the pixel density is improved in order to improve the image quality is a general image input device. It is often higher than the pixel density. In particular, the display density is higher than the display density of the display 17a, and if the display on the display 17a is made to coincide with each other on a pixel basis and printed, an extremely small image may be obtained.
[0045]
  Therefore, resolution conversion is performed in order to eliminate the difference in pixel density for each actual device while determining the reference pixel density in the operating system 12a. For example, when the resolution of the display 17a is 72 DPI, if the operating system 12a uses 360 DPI as a reference, the display driver 12b performs resolution conversion between the two. In the same situation, if the resolution of the color printer 17b is 720 DPI, the printer driver 12c performs resolution conversion.
[0046]
  The resolution conversion corresponds to an interpolation process because it is a process of increasing the number of constituent pixels in the image data, and the display driver 12b and the printer driver 12c implement the interpolation process as one of its functions. Here, the display driver 12b and the printer driver 12c can execute the first interpolation processing means C12 and the second interpolation processing means C13 described above, and further, the first superimposition ratio determining means C14, the image data superimposing means C15, and the image data. The data output means C16 is executed to superimpose ratios according to the properties of the image.
[0047]
  The display driver 12b and the printer driver 12c are stored in the hard disk 13b and are read and operated by the computer main body 12 at the time of activation. At the time of introduction, it is recorded on a medium such as the CD-ROM 13c1 or the floppy (R) disk 13a1 and installed. Therefore, these media constitute a medium in which an image data interpolation program is recorded.
  In the present embodiment, the image data interpolation apparatus is realized as the computer system 10, but such a computer system is not necessarily required, and any system that requires interpolation processing for similar image data may be used. . For example, as shown in FIG. 3, an image data interpolation device for performing interpolation processing is incorporated in the digital still camera 11b1, and the interpolated image data is displayed on the display 17a1 or printed on the color printer 17b1. Also good. As shown in FIG. 4, in the color printer 17b2 that inputs and prints image data without using a computer system, the image data input through the scanner 11a2, the digital still camera 11b2, the modem 14a2, or the like is automatically processed. It is also possible to perform a print process after performing resolution conversion.
In addition, the present invention can naturally be applied to various apparatuses that handle image data such as a color facsimile apparatus 18a as shown in FIG. 5 and a color copy apparatus 18b as shown in FIG.
[0048]
  FIG. 7 shows a software flow related to resolution conversion executed by the printer driver 12c described above. In FIG. 7, in step S102, original image data is acquired. When the application 12d reads an image from the scanner 11a, performs predetermined image processing, and performs print processing, print data with a predetermined resolution is acquired by the printer driver 12c via the operating system 12a. . This process is an image data acquisition process or an image data acquisition function when viewed as software. The various steps executed by the computer including the image data acquisition step are performed directly on the operating system 12a itself or hardware. It can be understood as not including. On the other hand, when it is considered that it is organically coupled with hardware such as a CPU, it corresponds to the image data acquisition means C11.
[0049]
  In step S104, in the read image data, the pixel to be interpolated is set as the target pixel, and the pixel data of the peripheral 5 × 5 pixel region around the target pixel is set as the reference pixel, and the histogram of the luminance value of the reference pixel is displayed. create. In step S106, the number of times that different luminance values appear in the obtained histogram is acquired, and it is determined whether or not the number of appearances is smaller than 15. As the number of different luminance values increases, the number of colors in the reference pixel increases. Therefore, here, a pixel whose luminance value appears less than 15 is regarded as a non-natural image, and the number of appearances is smaller than 15 in step S106. When it is determined, the superposition ratio (rate) of the first interpolation process is set to 1 in step S110.
[0050]
  When it is determined in step S106 that the number of appearances of the luminance value is not less than 15, rate is determined by the evaluation function F in step S108. Although details will be described later, the evaluation function F is a function of the luminance value width of the reference pixel, that is, the difference between the maximum luminance value Ymax and the minimum luminance value Ymin in the reference pixel. Therefore, the series of processing shown in steps S104, S106, S108, and S110 corresponds to the first superimposition ratio determining step or the first superimposition ratio determining function, and these are organically combined with hardware such as a CPU. Considering this, the first superposition ratio determining means C14 is configured.
[0051]
  When the rate is determined in step S108 or step S110, it is determined whether or not the rate is “0” in step S112. When the rate is “0”, the first interpolation process is not performed. ing. If it is not determined in step S112 that the rate is "0", the first interpolation process is executed in step S114. Accordingly, this processing corresponds to the first interpolation processing step or the first interpolation processing function, and the first interpolation processing means C12 is considered to be integrally combined with hardware such as a CPU. Will be composed.
[0052]
  Further, in step S116, it is determined whether or not the rate is “1”. When the rate is “1”, the second interpolation processing is not performed. If it is not determined in step S116 that the rate is "1", the second interpolation process is executed in step S118. Therefore, this processing corresponds to the second interpolation processing step or the second interpolation processing function, and considering that these are organically combined with hardware such as a CPU, the second interpolation processing means C13 is Will be composed. After determining the rate in this way and executing the first interpolation process or the second interpolation process, in step S120, based on the rate, pixel data is superimposed by the following formula to generate an interpolation pixel.
Superimposition data =
(First interpolation processing) × rate + (second interpolation processing) × (1-rate)
                                                  ... (1)
  Therefore, if this process corresponds to the image data superimposing step or the image data superimposing function and these are considered to be organically combined with hardware such as a CPU, the image data superimposing means C15 is configured. When the generation of the interpolation data for the target pixel is completed in this way, it is determined in step S122 whether or not the superimposition processing for all the target pixels of the input original image data is completed, and the process for all the target pixels is performed. The process after step S104 is repeated until the process ends. When the superimposition process for all the target pixels is completed, the interpolated image data is output in step S124. In the case of the printer driver 12c, print data is not obtained only by resolution conversion, but color conversion or halftone processing is necessary. Accordingly, outputting image data here means delivery of data to the next stage, and this processing corresponds to the image data output step or image data output function, which is integrated with hardware such as a CPU. Therefore, the image data output means C16 is configured.
[0053]
  Next, the first interpolation processing flow in step S114 of the above flow will be described. FIG. 8 shows the first interpolation processing flow. In FIG. 8, in step S202, in order to perform pattern matching, the pixel of interest is further extracted from the pixel data of the 5 × 5 pixel area acquired in step S104. Pixel data of a region of 3 × 3 pixels around the center is extracted. In step S204, the pixel data is converted into a binary pattern depending on whether or not the luminance value of the extracted pixel is larger than a predetermined threshold value described later.
[0054]
  In step S206, it is determined whether or not the binary pattern matches the predetermined edge pattern prepared in advance in step S204. Here, the edge pattern is a binary pattern when the 3 × 3 pixel region has an edge of “30 °”, “45 °”, etc., and the edge pattern and the 2 created in step S204 above. By determining whether or not the value pattern matches, it is determined whether or not the 3 × 3 pixel region is an edge.
[0055]
  If it is determined in step S206 that both patterns match, an interpolated pixel is generated in accordance with a rule determined in advance in accordance with the edge pattern in step S208. In the flow, the process interpolation process at the time of pattern matching is expressed as a pattern matching method. If it is not determined in step S206 that both patterns match, an interpolation pixel is generated by the near list method in step S210. When the interpolation processing for the target pixel is executed in this way, the data is stored in the RAM 12f in step S212, and the flow returns to the flow shown in FIG.
[0056]
  Next, more specific processing for the above flow will be described. In the present embodiment, it is determined whether the original image acquired in step S102 is computer graphics (non-natural image) or a photograph (natural image), and the determination result is reflected in the superposition ratio. . Of course, in order to prevent an error in determining the interpolation method based on the property determination of the interpolation target image, a configuration in which the rate is not fixed to “1” or “0” is possible. If it is smaller than 15, the image is first considered to be a non-natural image, and emphasis is placed on executing the second interpolation process on the non-natural image so as not to obscure the contour of the image.
[0057]
  Specifically, this determination uses a histogram of luminance values shown in FIG. 9. In step 104, the luminance value is obtained for each reference pixel in the 5 × 5 pixel area, and the number of pixels in the range that the luminance can take. Aggregate histograms. Then, the number of appearances of the luminance value, that is, how many luminance values whose distribution number is not “0” is counted, and the determination in step S106 is performed. Of course, it is natural that there are multiple colors with the same brightness among 16.7 million colors in one image. If you focus only on the comparison with a non-natural image, is there a lot of color or brightness? It is possible to compare whether the number is small or not, and it is also possible to determine whether or not the image is a non-natural image by counting the number of appearances of colors in addition to the luminance value.
[0058]
  If the pixel data for the reference pixel has luminance as a component element, the distribution can be obtained using the luminance value. However, even in the case of image data in which the luminance value is not a direct component value, the component value that indirectly represents the luminance is provided. Accordingly, the luminance value can be obtained by performing conversion from the color space where the luminance value is not a direct component value to the color space where the luminance value is a direct component value.
