JP4135365B2 - Image processing apparatus and method, recording medium, and program - Google Patents

Description

【数２】

【００８６】
図１０において、横軸は記憶色領域の中心からのマハラノビス距離を表わしており、縦軸は補正量（補正後の輝度値から補正前の輝度値を減算したもの（ΔＹ））を表わしている。
【００８７】
図に示されるように、補正量決定部４２は、例えば、記憶色領域の中心からマハラノビス距離が遠くなるにつれて、補正量が次第に小さくなるような滑らかな曲線を作成する。この例においては、補正を行う領域が「１」以内になるように正規化されており、必要に応じて、記憶色領域ではない領域も補正されるようになされている。例えば、記憶色領域の中心からマハラノビス距離上で「０．６」の距離の範囲までが記憶色領域とされている場合、記憶色領域の中心から「０．６」乃至「１．０」の距離にある記憶色領域ではない領域も補正が施される。
【００８８】
このように、マハラノビス距離が遠くなるにつれて補正量が次第に小さくなる補正カーブを利用することで、記憶色領域の周辺部も補正が施されることとなり、記憶色領域にのみ補正を施す場合に較べて、輝度の変化を滑らかにすることができ、擬似輪郭が発生するのを抑制することができる。すなわち、記憶色領域にのみ補正を施す場合、記憶色領域と記憶色領域でない領域の輝度の差が大きくなり、擬似輪郭が発生することがある。
【００８９】
当然、補正量決定部４２により作成される補正カーブは、図１０に示されるものに限定されるものではなく、諧調飛びが発生しない滑らかなカーブであれば、様々なカーブを補正カーブとして利用することができる。
【００９０】
補正量決定部４２により作成された補正カーブは、画像データと共に記憶色領域抽出部４３に供給される。
【００９１】
記憶色領域抽出部４３には、記憶色の色相値が予め設定されており、記憶色領域抽出部４３は、ステップＳ７０において、画素毎の色相値を確認し、ステップＳ７１に進み、設定されている色相値に基づいて、補正が必要な画素（記憶色領域の画素と、その周辺の画素）を抽出する。記憶色領域抽出部４３は、補正が必要な画素に関する情報を、補正量決定部４２により決定された補正カーブ、および画像データと共に補正処理部４４に出力する。
【００９２】
ステップＳ７２において、補正処理部４４は、供給された補正カーブに基づいて、記憶色領域抽出部４３により抽出された記憶色領域の画素と、必要に応じて、記憶色領域の周辺の画素を補正する。例えば、図１０に示されるような補正カーブが供給されてきた場合、補正処理部４４は、記憶色領域の中心に相当する画素の輝度値を約「０．８」だけ高く補正し、中心に相当する画素からマハラノビス距離が遠くなるにつれて、補正量が次第に小さくなるように輝度値を補正する。
【００９３】
補正処理部４４は、ステップＳ７３において、補正が必要な全ての画素を補正したか否かを判定し、補正してないと判定した場合、ステップＳ７０に戻りそれ以降の処理を繰り返し実行する。一方、補正処理部４４は、補正が必要な全ての画素を補正したと判定した場合、補正後の画像を彩度補正部２５に出力し、処理を終了する。
【００９４】
以上のように、画像のカテゴリに応じて補正方法を切り替えるようにしたので、記憶色領域の境界付近に擬似輪郭が発生するのを抑制することができる。すなわち、人物が比較的広い範囲に写っているポートレート画像が供給されており、その記憶色領域（肌色領域）のみを抽出し、被写体の顔の色を補正した場合、顔の領域と顔以外の領域の輝度値に大きな差異が生じてしまい、その境界部に擬似輪郭が生じてしまうことがある。
【００９５】
また、画像全体を補正する場合であっても、図９に示されるように、輝度値の中心付近に相当する画素に対しては、その輝度値が高くなるように補正し、中心付近より低い輝度値が設定されている画素に対しては、その輝度値がさらに低くなるように補正するようにしたので、画像全体のコントラストを保持することができる。
【００９６】
さらに、図１０に示されるように、記憶色領域を抽出してローカル補正を施す場合、記憶色領域の周辺の画素にも補正が施されるようにしたので、記憶色領域のみに補正を施す場合に較べて、記憶色領域と記憶色領域以外の領域に、急激な輝度値の変化が生じるのを抑制することができる。
【００９７】
次に、図１１のフローチャートを参照して、図４のステップＳ６において実行される彩度補正部２５の処理について説明する。
【００９８】