[0059]
  The color conversion between different color spaces is not uniquely determined by the conversion formula, but the color conversion with each component value as a coordinate is obtained and the color conversion is stored. It is necessary to perform sequential conversion with reference to the table. Then, strictly speaking, it is necessary to have a color conversion table of 16.7 million elements. As a result of considering the efficient use of storage resources, instead of preparing correspondences for all coordinate values, we usually prepare correspondences for appropriate discrete grid points and use interpolation operations together. Like to do. However, since such an interpolation calculation is possible through several multiplications and additions, the amount of calculation processing becomes enormous.
[0060]
  That is, if a full-size color conversion table is used, the processing amount is reduced, but the table size becomes an unrealistic problem, and if the table size is made a realistic size, the calculation processing amount becomes unrealistic. There are many cases. In view of such a situation, a simplified luminance value calculation method is employed in the present embodiment. That is, the interpolation target image data of the present embodiment employs the RGB color space, and each component value indicates the brightness of the color. Therefore, when each component value is viewed independently, the luminance is linear. There is a nature that corresponds to. Therefore, without considering the addition ratio of each color,
Y = (R + G + B) / 3
However, in this embodiment, the value value is calculated by the equation.
[0061]
  Of course, the calculation of the luminance value is not limited to such a method, and a conversion formula of the following expression for obtaining the luminance from the three primary colors of RGB may be adopted as used in the case of television or the like.
Y = 0.30R + 0.59G + 0.11B
In this way, the luminance value can be obtained by only three multiplications and two additions.
  In such a luminance value histogram, if there are more than 15 luminance values whose distribution number is not “0”, the superimposition ratio rate is determined by the evaluation function F previously determined in step S108. The solid line in FIG. 10 shows a specific example of the evaluation function F (y). The evaluation function has a value in a range where “y” is “0 ≦ y ≦ 255”, and “0 ≦ F (y)”. It fluctuates in the range of ≦ 1 ”. Further, “F (y) = 0” in the range of “0 ≦ y ≦ 64”, “F (y) = 1” in the range of “192 ≦ y ≦ 255”, and “64 ≦ y ≦ 192”. In the range “”, F (y) increases linearly from “0” to “1”. Here, the luminance value width “Ymax−Ymin” in FIG. 9 is substituted for “y”. Therefore, the rate increases as the luminance value width increases, that is, as the relationship between the reference pixels is more likely to be an edge.
[0062]
  In addition, the function for increasing the rate as the edge appears is not limited to the example shown in FIG. 10, and any function that increases monotonously with the luminance value width when the luminance value width is a variable may be used. Further, it is not always necessary to set the maximum value of the evaluation function to “1” and the minimum value to “0”, and the first value of the evaluation function is set to “0.7” as shown by the dotted line in FIG. It is also possible to prevent the execution of only the interpolation process, or to employ an evaluation function that changes smoothly over the luminance value range “0” to “255”.
[0063]
  In the case of a histogram as shown in FIG. 9, since there are 19 luminance values whose distribution number is not “0”, the rate is determined by the evaluation function shown in FIG. In S118, the first interpolation process and the second interpolation process are executed. FIG. 11A shows an example of a 3 × 3 pixel luminance pattern extracted in step S202 of FIG. In step S204, the binary pattern shown in FIG. 6B is created from the brightness pattern shown in FIG. 6A. First, the brightness value Yij is calculated for each pixel Pij based on the above brightness value calculation formula. To do.
[0064]
  Then, an intermediate value between the maximum value Ymax and the minimum value Ymin of these luminance values is obtained, and the intermediate value is set as the threshold value Yt. After the threshold value is calculated, each luminance value Yij is compared with the threshold value Yt, and the pixel value is converted into a binary pixel pattern by dividing the luminance value into those having a luminance value larger and smaller than the threshold value. Of course, the method of selecting the threshold value and the method of creating the binary pattern need not be limited to the above-described mode. If the intermediate value 128 of the luminance value is set as the threshold value or the minimum value of the luminance value is “45” or less, Various forms such as handling the minimum value as “45” by not using such an extreme luminance value can be adopted.
[0065]
  When the binary pattern shown in FIG. 11B is created, it is determined in step S206 whether or not the binary pattern matches an edge pattern prepared in advance. Here, various patterns can be adopted as the edge pattern prepared in advance. Specifically, the pattern shown in FIG. 11B matches an edge pattern prepared in advance. Here, the coincidence with the edge pattern in FIG. 5B means that the luminance value of the pixel located on the upper side in the j direction and the luminance value of the pixel located on the lower side in the j direction in FIG. This shows that the value difference tends to be large, and that there is not much difference in luminance value in the left-right direction on the paper. Accordingly, such a pattern is an edge parallel to the horizontal direction of the paper. Therefore, pixel interpolation processing is performed on the edge pattern according to a predetermined rule.
[0066]
  The predetermined rule includes various rules depending on the edge pattern. In general, the distance between the pixel and the pixel to be interpolated with respect to the gradation value of the predetermined pixel in the 3 × 3 pixel region is roughly described. The weighted average with the inverse ratio of is weighted is calculated. Here, various variations can be considered in terms of how to select a pixel for performing weighted averaging when generating an interpolation pixel, and how to reflect the angle, direction, etc. of the edge, Pixel interpolation processing is performed according to a rule predetermined for the edge pattern in the sense that a predetermined variation is associated with each edge pattern.
[0067]
  The edge pattern is a 3 × 3 pixel region, but the 5 × 5 pixel reference pixel is also used to determine the angle and direction of the edge. That is, the edge pattern needs to have various binary patterns as a database in advance, but if a predetermined pattern is prepared in advance in an area of 5 × 5 pixels, the number of the patterns becomes enormous. Therefore, in the present embodiment, a binary pattern is generated for a 3 × 3 pixel region in which the number of edge patterns prepared in advance is a realistic number, and whether or not the generated pattern matches the edge pattern. In order to further reflect the nature of the edge, a 5 × 5 pixel region is referred to.
[0068]
  Hereinafter, a pixel interpolation processing rule when it is determined that the binary pattern shown in FIG. FIG. 12A shows an area of 5 × 5 pixels as the reference pixel data, and the individual pixels are distinguished by attaching alphabets A to Y to the pixels. In this notation, the 3 × 3 pixel region used for the pattern matching is a pixel “G, H, I, L, M, N, Q, R, S”, and the pixel of interest is a pixel M. FIG. 12A shows a state after the binary pattern has already been formed in the 3 × 3 pixel region.
[0069]
  FIG. 12B shows the pixel after the interpolation process, and the target pixel M is interpolated into pixels a to i. The basic formula of calculation at the time of pixel generation is as shown in the following formula, and a weighted average is performed with the inverse ratio of the distance from the interpolation pixel to the original pixel as a weight. Here, the gradation value of the generated interpolation pixel is xbar, Pn is the gradation value of the original pixel (G, H, I, L, M, N, Q, R, S) to be referred to, and rn Is the distance from the generated interpolated pixel to each original pixel referenced.
[0070]
[Expression 1]

  The point that the gradation value of the interpolation pixel generated in this way is generated from the original pixel is a characteristic point. That is, binarization is performed on a pixel having a multi-gradation value, and pattern matching is tried on that. When pattern matching is performed, the gradation value of the interpolated pixel is generated from the gradation value of the surrounding pixels using a calculation formula corresponding to the matched pattern. In this way, while performing pattern matching determination, the generated pixels are generated from the surrounding original pixels based on the rules prepared based on the matched pattern, and therefore can be obtained by simple interpolation processing based on pattern matching. Not natural interpolation pixels can be generated. In addition, although the calculation which calculates | requires a brightness | luminance from each RGB value by the previous calculation was performed, the gradation value of the interpolation pixel to be generated actually has each RGB value, and for each value, the original pixel Calculation is performed by performing predetermined weighting using RGB values. Therefore, in a case where calculation is convenient, weighting may be derived in a state where luminance is used, and then calculation may be performed by applying the same weighting to RGB values.
[0071]
  As described above, the 3 × 3 pixel pattern shown in this example has little difference in luminance value in the horizontal direction (i direction), that is, an edge in the horizontal direction. Only the pixels are substituted, and in principle, the gradation values of the original pixels L, M, and N are substituted. Further, in the present embodiment, interpolation is performed to reflect a change in luminance value in a direction perpendicular to the horizontal edge, and the luminance value changes in a direction perpendicular to the horizontal edge. In this case, the gradation values of the original pixels Q, R, and S are substituted into Pn, and the state of the edge change is reflected in the gradation values of the interpolation pixels g, h, and i. That is, if the value of the pixel of interest M is larger (smaller) than the pixel R on the edge side, the pixel of interest will not be emphasized even if it is filled with the pixel R, so that the pixel of interest is used as it is.
[0072]
  Therefore, the luminance values of the pixel M and the pixel R are compared before the interpolation calculation, and if the luminance value of the pixel R is larger than the luminance value of the pixel M, the gradation values of the original pixels Q, R, and S Is substituted into the above equation to generate interpolated pixels g, h, i. As a result, the luminance values of the interpolated pixels g, h, i are larger than the interpolated pixels d, e, f, and the edges are more emphasized in the interpolated image data.