ステップＳ８１において、彩度補正部２５は、輝度補正部２４から供給されてきた画像の記憶色領域の彩度値と、補正後の彩度値（記憶色の彩度値）を取得する。例えば、彩度補正部２５には、記憶色である肌色の彩度値、空色の彩度値、緑色の彩度値などが予め設定されている。
【００９９】
ステップＳ８２において、彩度補正部２５は、画像に含まれる記憶色領域の彩度値を、記憶色の彩度値そのものに補正する補正カーブを決定する。
【０１００】
図１２は、彩度補正部２５により決定される補正カーブの例を示す図である。図１２において、横軸は補正前の彩度値（入力彩度値）を表わしており、縦軸は補正量（出力彩度値から入力彩度値を減算した量（ΔＣ））を表わしている。
【０１０１】
この例においては、入力彩度値が「０．２」の画素に対して、約「０．０６」だけ彩度値が加算され、その入力彩度値をほぼ中心として、入力彩度値の大小方向に補正量が次第に小さくなるような補正カーブとされている。
【０１０２】
図１３は、彩度補正部２５により決定される補正カーブの例を示す図である。図１３において、横軸は入力彩度値を表わしており、縦軸は出力彩度値を表わしている。
【０１０３】
図１３に示される補正カーブは、図１２の補正カーブと同様に、彩度値が「０．２」の画素に対して、「０．０６」だけ彩度値が加算され、「０．２６」の彩度値の画素が出力されるものとされている。また、入力彩度値「０．２」をほぼ中心として、入力彩度値の大小方向に補正量が次第に小さくなるような補正カーブとされている。
【０１０４】
図１４は、彩度補正部２５により決定される補正カーブの例を示す図である。図１４において、横軸は入力彩度値を表わしており、縦軸は彩度値のゲイン（出力彩度値／入力彩度値）を表わしている。
【０１０５】
図１４においては、図１２、および図１３の場合と同様に、入力彩度値「０．２」の画素が彩度値「０．２６」として出力されるように、ゲインとして「１．３」をとる補正カーブが表わされている。
【０１０６】
当然、彩度補正部２５により作成される補正カーブは、図１２乃至図１４に示されるものに限定されるものではなく、諧調飛びが発生しない滑らかなカーブであれば、様々なカーブを補正カーブとして利用することができる。
【０１０７】
図１１の説明に戻り、彩度補正部２５は、ステップＳ８３において、画素毎の色相値を確認し、ステップＳ８４において、確認した画素が記憶色領域の画素か否かを判定する。
【０１０８】
彩度補正部２５は、ステップＳ８４において、確認した画素が記憶色領域の画素であると判定した場合、ステップＳ８５に進み、ステップＳ８３で決定した補正カーブに基づいて、その画素を補正する。従って、図１２、および図１３に示されるような補正カーブが決定されている場合、彩度値が「０．２」の記憶色領域の画素は、「０．２６」の彩度値の画素に補正される。
【０１０９】
一方、ステップＳ８４において、確認した画素が記憶色領域の画素でないと判定された場合、ステップＳ８５の処理はスキップされる。
【０１１０】
ステップＳ８６において、彩度補正部２５は、全ての画素を確認したか否かを判定し、全ての画素を確認していないと判定した場合、ステップＳ８３に戻り、それ以降の処理を繰り返し実行し、一方、全ての画素を確認したと判定した場合、彩度を補正した画像データを色相補正部２６に出力する。そして、図４のステップＳ７以降の処理が実行される。
【０１１１】
次に、図１５のフローチャートを参照して、図４のステップＳ７において実行される色相補正部２６の処理について説明する。
【０１１２】
ステップＳ９１において、色相補正部２６は、彩度補正部２５から供給されてきた画像の記憶色領域の色相値と、補正後の色相値（記憶色の色相値）を取得する。例えば、色相補正部２６には、記憶色である肌色の色相値、空色の色相値、緑色の色相値などが予め設定されている。
【０１１３】
ステップＳ９２において、色相補正部２６は、画像に含まれる記憶色領域の色相値を、記憶色の色相値に補正する補正カーブを決定する。
【０１１４】
図１６は、色相補正部２６により決定される、肌色領域に対する補正カーブの例を示す図である。図１６において、横軸は補正前の色相値（入力色相値）を表わしており、縦軸は補正量（出力色相値から入力色相値を減算した量（ΔＨ））を表わしている。
【０１１５】
この例においては、「８０乃至１５０」の範囲の色相値により指定される画素に対して補正が施されるとされており、補正カーブは、２つのべき関数が組み合わされて作成されている。この補正カーブによれば、例えば、入力色相値が「１２０」の肌色の画素は、色相値が約「１０」だけ加算され、補正される。
【０１１６】
図１７は、色相補正部２６により決定される肌色領域に対する補正カーブの例を示す図であり、図1６に示される補正カーブに対して、横軸を、入力色相値を表わすものとし、縦軸を、出力色相値を表わすものとしたものである。
【０１１７】
従って、図１７に示される補正カーブは、図１６と同様に、「８０乃至１５０」の範囲の色相値をとる肌色の画素に対して補正が施されるとされており、この補正カーブによれば、例えば、入力色相値が「１２０」の画素は、色相値が約「１０」だけ加算され、補正される。
【０１１８】
図１８、および図１９に示される補正カーブは、図１６、および図１７に示される補正カーブに対して、それぞれ、べき関数の指数を変更したものであり、このように、様々な補正カーブを色相値の補正に利用することができる。
【０１１９】
図２０は、色相補正部２６により決定される緑色領域に対する補正カーブの例を示す図である。図２０において、横軸は補正前の色相値（入力色相値）を表わしており、縦軸は補正量（ΔＨ）を表わしている。
【０１２０】
この例においては、「１６５乃至２５５」の範囲の色相値の画素に対して補正が施されるとされており、補正カーブは、２つのべき関数が組み合わされて作成されている。この補正カーブによれば、例えば、入力色相値が「２０５」の画素は、色相値が約「４」だけ減算され、補正される。
【０１２１】
図２１は、色相補正部２６により決定される緑色領域に対する補正カーブの例を示す図であり、図２０に示される補正カーブに対して、横軸を、入力色相値を表わすものとし、縦軸を、出力色相値を表わすものとしたものである。
【０１２２】
図２２は、色相補正部２６により決定される空色領域に対する補正カーブの例を示す図である。図２２において、横軸は補正前の色相値（入力色相値）を表わしており、縦軸は補正量（ΔＨ）を表わしている。
【０１２３】
この例においては、「２８０乃至３４０」の範囲の色相値により指定される画素に対して補正が施されるとされており、補正カーブは、２つのべき関数が組み合わされて作成されている。この補正カーブによれば、例えば、入力色相値が「２９０」の画素は、色相値が約「５」だけ加算され、補正される。
【０１２４】
図２３は、色相補正部２６により決定される空色領域に対する補正カーブの例を示す図であり、図２２に示される補正カーブに対して、横軸を、入力色相値を表わすものとし、縦軸を、出力色相値を表わすものとしたものである。
【０１２５】
当然、色相補正部２６により作成される補正カーブは、図１６乃至図２３に示されるものに限定されるものではなく、諧調飛びが発生しない滑らかなカーブであれば、様々なカーブを補正カーブとして利用することができる。
【０１２６】
図１５の説明に戻り、色相補正部２６は、ステップＳ９３において、画素毎の色相値を確認し、ステップＳ９４において、確認した画素が記憶色の画素か否かを判定する。
【０１２７】
色相補正部２６は、ステップＳ９４において、確認した画素が記憶色の画素であると判定した場合、ステップＳ９５に進み、ステップＳ９３で決定した補正カーブに基づいて、その画素を補正する。従って、図１６、および図１７に示されるような補正カーブが決定されている場合、色相値が「１２０」の記憶色の画素は、ほぼ「１３０」の色相値の画素に補正される。
【０１２８】
一方、ステップＳ９４において、確認した画素が記憶色の画素でないと判定された場合、ステップＳ９５の処理はスキップされる。
【０１２９】
ステップＳ９６において、色相補正部２６は、画像の全ての画素を確認したか否かを判定し、全ての画素を確認していないと判定した場合、ステップＳ９３に戻り、それ以降の処理を繰り返し実行し、一方、全ての画素を確認したと判定した場合、画像データを変換部２７に出力する。そして、図４のステップＳ８以降の処理が実行され、輝度値、彩度値、および色相値が補正された画像が変換部２７によりRGB色空間のものに変換され、出力される。
【０１３０】
以上においては、基本的に、画像の撮影モードに基づいて画像のカテゴリを判別し、それに基づいてグローバル補正、またはローカル補正のいずれかを選択させ、補正量を決定するとしたが、画像を観察するユーザの年齢や性別等によっても記憶色の認識の仕方が異なるため、例えば、ディジタルスチルカメラ４にユーザの年齢や性別等の様々なユーザ情報が設定され、それがヘッダ情報に含まれている場合、そのような様々な情報に基づいて、補正方法や補正量を決定するようにしても良い。
【０１３１】
上述した一連の処理は、ハードウェアにより実行させることもできるが、ソフトウェアにより実行させることもできる。この場合、そのソフトウェアを実行させる装置は、例えば、図２４に示されるようなパーソナルコンピュータにより構成される。
【０１３２】