  The specific judgment conditions in this case are:
((M = edge) & (R> M)) or ((M ≠ edge) & (R <M))
And the calculation formula at that time is
a = d = (Lx2 + Mx2 + N) / 5
b = e = (L + Mx2 + N) / 4
c = f = (L + Mx2 + Nx2) / 5
g = (Qx4 + Rx6 + Sx3) / 13
h = (Q + Rx2 + S) / 4
i = (Qx3 + Rx6 + Sx4) / 13
It becomes. On the other hand, when the luminance value of the pixel R is smaller than the luminance value of the pixel M, only the gradation values of the original pixels L, M, and N are used without performing enhancement processing.
[0073]
  The specific judgment conditions in this case are:
not (((M = edge) & (R> M)) or ((M ≠ edge) & (R <M))).
a = d = g = (Lx2 + Mx2 + N) / 5
b = e = h = (L + Mx2 + N) / 4
c = f = i = (L + Mx2 + Nx2) / 5
It becomes. In calculating the gradation value of the interpolated pixel using the above formula, if the inverse ratio of the distance is used as the weight when the original pixel M is substituted for Pn, the value becomes too large. It is supposed to be.
[0074]
  In the present embodiment, the interpolation processing is executed with respect to the horizontal edge in consideration of the change in the luminance value in the vertical direction according to the above rules. However, in addition to the horizontal edge, various edge patterns are processed. FIG. 13 shows an example. When the pattern shown in FIG. 9A is found in step S206, an interpolation pixel shown in FIG. 5B is generated with reference to a predetermined pixel outside the 3 × 3 pixel area. According to the rule in the same pattern, the original pixels L, M, and R are used in the calculation of the gradation values of the interpolated pixels a, d, e, g, h, and i.
[0075]
  The gradation value calculation of the interpolation pixels b, c, and f changes which pixel is the original pixel based on the luminance value of a predetermined pixel included in the 5 × 5 pixel area. That is, if the pattern is a right-angled edge, the interpolated pixels b, c, and f are generated using the original pixels L, M, and R. If the pattern is not a right-angled edge, the original pixels G, H, N, and S are used. To generate interpolation pixels b, c, and f.
  First, the judgment condition that this pattern is a right angle is:
(((M = K = edge) & (F = not edge)) or ((M = W = edge) & (X = not edge)))
            or
(((M = K = not an edge) & (F = edge)) or ((M = W = not an edge) & (X = edge)))
And the calculation formula at that time is
c = (L + Mx2 + R) / 4
b = (Lx4 + Mx6 + Rx3) / 13
f = (Lx3 + Mx6 + Rx4) / 13
a = (Lx2 + Mx2 + R) / 5
d = (Lx3 + Mx3 + Rx2) / 8
e = (L + Mx2 + R) / 4
g = (L + M + R) / 3
h = (Lx2 + Mx3 + Rx3) / 8
i = (L + Mx2 + Rx2) / 5
It becomes. Next, the condition judgment as the first pattern close to a right angle other than this right angle is
(((M = F = edge) & (A = not edge)) or ((M = X = edge) & (Y = not edge)))
            or
(((M = F = not an edge) & (A = edge)) or ((M = X = not an edge) & (Y = edge)))
And the calculation formula at that time is
c = (G + Hx2 + Nx2 + S) / 6
b = (Lx4 + Mx6 + Rx3) / 13
f = (Lx3 + Mx6 + Rx4) / 13
And a, d, e, g, h, i are the same calculation formula. And the calculation formula when neither of the above two condition judgments is satisfied is as follows:
c = (G + Hx2 + Nx2 + S) / 6
b = (Gx15 + Hx30 + Nx20 + Sx12) / 77
f = (Gx12 + Hx20 + Nx30 + Sx15) / 77
In this case, the same calculation formula is used for a, d, e, g, h, and i.
[0076]
  Further, when the pattern shown in FIG. 13C is found, an interpolation pixel shown in FIG. 13D is generated. That is, since the pattern shown in FIG. 5C seems to be a part of a straight line connecting the diagonal lines of the 3 × 3 pixel region (edge of 45 degrees), the original pixels G, M, and S constituting these straight lines are used. Interpolated pixels a, e, i are generated. Further, the edge angle and the line thickness of the original pixel are determined based on the luminance value of a predetermined pixel included in the 5 × 5 pixel region, and the interpolation pixels b, c, f, d, g, The original pixel used for generating h is changed between “G, M, S” and “H, N”, or changed between “G, M, S” and “L, R”.
[0077]
  The processing branches into three depending on whether the detected 45-degree edge, part of the 45-degree edge, or part of the 30-degree edge. This judgment condition is shown in FIG. In addition, the description enclosed with the broken line in the figure is the content intended by each determination.
  The calculation formula when process 1 is selected is
b = (Gx5 + Mx10 + Sx4) / 19
c = (G + Mx2 + S) / 4
f = (Gx4 + Mx10 + Sx5) / 19
And the calculation formula when process 2 is selected is
b = (Gx5 + Mx10 + Sx4) / 19
c = (H + N) / 2
f = (Gx4 + Mx10 + Sx5) / 19
The calculation formula when process 3 is selected is
b = (Hx3 + Nx2) / 5
c = (H + N) / 2
f = (Hx2 + Nx3) / 5
It becomes. The calculation formula for other pixels is as follows:
a = (Gx3 + Mx6 + Sx2) / 11
d = (Gx5 + Mx10 + Sx4) / 19
e = (G + Mx2 + S) / 4
g = (G + Mx2 + R) / 4
h = (Gx4 + Mx10 + Sx5) / 19
i = (Gx2 + Mx6 + Sx3) / 11
It becomes.
[0078]
  When the pattern shown in FIG. 13E is found, the interpolated pixel shown in FIG. 13F is generated. That is, it seems to be a straight line (an edge of 30 degrees) having a shallower angle than the straight line shown in FIG. 3C, and therefore, the interpolation pixels a, b, c, e and f are generated. Further, based on the luminance value of a predetermined pixel included in the 5 × 5 pixel area, the edge angle of the original pixel, the protruding pixel, and the like are discriminated and used to generate the interpolated pixels d, g, h, i according to the discrimination. The original pixel to be changed is changed between “G, M, N” and “L, R, S”.
[0079]
  More specifically, the processing branches into four for the detected 30-degree edge. This determination condition is shown in FIG. In addition, the description enclosed with the broken line in the figure is the content intended by each determination.
  The calculation formula when process 1 is selected is
d = (G + Mx2 + N) / 4
g = (G + Mx2 + S) / 4
h = (Gx3 + Mx8 + Nx5) / 16
i = (G + Mx3 + Nx3) / 7
And the calculation formula when process 2 is selected is
d = (G + Mx2 + N) / 4
g = (Lx2 + Rx2 + S) / 5
h = (Lx4 + Rx6 + Sx3) / 13
i = (Lx3 + Rx6 + Sx4) / 13
The calculation formula when process 3 is selected is
d = (G + Mx2 + N) / 4
g = (Lx2 + Rx2 + S) / 5
h = (Lx4 + Rx6 + Sx3) / 13
i = (G + Mx3 + Nx3) / 7
The calculation formula when process 4 is selected is
d = (Lx15 + Rx10 + Sx6) / 31
g = (Lx2 + Rx2 + S) / 5
h = (Lx4 + Rx6 + Sx3) / 13
i = (Lx3 + Rx6 + Sx4) / 13
It becomes. The calculation formula for other pixels is as follows:
a = (Gx4 + Mx6 + Nx3) / 13
b = (Gx3 + Mx6 + Nx4) / 13
c = (G + Mx2 + Nx2) / 5
e = (Gx4 + Mx6 + Nx3) / 13
f = (Gx2 + Mx5 + Nx5) / 12
It becomes.
[0080]
  The above-described edge pattern, 5 × 5 pixel area reference method, original pixel selection method to be used, etc. are examples, and various modes can be adopted, and the magnification for generating an interpolation pixel is also described above. As described above, various magnifications such as 2 times and 4 times can be adopted in addition to a mode in which pixels are three times in length and width.
  First, an example of double interpolation will be described. 38 to 42 show the correspondence when the interpolation pixel a to the interpolation pixel d are generated from the original pixel M, and FIG. 38 shows the case where the horizontal edge is found by pattern matching. Also in this case, if the value of the target pixel M is larger (smaller) than the pixel R on the edge side, the target pixel is used as it is because the pixel R does not enhance. More specific judgment conditions are
((M = edge) & (R> M)) or ((M = not edge) & (R <M))
If this is true, enhancement processing is performed. In that case, the formula is
a = b = M
c = d = R
In other cases, the emphasis process is not performed.
a = b = c = d = M
It becomes.
[0081]
  Next, FIG. 39 shows a case where a corner edge is found by pattern matching.