図２４において、CPU（Central Processing Unit）５１は、ROM（Read Only Memory）５２に記憶されているプログラム、または、記憶部５８からRAM（Random Access Memory）５３にロードされたプログラムに従って各種の処理を実行する。RAM５３にはまた、CPU５１が各種の処理を実行する上において必要なデータなどが適宜記憶される。
【０１３３】
CPU５１、ROM５２、およびRAM５３は、バス５４を介して相互に接続されている。このバス５４にはまた、入出力インタフェース５５も接続されている。
【０１３４】
入出力インタフェース５５には、キーボード、マウスなどよりなる入力部５６、CRT，LCDなどよりなるディスプレイ、並びにスピーカなどよりなる出力部５７、ハードディスクなどより構成される記憶部５８、モデム、ターミナルアダプタなどより構成される通信部５９が接続されている。通信部５９は、ネットワークを介しての通信処理を行う。
【０１３５】
入出力インタフェース５５にはまた、必要に応じてドライブ５０が接続され、磁気ディスク６１、光ディスク６２、光磁気ディスク６３、或いは半導体メモリ６４などが適宜装着され、それから読み出されたコンピュータプログラムが、必要に応じて記憶部５８にインストールされる。
【０１３６】
一連の処理をソフトウェアにより実行させる場合には、そのソフトウェアを構成するプログラムが、専用のハードウェアに組み込まれているコンピュータ、または、各種のプログラムをインストールすることで、各種の機能を実行することが可能な、例えば、汎用のパーソナルコンピュータなどに、ネットワークや記録媒体からインストールされる。
【０１３７】
この記録媒体は、図２４に示されるように、装置本体とは別に、ユーザにプログラムを提供するために配布される、プログラムが記録されている磁気ディスク６１（フロッピディスクを含む）、光ディスク６２（CD-ROM(Compact Disk-Read Only Memory)，DVD(Digital Versatile Disk)を含む）、光磁気ディスク６３（MD（登録商標）(Mini-Disk)を含む）、もしくは半導体メモリ６４などよりなるパッケージメディアにより構成されるだけでなく、装置本体に予め組み込まれた状態でユーザに提供される、プログラムが記録されているROM５２や、記憶部５８に含まれるハードディスクなどで構成される。
【０１３８】
なお、本明細書において、記録媒体に記録されるプログラムを記述するステップは、記載された順序に従って時系列的に行われる処理はもちろん、必ずしも時系列的に処理されなくとも、並列的あるいは個別に実行される処理をも含むものである。
【０１３９】
【発明の効果】
以上のように、本発明の画像処理装置および方法、記録媒体、並びにプログラムによれば、擬似輪郭等を発生させることなく、画像の色をより好適に補正することができる。
【図面の簡単な説明】
【図１】本発明の画像処理装置が適用される機器の例を示す図である。
【図２】本発明を適用した画像処理装置の構成例を示すブロック図である。
【図３】図２の輝度補正部の構成例を示すブロック図である。
【図４】図２の画像処理装置の処理を説明するフローチャートである。
【図５】図２のカテゴリ判別部の処理を説明するフローチャートである。
【図６】図２のカテゴリ判別部の他の処理を説明するフローチャートである。
【図７】図２のカテゴリ判別部のさらに他の処理を説明するフローチャートである。
【図８】図２の輝度補正部の処理を説明するフローチャートである。
【図９】図８の処理により作成される補正カーブの例を示す図である。
【図１０】図８の処理により作成される補正カーブの他の例を示す図である。
【図１１】図２の彩度補正部の処理を説明するフローチャートである。
【図１２】図１１の処理により作成される補正カーブの例を示す図である。
【図１３】図１１の処理により作成される補正カーブの他の例を示す図である。
【図１４】図１１の処理により作成される補正カーブのさらに他の例を示す図である。
【図１５】図２の色相補正部の処理を説明するフローチャートである。
【図１６】図１５の処理により作成される補正カーブの例を示す図である。
【図１７】図１５の処理により作成される補正カーブの他の例を示す図である。
【図１８】図１５の処理により作成される補正カーブのさらに他の例を示す図である。
【図１９】図１５の処理により作成される補正カーブの例を示す図である。
【図２０】図１５の処理により作成される補正カーブの他の例を示す図である。
【図２１】図１５の処理により作成される補正カーブのさらに他の例を示す図である。
【図２２】図１５の処理により作成される補正カーブの例を示す図である。
【図２３】図１５の処理により作成される補正カーブの他の例を示す図である。
【図２４】パーソナルコンピュータの構成例を示すブロック図である。