(((M = K = edge) & (F = not edge)) or ((M = W = edge) & (X = not edge)))
                              or (((M = K = not an edge) & (F = edge)) or ((M = W = not an edge) & (X = edge)))
To emphasize that if this is true, then
b = (L + R) / 2
If this is not the case,
b = (Gx2 + Hx3 + Nx3 + Sx2) / 10
Calculate with the following formula. The calculation formula for other pixels is
a = (Lx3 + Rx2) / 5
c = (L + R) / 2
d = (Lx2 + Rx3) / 5
It becomes.
[0082]
  FIG. 40 shows a case where an edge of 45 degrees is found by pattern matching. In order to determine whether it is a part of a 45 degree edge or a part of a 30 degree edge, the condition judgment shown in FIG. Branches into two processes. Here, the calculation formula in the case of processing 1 is
b = (G + S) / 2
And the formula for processing 2 is
b = (H + N) / 2
It becomes. The calculation formula of other pixels is
a = (Gx2 + S) / 3
c = (G + S) / 2
d = (G + Sx2) / 3
It becomes.
[0083]
  FIG. 42 shows a case where an edge of 30 degrees is found by pattern matching, and branches to three processes according to the condition judgment shown in FIG. Here, the calculation formula in the case of processing 1 is
c = (G + N) / 2
d = (G + Nx2) / 3
And the formula for processing 2 is
c = (Lx3 + Rx3 + Sx2) / 8
d = (Lx2 + Rx3 + Sx3) / 8
And the formula for processing 3 is
c = (Lx3 + Rx3 + Sx2) / 8
d = (G + Nx2) / 3
It becomes. The calculation formula of other pixels is
a = (Gx3 + Nx2) / 5
b = (Gx2 + Nx3) / 5
It becomes.
[0084]
  When it is not determined that the edge pattern prepared in advance in the 3 × 3 pixel area matches the binary pattern, such a pattern matching method is not used as the first interpolation processing, and the near list method is executed. The In the near list method, as shown in FIG. 14, the distance between the four surrounding grid points Pij, Pi + 1j, Pij + 1, Pi + 1j + 1 and the point Puv to be interpolated is obtained, and the data of the nearest grid point is transferred as it is. This can be expressed as a general formula:
Puv = Pij
Here, i = [u + 0.5] and j = [v + 0.5]. In addition, [] has shown taking an integer part with a Gauss symbol.
[0085]
  FIG. 15 shows a situation in which the number of pixels is interpolated by three times in the vertical and horizontal directions by the near list method. It is assumed that there are four corner pixels (□ Δ ○ ●) before interpolation, and the data of the nearest pixel among these pixels is transferred as it is to the pixel generated by interpolation. That is, in this example, the pixels adjacent to the pixels at the four corners are respectively copied. Further, when such processing is performed, an original image in which black pixels are diagonally arranged with a white pixel as a background as shown in FIG. 16 is diagonally expanded while black pixels are enlarged three times vertically and horizontally as shown in FIG. Will be placed in the direction. The near list method has a feature that the edge of an image is maintained as it is. Therefore, when enlarged, the edge of the jaggy is conspicuous, but the edge is held as an edge.
[0086]
  After the interpolation pixel is generated by the first interpolation process in this way, the second interpolation process is further executed unless the rate is “1”. In the present embodiment, interpolation processing by a so-called cubic method is executed as the second interpolation processing, and this processing will be described below. As shown in FIG. 18, the cubic method uses data of a total of 16 lattice points including not only four lattice points surrounding the point Puv to be interpolated but also surrounding lattice points.
[0087]
  When a total of 16 grid points surrounding the interpolation point Puv have values, the interpolation point Puv is determined by the influence thereof. For example, if interpolation is to be performed using a linear expression, weighted addition may be performed in inverse proportion to the distance from two grid points sandwiching the interpolation point. If attention is paid to the X-axis direction, the distance from the interpolation point Puv to the above 16 grid points is x1, the distance to the left outer grid point, the distance to the left inner grid point, x2, While expressing the distance x3 to the grid point and the distance x4 to the right outer grid point, the degree of influence corresponding to such a distance is represented by a function f (x). Further, when paying attention to the Y-axis direction, the distance from the interpolation point Puv to the 16 grid points is y1, the distance to the upper outer grid point, y2 to the upper inner grid point, and the lower inner grid point. While representing the distance y3 to the point and the distance y4 to the lower outside lattice point, the degree of influence can be similarly expressed by the function f (y).
[0088]
  Since the 16 grid points contribute to the interpolation point Puv with an influence degree corresponding to the distance as described above, the influence degree in the X-axis direction and the Y-axis direction with respect to data is accumulated in all the grid points. The formula is as follows.
[0089]
[Expression 2]

  If the degree of influence according to the distance is expressed by a cubic convolution function,
f (t) = {sin (πt)} / πt
It becomes. The distances x1 to x4 and y1 to y4 described above are calculated as follows using the absolute values of the coordinate values (u, v) of the lattice points Puv.
x1 = 1+ (u- | u |) y1 = 1+ (v- | v |)
x2 = (u- | u |) y2 = (v- | v |)
x3 = 1- (u- | u |) y3 = 1- (v- | v |)
x4 = 2- (u- | u |) y4 = 2- (v- | v |)
  If we expand about P under the above assumptions,
[0090]
[Equation 3]

It becomes. Note that the degree of influence f (t) corresponding to the distance is approximated by the following cubic expression as called a cubic convolution function.
[0091]
[Expression 4]

This cubic method has a feature that it gradually changes from one lattice point to the other lattice point, and the degree of change becomes a so-called cubic function.
[0092]
  19 and 20 show specific examples when interpolation is performed by the cubic method. In order to facilitate understanding, a model in which there is no change in data in the vertical direction and an edge occurs in the horizontal direction will be described. Also, the number of pixels to be interpolated is 3 points.
First, specific numerical values in FIG. 20 will be described. The gradation value of the pixel before interpolation is shown as “Original” in the left column, and four pixels (P0, P1, P2, P3) having the gradation value “64” are arranged, and the pixel having the gradation value “128”. Five pixels (P5, P6, P7, P8, P9) having a gradation value “192” are arranged with one point between (P4). In this case, the edge is a pixel portion having a gradation value of “128”.
[0093]
  When three pixels (Pn1, Pn2, Pn3) are interpolated between the pixels, the distance between the interpolated pixels is “0.25”, and the above-described x1 to x4 are for each interpolation point. It becomes the numerical value of the middle column of the table. f (x1) to f (x4) are also calculated uniquely corresponding to x1 to x4. For example, x1, x2, x3, and x4 are “1.25” and “0.25”, respectively. , “0.75”, and “1.75”, f (t) corresponding thereto are roughly “−0.14”, “0.89”, “0.30”, “−0.05”. " In addition, when x1, x2, x3, and x4 are “1.50”, “0.50”, “0.50”, and “1.50”, respectively, 0.125 "," 0.625 "," 0.625 ", and" -0.125 ". In addition, when x1, x2, x3, and x4 are “1.75”, “0.75”, “0.25”, and “1.25”, respectively, −0.05 ”,“ 0.30 ”,“ 0.89 ”, and“ −0.14 ”. The result of calculating the gradation value of the interpolation point using the above results is shown in the right column of the table and is also shown in a graph in FIG. The meaning of this graph will be described in detail later.
[0094]
  If it is assumed that there is no change in the data in the vertical direction, the calculation is simplified, and only the data of four grid points (P1, P2, P3, P4) arranged in the horizontal direction are referred to, and Using the influence degree f (t) corresponding to the distance to the grid point, it can be calculated as follows.
P = P1 ・ f (x1) + P21f (x2) + P3 ・ f (x3) + P4 ・ f (x4)
Therefore, when calculating the interpolation point P21,
P21 = 64 * f (1.25) + 64 * f (0.25) + 64 * f (0.75) + 128 * f (1.75)
= 64 * (-0.14063) + 64 * (0.890625) + 64 * (0.296875) +128 * (-0.04688)
= 61
It becomes.
[0095]
  Since the cubic method can be expressed in a cubic function, the quality of the interpolation result can be influenced by adjusting the shape of the curve.
As an example of the adjustment,
0 <t <0.5 f (t) =-(8/7) t ** 3- (4/7) t ** 2 + 1
0.5 <t <1 f (t) = (1-t) (10/7)
1 <t <1.5 f (t) = (8/7) (t-1) ** 3+ (4/7) (t-1) ** 2- (t-1)
1.5 <t <2 f (t) = (3/7) (t-2)
Will be called the hybrid bicubic method or the M (modified) cubic method.
[0096]
  FIG. 21 shows a specific example when interpolation is performed by the M cubic method, and shows a result of interpolation for the same hypothetical model as in the cubic method. FIG. 19 also shows the result of interpolation processing by the M cubic method. In this example, the cubic function-like curve becomes slightly steep, and the entire image becomes sharp. Regardless of whether the M cubic method or the normal cubic method is used, if the interpolation pixel is generated by the cubic method in the second interpolation processing, the variable of the superimposition data calculation formula, that is, the first interpolation processing data is used. , Rate, and all of the second interpolation processing data are calculated, the interpolation data of the interpolation pixel is calculated by substituting the values into the same equation.