【符号の説明】
１１ 画像処理装置， ２１ 画像解析部， ２２ カテゴリ判別部， ２３変換部， ２４ 輝度補正部， ２５ 彩度補正部， ２６ 色相補正部， ２７ 変換部， ２８ 出力部， ４１ 補正切替部， ４２ 補正量決定部，４３ 記憶色領域抽出部， ４４ 補正処理部， ６１ 磁気ディスク， ６２ 光ディスク， ６３ 光磁気ディスク， ６４ 半導体メモリ[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing apparatus and method, a recording medium, and a program, and more particularly, to an image processing apparatus and method and recording capable of more suitably correcting the color of an image without generating a pseudo contour or the like. The present invention relates to a medium and a program.
[0002]
[Prior art]
In recent years, devices capable of handling digital color images, such as storing images taken with a digital still camera in a personal computer or outputting the stored images to a television receiver or printer, are rapidly spreading. .
[0003]
Among these devices, there are devices that correct a color close to a so-called memory color included in an image to the memory color itself and output the obtained image.
[0004]
Memory colors are colors that remain in human memory, such as human skin color, sky blue, and plant green, and the viewer's eyes for these colors are sensitive to the viewer. It is said that an image obtained by correcting a memory color is recognized as a better-looking image than an image obtained by photographing.
[0005]
By the way, in a device that outputs an image in which a color close to the memory color is corrected to the memory color itself, generally, a method of extracting a color area close to the memory color and correcting only that area is employed. . Therefore, for example, when a person is shown in the image, the color of the face area of the subject is corrected to the skin color of the memory color and output.
[0006]
[Problems to be solved by the invention]
However, when a memory color area is extracted in this way and local correction is performed only on that area, the color change from the area that has not been corrected becomes abrupt and a pseudo contour is generated at the boundary. There was a problem of doing it.