[0097]
  Hereinafter, a state in which the superimposition process is performed on the image data as illustrated in FIG. Here, it is assumed that the number of appearances of the luminance value Y is greater than 15 in the 5 × 5 region around the target pixel in the description, and the rate is not “0” or “1”. Therefore, after the rate is calculated in step S108, the first interpolation process is executed in step S114. In step S202, a 3 × 3 pixel area (solid line area) around the target pixel shown in FIG. 22A is extracted, and in step S204, a binary pattern shown in FIG. 22B is created from the area data.
[0098]
  This binary pattern matches a predetermined edge pattern prepared in advance, and processing by the pattern matching method is performed in step S208 to generate an interpolation pixel shown in FIG. At this time, the 3 × 3 pixel shown in FIG. 22A is a diagonal edge and has a luminance gradient toward the lower left. Therefore, in the interpolation pixel shown in FIG. 22C, the edge is emphasized with a luminance gradient toward the lower left while retaining the edge in the diagonal direction reflecting the properties of the 3 × 3 pixel.
[0099]
  On the other hand, in step S118, the interpolated pixel shown in FIG. 22 (d) is generated using the pixel data shown in FIG. 22 (a). Here, as described above, 16 pixel data around the interpolation pixel are used to generate the interpolation pixel. For example, in order to generate the upper right pixel shown as a black dot in FIG. 16 pixels consisting of the solid line and dotted line areas shown are used. In such an interpolated pixel, the edge is ambiguous compared to the pattern matching method while reflecting a subtle change between the pixels in the solid line and dotted line regions due to the nature of the cubic method. That is, although there is an edge in the diagonal direction of the interpolated pixel, the luminance gradient in the lower left direction becomes gentle and the edge is ambiguous.
[0100]
  After the interpolation pixels are generated by the first interpolation process and the second interpolation process in this way, each interpolation pixel by the first interpolation process is multiplied by rate, and each interpolation pixel by the second interpolation process is (1). -Rate) is multiplied to generate a superimposed pixel as shown in FIG. As shown in FIG. 4E, when the superimposed interpolation pixel for the target pixel is generated, all the target pixels having new coordinate values are interpolated through the determination in step S122, and the interpolated image data is converted into the next stage in step S124. Deliver to processing. However, depending on the interpolation magnification, the amount of interpolated image data may be extremely large, or the memory area that can be used by the printer driver 12c may not be so large in the first place. In such a case, the data may be output separately for each fixed amount of data.
[0101]
  Finally, a summary of this embodiment is shown in FIG. The digital image obtained by scanning the photograph by the scanner 11a is the original image data, which is to be processed in the subsequent process. The means for obtaining the original image data is the image data obtaining means C11.
  In step S104, the original image data is read, and a histogram of luminance values is created in a peripheral 5 × 5 pixel area centered on the target pixel. In step S106, the number of times the different luminance values appear in the obtained histogram is acquired, and if it is less than a predetermined threshold value, it is determined that the nature of the image is a simple non-natural image and the first interpolation is performed. The processing superimposition ratio (rate) is set to 1, but if it is large, the nature of the image is determined to be a natural image or a non-natural image mixed image, and the rate is determined by the evaluation function F. This evaluation function F is a function of the difference between the maximum luminance value Ymax and the minimum luminance value Ymin in the above region, and evaluates the edge likeness. The means for obtaining the superposition ratio from the image properties in this way is the first superposition ratio determination means C14.
[0102]
  In step S114, an interpolation process that tends to preserve edges is performed in a predetermined region centered on the target pixel by the pattern matching method or the near list method, and this is the first interpolation processing means C12.
  On the other hand, in step S118, the interpolation processing is performed by the cubic method in which the edge is ambiguous compared to the pattern matching method while reflecting a subtle change between pixels, and this is the second interpolation processing means C13.
[0103]
  In step S120, the pixel data that has been interpolated separately is weighted and added according to the calculation formula (1) using the superposition ratio (rate) determined as described above to generate an interpolation pixel. To do. The sum of the superposition ratios is naturally “1”. Of course, this means is the image data superimposing means C15.
  Thereafter, when the pixel of interest is sequentially moved and processing is performed for all the pixels, an interpolation image is created. Accordingly, in step S124, the completed interpolation image data is output. This process is the image data output means C16. As an example, if it is output to the printer 17b, a printed matter is obtained.
[0104]
  Next, another embodiment of the present invention will be described. In this embodiment, an image data interpolation program is used for the printing system. Note that components and the like common to the above-described embodiment are referred to as they are.
  The color printer 17b described above outputs the processing result of the application 12d as print data via the printer driver 12c. The color printer 17b prints a corresponding image by adding dots on the printing paper using color ink. To do.
[0105]
  23 to 25 show a schematic configuration of a color inkjet printer 21 as an example of such a color printer. The color inkjet printer 21 includes a print head 21a composed of three print head units, a print head controller 21b for controlling the print head 21a, a print head digit moving motor 21c for moving the print head 21a in the digit direction, A dot printing mechanism comprising a paper feed motor 21d for feeding print paper in the line direction, and a print controller 21b, a print head digit movement motor 21c, and a printer controller 21e which is an interface with external devices in the paper feed motor 21d; It is possible to print an image while scanning the print head 21a on a recording medium that is a printing sheet in accordance with the print data.
[0106]
  FIG. 24 shows a specific configuration of the print head 21a, and FIG. 25 shows an operation during ink ejection. The print head 21a is formed with a fine pipe line 21a3 extending from the color ink tank 21a1 to the nozzle 21a2, and an ink chamber 21a4 is formed at the end of the pipe line 21a3. The wall surface of the ink chamber 21a4 is formed of a flexible material, and a piezoelectric element 21a5 that is an electrostrictive element is provided on the wall surface. The piezo element 21a5 has a crystal structure that is distorted by applying a voltage and performs high-speed electro-mechanical energy conversion. The distorted operation of the crystal structure pushes the wall surface of the ink chamber 21a4, thereby the ink chamber 21a4. Reduce the volume of the. Then, a predetermined amount of color ink particles are ejected vigorously from the nozzle 21a2 communicating with the ink chamber 21a4. This pump structure is called a micro pump mechanism.
[0107]
  In addition, two independent rows of nozzles 21a2 are formed in one print head unit, and color ink is supplied independently to the nozzles 21a2 in each row. Therefore, two print nozzle units are provided for each of the three print head units, and it is possible to use six color inks to the maximum. In the example shown in FIG. 23, two rows in the print head unit in the left row are used for black ink, only one row in the middle print head unit is used for cyan ink, and in the print head unit in the right row. The left and right two rows are used for magenta ink and yellow ink, respectively.
[0108]
  On the other hand, the interval in the vertical direction of the nozzles 21a2 formed in the print head 21a does not coincide with the printing resolution, and in general, the nozzles 21a2 are formed at intervals larger than the printing resolution. Nevertheless, higher resolution printing is possible because the paper feed motor 21d is controlled in the paper feed direction. For example, if the paper is fed in eight steps between the intervals of the nozzles 21a2 and printing is performed by operating the print head 21a in the shift direction for each step, the resolution is improved. Of course, it can be said that depending on the timing control, the color feed direction can be said to be the resolution if the color ink is ejected at an arbitrary interval. In a strict sense, the dot diameter of the color ink can also be an element of resolution, but will be ignored here for the sake of easy understanding.
[0109]
  In the present embodiment, printing is executed based on image data acquired by the image input device of the computer system 10 on the premise of the hardware system as described above. At this time, if there is a difference between the resolution of the original image data and the resolution of the color printer 17b, an interpolation process is executed. Here, changes in resolution and gradation when print data is output to the color printer 17b when the application 12d executes print processing will be described.
[0110]
  FIG. 26 shows the flow of image data. The application 12d generates a print request to the operating system 12a, and at that time, delivers the output size, printing paper, printing speed, ink type, and RGB 256-gradation image data. Then, the operating system 12a delivers this output size, printing paper, printing speed, ink type, and image data to the printer driver 12c. Here, the operating system 12a outputs the operation results of the keyboard 15a and the mouse 15b to the printer driver 12c while displaying on the display 17a via the display driver 12b, so that the printer driver 12c has the specified output size. Image interpolation processing is executed to generate print data. Normally, this print data is CMYK 2 gradation, and is output from the hardware port to the color printer 17b via the operating system 12a.
[0111]
  As described above, in this embodiment, the print control program is executed by the computer system 10 and print data is output to the color printer 17b. However, the target printing apparatus is connected to the above-described ink jet type color printer 21. It is not limited. For example, the color printer 21 is of an ink jet type that employs a micropump mechanism, but a printer other than the micropump mechanism can also be employed. As shown in FIG. 27, a heater 21a8 is provided on the wall surface of the pipe 21a7 in the vicinity of the nozzle 21a6, and the heater 21a8 is heated to generate bubbles, and a bubble jet (R ) Type pump mechanism is also in practical use.
[0112]
  As another mechanism, FIG. 28 shows a schematic configuration of main parts of a so-called electrophotographic color printer 22. A charging device 22b, an exposure device 22c, a developing device 22d, and a transfer device 22e are arranged on the periphery of the rotating drum 22a as a photoconductor corresponding to the rotation direction, and the peripheral surface of the rotating drum 22a is made uniform by the charging device 22b. Then, the image portion is charged by the exposure device 22c, the toner is attached to the uncharged portion by the developing device 22d, and the transfer device 22e transfers the toner onto paper as a recording medium. Thereafter, the toner is passed between the heater 22f and the roller 22g to melt and fix the toner on the paper. Then, since these are a set and printing with toner of one color is performed, a total of four colors are individually provided.
[0113]
  That is, the specific configuration of the printing apparatus is not particularly limited, and various changes can be made not only in the application range of such an individual printing method but also in the application mode.
  Next, a process for executing optimal image processing corresponding to the output resolution using the above-described printing system will be described.
FIG. 29 is a block diagram showing a schematic configuration of this printing system. The print quality acquisition unit C21 acquires the print quality to be printed in the executed print job, and the first interpolation processing unit C22 and the second interpolation processing unit C23 interpolate to increase the number of constituent pixels in the image data. Process. Here, the first interpolation processing means C22 can execute the pattern matching method and the near list method as the interpolation processing, and the second interpolation processing means C23 can execute the cubic method as the interpolation processing. .
[0114]
  The second superimposition ratio determining means C24 determines the superimposition ratio of the interpolation pixels by the first interpolation processing means C22 and the second interpolation processing means C23 based on the print quality acquired by the print quality acquisition means C21. When the second superimposition ratio determining means C24 determines the superposition ratio, the image data superimposing means C25 superimposes the interpolated pixel data by the first interpolation processing means C22 and the second interpolation processing means C23 at the superimposition ratio. The print control processing unit C26 generates print data based on the pixel data superimposed by the image data superimposing unit C25 and executes printing.
[0115]
  FIG. 30 shows a software flow related to the printing process executed by the printer driver 12c described above. In FIG. 30, in step S302, original image data is acquired. For example, when the application 12d reads an image from the scanner 11a, performs predetermined image processing, and then performs print processing, image data with a predetermined resolution is transferred to the printer driver 12c via the operating system 12a. The stage is applicable.
[0116]
  In step S303, print settings are input. When the printing process is executed by the application 12d, if the operating system 12a provides a GUI environment, a window for printing operation is displayed as shown in FIG. Various parameters or the like can be adopted here. In the present embodiment, there are “(print) number of copies”, “start page”, “end page”, and the like. In addition to the “OK” button and the “Cancel” button, “Printer setting” buttons are also provided as operation instruction buttons.
[0117]
  When “printer setting” is instructed, a window display as shown in FIG. 32 is performed. This window display is prepared for making various settings according to the function of each printer, and the superposition ratio is changed according to the setting contents in this window as will be described later. In this example, one of “360 dpi” and “720 dpi” can be selected as “(print) resolution”. Also, the size of “A4” or “B5” as “paper” and the quality of “plain paper” or “glossy paper”, “fast” or “slow” as “printing speed”, “pigment” as “ink” “Dye” can be selected. Of course, this print setting is merely an example, and it is not necessary to configure the user to select all of them. Various modes such as performing ink communication by performing bidirectional communication with the color printer 17b can be employed. is there.
[0118]
  In the present embodiment, the above-described setting contents are stored in a setting file under the management of the operating system 12a. If the setting has already been made, the setting file is read with reference to the setting file. When the operator changes the setting content in accordance with the printing operation, the changed setting content is read as the print setting. Such processing in step S303 is the above-described print quality acquisition process or print quality acquisition function when viewed as software. The various steps executed by the computer including the print quality acquisition process are the operating system 12a itself. And can be understood as not directly including hardware. On the other hand, if it is considered that it is organically combined with hardware such as a CPU, it corresponds to the print quality acquisition means C21.
[0119]
  In step S304, in the read image data, the pixel to be interpolated is set as the target pixel, the pixel data of the peripheral 5 × 5 pixel region around the target pixel is set as the reference pixel, and the histogram of the luminance value of the reference pixel is displayed. create. In step S306, the number of times that different luminance values appear in the obtained histogram is acquired, and it is determined whether or not the number of appearances is smaller than 15. As the number of different luminance values increases, the number of colors in the reference pixel increases. Therefore, in this case, a non-natural image having a luminance value appearance count smaller than 15 is considered to be a non-natural image. When it is determined, the superposition ratio (rate) of the first interpolation processing is set to 1 in step S310.
[0120]
  If it is determined in step S306 that the number of appearances of the luminance value is not less than 15, the evaluation function F is determined based on the print setting contents input in step S303 in step S307. In the determination of the evaluation function F in step S307, the evaluation function F is determined so that the superposition ratio of the second interpolation processing is increased when the print quality is high. Although details will be described later, in this embodiment, two types of evaluation functions are prepared for cases where the print quality is high and low, and the evaluation function value when the print quality is high is the evaluation function when the print quality is low. It is smaller than the value. The evaluation function F is a function of the difference between the luminance value width of the reference pixel, that is, the maximum luminance value Ymax and the minimum luminance value Ymin in the reference pixel.
[0121]
  In step S308, the rate is determined by substituting the difference between the maximum luminance value Ymax and the minimum luminance value Ymin based on the histogram for the evaluation function F determined in step S307. Therefore, the series of processing shown in steps S304, S306, S307, S308, and S310 corresponds to the second superimposition ratio determining step or the second superimposition ratio determining function, and these are organically combined with hardware such as a CPU. If it thinks as a thing, the 2nd superimposition ratio determination means C24 will be comprised.
[0122]
  When the rate is determined in step S308 or step S310, it is determined in step S312 whether or not the rate is “0”. When the rate is “0”, the first interpolation process is not performed. ing. If it is not determined in step S312 that the rate is “0”, the first interpolation process is executed in step S314. Accordingly, this processing corresponds to the first interpolation processing step or the first interpolation processing function, and the first interpolation processing means C22 is considered to be organically coupled with hardware such as a CPU. Will be composed.
[0123]
  Further, in step S316, it is determined whether or not the rate is “1”. When the rate is “1”, the second interpolation process is not performed. If it is not determined in step S316 that the rate is “1”, the second interpolation process is executed in step S318. Therefore, this processing corresponds to the second interpolation processing step or the second interpolation processing function, and considering that these are organically combined with hardware such as a CPU, the second interpolation processing means C23 is provided. Will be composed. After determining the rate in this way and executing the first interpolation process or the second interpolation process, the pixel data is superimposed by the above-described equation (1) based on the rate in step S320, and the interpolation pixel is determined. Generate.
Superimposition data =
(First interpolation processing) × rate + (second interpolation processing) × (1-rate)
                                                  ... (1)
  Therefore, if this process corresponds to the image data superimposing step or the image data superimposing function and these are considered to be organically coupled with hardware such as a CPU, the image data superimposing means C25 is configured. When the generation of the interpolation data for the target pixel is completed in this way, it is determined in step S322 whether or not the superimposition process for all the target pixels of the input original image data is completed, and the process for all the target pixels is performed. The process from step S304 onward is repeated until the process ends.
[0124]
  When the superimposition processing for all the target pixels is completed, printing based on the image data interpolated in step S324 is performed. In the case of the printer driver 12c, print data is not obtained only by resolution conversion, but color conversion or halftone processing is necessary. Therefore, here, after executing such conversion processing and the like, print data is output to the color printer 17b, and this processing corresponds to the above-mentioned print control processing step or print control processing function, which is a hardware such as a CPU. If it is considered that it is organically combined with the wear, the print control processing means C26 is configured.
[0125]
  Next, more specific processing for the above flow will be described. In the present embodiment, it is determined whether the original image acquired in step S302 is computer graphics (non-natural image) or a photograph (natural image), and not only the print quality but also the determination result is reflected in the superposition ratio. It is supposed to let you. Of course, in order to prevent an error in determining the interpolation method, a configuration in which the rate is not fixed to “1” or “0” is possible. However, in this embodiment, if the number of appearances of the luminance value is smaller than 15, first, it is unnatural. The emphasis is on executing the second interpolation process on the non-natural image so as not to obscure the contour of the image.
[0126]
  Specifically, this determination uses a histogram of luminance values shown in FIG. 14. In step 104, the luminance value is obtained for each reference pixel in a 5 × 5 pixel area, and the number of pixels in the range that the luminance can take. Aggregate histograms. Then, the number of appearance of the luminance value, that is, how many luminance values whose distribution number is not “0” is counted, and the determination in step S306 is performed.