[0007]
The present invention has been made in view of such a situation, and is capable of more suitably correcting the color of an image without generating a pseudo contour or the like.
[0008]
[Means for Solving the Problems]
  The image processing apparatus of the present inventionBased on the size of a plurality of predetermined memory color areas in the entire image data,Discriminating means for discriminating the category and images discriminated by the discriminating meansFor the entire dataDepending on the category, only correction of the memory color area that approximates the predetermined memory color, or imagedataSelection means for selecting one of the corrections of the whole, and selection by the selection meansDepending on the image dataCorrection means for correctingThe
  When the size of the skin color area as the memory color area is larger than the first threshold value and the size of the green color area and the sky blue area is smaller than the second threshold value, the determination means determines the category of the image data as a portrait. It can be discriminated from an image.
  When the size of the skin color area as the memory color area is larger than the first threshold value and the size of the green color area and the sky blue area is larger than the second threshold value, the determination means determines the category of the image data as a portrait. & Can be discriminated from landscape images.
  The discrimination means has a size of the skin color area as the memory color area that is smaller than a first threshold value and larger than a second threshold value that is smaller than the first threshold value. When the size is smaller than the third threshold, the category of the image data can be determined as a small person image.
  The discrimination means has a size of the skin color area as the memory color area that is smaller than a first threshold value and larger than a second threshold value that is smaller than the first threshold value. When the size is larger than the third threshold, the category of the image data can be determined as a small person & landscape image.
  The discriminating unit determines the category of the image data when the size of the skin color area as the memory color area is smaller than a first threshold and the sizes of the green area and the sky blue area are larger than a second threshold. Can be determined.
[0009]
  The selection means is an imagedataAs the brightness correction, only the memory color area or the imagedataOne of the entire corrections of the image is selected, and the correction unit selects the image according to the selection by the selection unit.dataThe luminance can be corrected.
[0010]
Extraction means for extracting the memory color area is further provided, and the correction means can correct the saturation and hue of the memory color area extracted by the extraction means.
[0011]
  The discriminating means is an imageFor the entire dataCategoryfurther,Information described in the image data headerOn the basis of theIt can be determined.
[0012]
  An image processing method of the image processing apparatus according to the present invention includes:Based on the size of a plurality of predetermined memory color areas in the entire image data,Discrimination step for discriminating categories and images discriminated by the processing of the discrimination stepFor the entire dataDepending on the category, only correction of the memory color area that approximates the predetermined memory color, or imagedataA selection step to select one of the total corrections, and selection by the processing of the selection stepDepending on the image dataA correction step for correctingIncluding.
[0013]
  The program recorded on the recording medium of the present invention is:Based on the size of a plurality of predetermined memory color areas in the entire image data,Discrimination step for discriminating categories and images discriminated by the processing of the discrimination stepFor the entire dataDepending on the category, only correction of the memory color area that approximates the predetermined memory color, or imagedataA selection step to select one of the total corrections, and selection by the processing of the selection stepDepending on the image dataA correction step for correctingIncluding.
[0014]
  The program of the present inventionBased on the size of a plurality of predetermined memory color areas in the entire image data,Discrimination step for discriminating categories and images discriminated by the processing of the discrimination stepFor the entire dataDepending on the category, only correction of the memory color area that approximates the predetermined memory color, or imagedataA selection step to select one of the total corrections, and selection by the processing of the selection stepDepending on the image dataCorrection step to correctLet
[0015]
  In the image processing apparatus and method and the program of the present invention,Based on the size of a plurality of predetermined memory color areas in the entire image data,Category is determined and the image is determinedFor the entire dataDepending on the category, only correction of the memory color area that approximates the predetermined memory color, or imagedataOne of the total corrections is selected. And the selection resultDepending on the image dataIs corrected.
[0016]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a diagram illustrating an example of a device to which an image processing apparatus according to the present invention can be applied.