  In this way, if the luminance value histogram has more than 15 luminance values whose distribution number is not “0”, the superimposition ratio of the first interpolation processing is determined by the evaluation function. In this embodiment, two types of evaluation functions are prepared in advance, one for high print quality and one for low print quality, and one of the setting contents set in the window shown in FIG. 32 shows high quality. If there is, it is determined in step S306 to use the high print quality evaluation function. FIG. 33 shows an example of the evaluation function F (y), where FIG. F1 is an evaluation function for low print quality, and FIG. 33 is an evaluation function for high print quality.
[0127]
  The evaluation function F1 (y) has a value in a range where “y” is “0 ≦ y ≦ 255” and fluctuates within a range of “0 ≦ F (y) ≦ 1”. Further, “F (y) = 0” in the range of “0 ≦ y ≦ 64”, “F (y) = 1” in the range of “192 ≦ y ≦ 255”, and “64 ≦ y ≦ 192”. In the range “”, F (y) increases linearly from “0” to “1”. The evaluation function F2 (y) has a value in the range of “0 ≦ y ≦ 255” for “y”, and fluctuates in the range of “0 ≦ F (y) ≦ 0.7”. In the range of “0 ≦ y ≦ 64”, “F (y) = 0”, and in the range of “192 ≦ y ≦ 255”, “F (y) = 0.7”, and “64 ≦ y In the range of ≦ 192, F (y) increases linearly from “0” to “0.7”.
[0128]
  Here, the luminance value width “Ymax−Ymin” in FIG. 14 is substituted for “y”. Therefore, the rate increases as the luminance value width increases, that is, as the relationship between the reference pixels is more likely to be an edge. Furthermore, the value of the evaluation function F1 is always greater than or equal to the value of the evaluation function F2 even with the same luminance value width. That is, in the case of high quality printing with the same luminance value width, the superposition ratio of the second interpolation processing that performs the interpolation processing without losing the gradation of the image is increased.
[0129]
  Interpolation that does not impair the gradation of the original image when printing with high print resolution, high print paper quality, or low speed print, or in combination with high print quality. For example, a subtle gradation change is reproduced as an effect visible to the user in the print result. On the other hand, when printing is performed with low print quality, even if printing is performed using print data that can reproduce a subtle gradation change, the printed result is not reproduced as an effect visible to the user. Therefore, it can be said that it is worth performing the second interpolation process at the time of high print quality. In this embodiment, the superimposition ratio of the second interpolation process is increased by using the evaluation function F2.
[0130]
  In the present embodiment, the type of ink also affects the selection of the evaluation function. Here, “pigment” and “dye” can be selected as the type of ink, and it cannot be generally determined that either one is of high quality, but “pigment” tends not to bleed compared to “dye”. There is a tendency that the edges of the “pigment” are more conspicuous in the print result when comparing the two. Therefore, it can be said that the type of ink reflects the print quality. In this embodiment, when the same print result is obtained with both inks, the second interpolation process is performed in the case of “pigment” compared to “dye”. The ratio may be increased. Therefore, in the present embodiment, when the ink is a “pigment”, the evaluation function F2 is used.
[0131]
  Of course, the method of reflecting the evaluation function and the acquired print setting in this embodiment is merely an example, and various aspects can be adopted. For example, the evaluation function does not need to be two types as described above, and a plurality of types can be prepared. When there are four types of print setting items that affect print quality as in the above embodiment, four evaluation functions are prepared, and an evaluation function having a smaller maximum value is selected each time one item set to high quality is added. It is possible to do this.
[0132]
  In addition, it is not necessary to prepare a plurality of evaluation functions in advance as described above. One evaluation function F1 described above is prepared, and the evaluation function is used which is "0.7 times" during high quality printing. Thus, the rate can be determined, or the evaluation function can be "0.9 times" every time one item set to high quality is added. Furthermore, the shape of the function need not be limited to the example shown in FIG. 33, and may be a function that increases monotonously with the luminance value width when the luminance value width is a variable, and covers the luminance value widths “0” to “255”. It is also possible to employ a smoothly changing evaluation function.
[0133]
  When the evaluation function to be used is determined in step S307 as described above, the superposition ratio rate is determined by the determined evaluation function in step S308. In the case of the histogram as shown in FIG. 14, since there are 19 luminance values whose distribution number is not “0”, the process of step S307 is executed after the determination of step S306. Also, as shown in FIG. 32, when “720 dpi” for resolution, “glossy paper” for paper, “fast” for printing speed, and “pigment” for ink are selected, resolution, paper, and ink are selected. In step S307, the evaluation function F2 is selected because the print setting for selecting the evaluation function for high-quality printing has been made. In step S308, the rate is determined by the evaluation function F2, and the first interpolation process and the second interpolation process are executed in steps S314 and S318, respectively.
[0134]
  After the interpolation pixel is generated by the first interpolation process, the second interpolation process is further executed unless the rate is “1”.
  Regardless of whether the M cubic method or the normal cubic method is used, if the interpolation pixel is generated by the cubic method in the second interpolation processing, the variable of the superposition data calculation formula (1), that is, the first Interpolation process data, rate, and second interpolation process data are all calculated. By substituting a value into Equation (1), the superimposed data of the interpolation pixel is calculated.
[0135]
  Here, as shown in FIG. 32 described above, the print quality in this state is high, and a slight gradation change can be reflected in the print result. At this time, since the evaluation function is F2, the superposition ratio of the cubic method is high.
  On the other hand, in the window shown in FIG. 32, the print quality is low when “360 dpi” for resolution, “plain paper” for paper, “fast” for printing speed, and “dye” for ink are selected. The effect of the second interpolation process is not noticeable. In this state, the evaluation function F1 is used as described above, and the superposition ratio of the second interpolation process is lower than that in the setting state shown in FIG. At this time, the superposition ratio of the pattern matching method that emphasizes the edge or the near list method is higher than the cubic method that reproduces the gradation value, and it is possible to obtain a sharp print result although the print quality is low.
[0136]
  Finally, a summary of the present invention is shown in FIG.
  In the “Printer Settings” window display, various settings can be made according to the function of each printer, and the setting contents are stored in a setting file. If the setting has already been made, the setting file is read with reference to the setting file. When the setting content is changed in accordance with the printing operation, the changed setting content is read. In step S303, the print setting is input in this way, and this corresponds to the print quality acquisition unit C21.
[0137]
  On the other hand, in step S304, the original image data is read, and a histogram of luminance values is created in a peripheral 5 × 5 pixel area centered on the target pixel. In step S306, the number of times the different luminance values appear in the obtained histogram is acquired, and if it is less than a predetermined threshold, it is determined that the nature of the image is a simple non-natural image and the first interpolation is performed. The process superposition ratio (rate) is set to 1, but if it is large, the image function is determined to be a natural image or a non-natural image mixed image, and the evaluation function is determined in order to determine the rate by the evaluation function F. The evaluation function F is determined according to the print quality, and includes an evaluation function F1 for low image quality and an evaluation function F2 for high image quality. The evaluation function F is a function of the difference between the maximum luminance value Ymax and the minimum luminance value Ymin in the above-described region, and evaluates the edge likeness. When the edge likelihood is high, the ratio of the interpolation processing with a relatively low calculation load that is easy to emphasize the edge with the evaluation function F1 is increased. Otherwise, the ratio of the interpolation processing with a high calculation load that emphasizes the smoothness with the evaluation function F2. To increase. The means for obtaining the superposition ratio from the print quality in this way is the second superposition ratio determining means C24.
[0138]
  In step S314, an interpolation process that tends to preserve edges is performed by a pattern matching method or a near list method in a predetermined region centered on the pixel of interest, and this is the first interpolation processing means C22.
  On the other hand, in step S318, the interpolation processing is performed by the cubic method in which the edge is ambiguous compared to the pattern matching method while reflecting a subtle change between pixels, and this is the second interpolation processing means C23.
[0139]
  In step S320, weighted addition is performed on the pixel data individually interpolated by the calculation formula (1) using the superposition ratio (rate) determined as described above to generate an interpolation pixel. To do. The sum of the superposition ratios is naturally “1”. Of course, this means is the image data superimposing means C25.
  Thereafter, when the pixel of interest is sequentially moved and processing is performed for all the pixels, an interpolation image is created, and in step S324, the completed interpolation image data is used and printed by the printer 17b. This process is the print control processing means C26.
[0140]
  Thus, in the present invention, the first interpolation process and the second interpolation process are superimposed based on a predetermined evaluation function. Therefore, although the edge of the superimposed pixel is ambiguous compared to the case of only the pattern matching method, the edge is sharper than that of the case of only the cubic method. In addition, subtle gradation changes are reduced compared to the cubic method alone, but the gradation changes are richer than in the pattern matching method alone. In addition, since the evaluation function is a function of the luminance value width, it is possible to determine the interpolation processing ratio according to the property of the image. Furthermore, the overlap ratio of the interpolation process more suitable for the print quality that directly affects the effect of the interpolation process is increased. As a result, the advantages of the individual interpolation processes become more noticeable. In addition, there is no such thing as highlighting the disadvantages of both. Therefore, it is possible to perform an accurate interpolation process according to the print quality while preventing an error in determining an interpolation method based on the property determination of the interpolation target image.
[Brief description of the drawings]
FIG. 1 is a block diagram of an image data interpolation apparatus according to an embodiment of the present invention.
FIG. 2 is a block diagram of specific hardware of the image data interpolation device.
FIG. 3 is a schematic view showing another application example of the image data interpolation apparatus of the present invention.
FIG. 4 is a schematic view showing another application example of the image data interpolation apparatus of the present invention.
FIG. 5 is a schematic view showing another application example of the image data interpolation apparatus of the present invention.
FIG. 6 is a schematic view showing another application example of the image data interpolation apparatus of the present invention.
FIG. 7 is a flowchart of processing related to resolution conversion executed by a printer driver.
FIG. 8 is a first interpolation processing flowchart.
FIG. 9 is a diagram showing a histogram of luminance values.
FIG. 10 is a diagram illustrating a specific example of an evaluation function F (y).
FIG. 11 is a diagram illustrating an example of a luminance pattern of 3 × 3 pixels.
FIG. 12 is a diagram illustrating a 5 × 5 pixel region that is a reference pixel;
FIG. 13 is a diagram illustrating a specific example of an edge pattern.
FIG. 14 is a conceptual diagram of the near list method.
FIG. 15 is a diagram showing a situation in which data at each grid point is transferred by the near list method.
FIG. 16 is a schematic diagram showing a state before interpolation in the near list method.
FIG. 17 is a schematic diagram illustrating a situation after interpolation in the near list method.
FIG. 18 is a conceptual diagram of the cubic method.
FIG. 19 is a diagram showing a change state of data at the time of concrete application of the cubic method.
FIG. 20 is a diagram illustrating a specific application example of the cubic method.
FIG. 21 is a diagram illustrating a specific application example of the M cubic method.
FIG. 22 is a diagram illustrating a specific state in which superimposition processing is performed on image data.
FIG. 23 is a schematic block diagram of an inkjet color printer.
FIG. 24 is a schematic explanatory diagram of a print head unit in the color printer.
FIG. 25 is a schematic explanatory diagram illustrating a situation in which color ink is ejected by the print head unit.
FIG. 26 is a flowchart showing the flow of image data in the printing system.
FIG. 27 is a schematic explanatory diagram illustrating a state in which color ink is ejected by a bubble jet (R) type print head.
FIG. 28 is a schematic explanatory diagram of an electrophotographic printer.
FIG. 29 is a block diagram illustrating a schematic configuration of a printing system.
FIG. 30 is a flowchart of processing related to printing processing executed by a printer driver.
FIG. 31 is a diagram illustrating an operation window for print processing.
FIG. 32 is a diagram illustrating an operation window for setting a printer.
FIG. 33 is a diagram illustrating a specific example of an evaluation function F (y).
FIG. 34 is a diagram showing a schematic configuration of the present invention.
FIG. 35 is a diagram showing a schematic configuration of the present invention.
FIG. 36 is a diagram illustrating condition determination for a 45-degree edge.
FIG. 37 is a diagram showing condition determination for an edge of 30 degrees.
FIG. 38 is a diagram illustrating a correspondence between an original pixel and an interpolation pixel when a horizontal edge is found in the case of double interpolation.
FIG. 39 is a diagram illustrating a correspondence between an original pixel and an interpolation pixel when a right-angled edge is found in the case of double interpolation.
FIG. 40 is a diagram illustrating a correspondence between an original pixel and an interpolation pixel when a 45-degree edge is found in the case of double interpolation.
FIG. 41 is a diagram illustrating condition determination for a 45-degree edge.
FIG. 42 is a diagram illustrating a correspondence between an original pixel and an interpolation pixel when a 30-degree edge is found in the case of double interpolation.
FIG. 43 is a diagram illustrating condition determination for a 30-degree edge.
[Explanation of symbols]
  10. Computer system
  11a ... Scanner
  11b ... Digital still camera
  11c ... Video camera
  12 ... Computer body
  12a ... Operating system
  12b ... Display driver
  12c ... Printer driver
  12d Application
  13a: floppy (R) disk drive
  13a1 Floppy (R) disk
  13b ... Hard disk
  13c ... CD-ROM drive
  13c1 CD-ROM
  14a ... modem
  15a ... Keyboard
  15b ... Mouse
  17a ... Display
  17b ... Color printer
  18a ... Color facsimile machine
  18b ... color copying machine
  21 ... Color inkjet printer
  22 Color printer

Claims (18)

An image data interpolation program for performing pixel interpolation on image data in which an image is expressed in multiple gradations by a plurality of pixels on a computer,
Get the above image data,
Based on the reference pixels around the pixel to be interpolated in the image data, a feature amount representing the degree of unnaturalness of the image is acquired,
And, if it is determined that an edge pattern exists in the reference pixels around the pixel to be interpolated on the condition that the acquired feature amount exceeds a predetermined value, the rule is different for each edge pattern and defined in advance. Among the selection rules of the surrounding pixels used for the weighted average for generating the pixel to be interpolated, the surrounding pixels are selected according to the selection rule corresponding to the existing edge pattern, and pixel interpolation using the selected pixel is performed. Do,
An image data interpolation program for causing a computer to execute processing. In the image data interpolation program according to claim 1,
An image data interpolation program characterized in that the feature amount is determined by an evaluation function depending on the data of the reference pixel. In the image data interpolation program according to claim 1 or 2,
An image data interpolation program for determining the feature amount based on the number of appearances of different gradation values in the reference pixel. In the image data interpolation program according to claim 3,
An image data interpolation program characterized in that the feature amount becomes a maximum value when the number of appearances of different gradation values in the reference pixel is smaller than a predetermined threshold value. In the image data interpolation program according to any one of claims 3 and 4,
An image data interpolation program for increasing the feature amount as the gradation value width of the reference pixel increases. In the image data interpolation program according to any one of claims 3 to 5,
An image data interpolation program, wherein the gradation value of the reference pixel is a luminance value of the reference pixel. An image data interpolation method for performing pixel interpolation on image data in which an image is expressed in multiple gradations by a plurality of pixels,
The image data is acquired, a feature amount representing the degree of non-natural image quality of the image is acquired based on reference pixels around the pixel to be interpolated in the image data, and the acquired feature amount exceeds a predetermined value. If it is determined that there is an edge pattern in the reference pixels around the pixel to be interpolated, it is used for the weighted average for generating the pixel to be interpolated, which is defined in advance with different rules for each edge pattern. An image data interpolation method comprising: selecting peripheral pixels according to a selection rule corresponding to the existing edge pattern from the selection rules of peripheral pixels to be performed, and performing pixel interpolation using the selected pixels. In the image data interpolation method according to claim 7,
An image data interpolation method, wherein the feature amount is determined by an evaluation function depending on data of the reference pixel. In the image data interpolation method according to any one of claims 7 and 8,
An image data interpolation method, wherein the feature amount is determined based on the number of appearances of different gradation values in the reference pixel. In the image data interpolation method according to claim 9,
An image data interpolation method, wherein the feature amount is a maximum value when the number of appearances of different gradation values in the reference pixel is smaller than a predetermined threshold value. In the image data interpolation method according to any one of claims 9 and 10,
An image data interpolation method, wherein the feature amount is increased as the gradation value width of the reference pixel is larger. In the image data interpolation method according to any one of claims 9 to 11,
An image data interpolation method, wherein the gradation value of the reference pixel is a luminance value of the reference pixel. An image data interpolation device that performs pixel interpolation on image data in which an image is expressed in multiple gradations by a plurality of pixels,
The image data is acquired, a feature amount representing the degree of non-natural image quality of the image is acquired based on reference pixels around the pixel to be interpolated in the image data, and the acquired feature amount exceeds a predetermined value. If it is determined that there is an edge pattern in the reference pixels around the pixel to be interpolated, it is used for the weighted average for generating the pixel to be interpolated, which is defined in advance with different rules for each edge pattern. An image data interpolating apparatus comprising: selecting peripheral pixels according to a selection rule corresponding to the existing edge pattern, and performing pixel interpolation using the selected pixels. In the image data interpolating device according to claim 13,
An image data interpolation apparatus, wherein the feature amount is determined by an evaluation function depending on data of the reference pixel. In the image data interpolation device according to any one of claims 13 and 14,
An image data interpolation apparatus, wherein the feature amount is determined based on the number of appearances of different gradation values in the reference pixel. In the image data interpolation device according to claim 15,
The image data interpolating apparatus according to claim 1, wherein the feature amount becomes a maximum value when the number of appearances of different gradation values in the reference pixel is smaller than a predetermined threshold value. The image data interpolation device according to any one of claims 15 and 16,
An image data interpolation apparatus characterized in that the feature amount is increased as the gradation value width of the reference pixel is increased. In the image data interpolation device according to any one of claims 15 to 17,
An image data interpolation apparatus, wherein the gradation value of the reference pixel is a luminance value of the reference pixel.

Images