[0017]
As shown in FIG. 1, the image processing apparatus 11 (see FIG. 2) includes, for example, a personal computer 1, a projector 2, a digital video camera 3, a digital still camera 4, a PDA (Personal Digital Assistants) 5, a mobile phone 6, In addition, the present invention can be applied to the image data processing unit of each device such as the printer 7.
[0018]
FIG. 2 is a block diagram showing a configuration example of the image printing apparatus 11 to which the present invention is applied.
[0019]
The image analysis unit 21 acquires an image file captured and stored by the digital still camera 4 or the like, and analyzes header information of the image file. In the header information, for example, Exif (Exchangeable image file format for Digital Still Cameras) indicating information such as the model name of the camera that has taken the image, the shooting date and time, the shutter speed, and the shooting mode is described. Analyze information such as this Exif. For example, when the shooting mode is set by the user of the digital still camera 4 and an image is shot, the image analysis unit 21 reads the set mode from the header information and notifies the category determination unit 22 of the shooting mode of the image. To do.
[0020]
Further, the image analysis unit 21 calculates a statistic of the memory color included in the image data and notifies the category determination unit 22 of it. Specifically, the image analysis unit 21 approximates to a predetermined memory color such as a human skin area (skin color area), a landscape sky area (sky blue area), a green area of plants (green area), for example. The ratio of the color area (hereinafter, simply referred to as a memory color area) in the entire image is calculated and notified to the category determination unit 22. The acquired image data is also output to the category determination unit 22.
[0021]
The category determination unit 22 determines the category of the image based on the notification from the image analysis unit 21. For example, the category determination unit 22 determines whether the size (amount) of the skin color area included in the image is larger than a predetermined threshold or whether the size of the sky blue area is larger than the threshold. Determine the category.
[0022]
As will be described later, the image supplied from the image analysis unit 21 includes, for example, a “portrait image” mainly including a person as a subject, a “landscape image” mainly including a landscape, and a ratio of persons. Are classified into categories such as “small person images”. The category determined by the category determination unit 22 is used when selecting a correction method.
[0023]
The result of category discrimination by the category discrimination unit 22 is notified to the luminance correction unit 24. Further, the image data supplied to the category determination unit 22 is output to the conversion unit 23.
[0024]
The conversion unit 23 designates the supplied image data designated by R (red), G (green), and B (blue) by Y (luminance), C (saturation), and H (hue). The image data is converted into image data, and the image data obtained by converting the color space is output to the luminance correction unit 24. By changing the color space, for example, if only the saturation is to be changed, only the saturation value may be changed, and the color can be changed more directly and easily than in the RGB space. That is, when changing the saturation in the RGB space, it is necessary to change each value of RGB.
[0025]
The luminance correction unit 24 determines the correction amount based on the category of the image notified from the category determination unit 22, the luminance value of the image data supplied from the conversion unit 23, the luminance value of the memory color, and the like, for example, the storage The luminance value of the color area is corrected for each pixel. The image data whose luminance has been corrected by the luminance correction unit 24 is output to the saturation correction unit 25. The detailed configuration and processing of the luminance correction unit 24 will be described later.
[0026]
The saturation correction unit 25 determines a correction amount based on the saturation of the image data supplied from the luminance correction unit 24 and the saturation of the memory color set in advance, and the saturation of the memory color area is determined for each pixel. To correct. The image data whose saturation is corrected by the saturation correction unit 25 is output to the hue correction unit 26.
[0027]
The hue correction unit 26 determines a correction amount based on the hue of the image data supplied from the saturation correction unit 25 and the hue of the preset memory color, and corrects the hue of the memory color area for each pixel. . The image data whose hue has been corrected by the hue correction unit 26 is output to the conversion unit 27.
[0028]
The conversion unit 27 acquires image data whose luminance, saturation, and hue are corrected from the hue correction unit 26, and converts the acquired image data into an RGB color space. The image data converted into the RGB color space is supplied to the output unit 28, and is output to the subsequent device of the image processing device 11 by the output unit 28. That is, when the image processing apparatus 11 is provided in the personal computer 1 of FIG. 1, an image whose color is corrected by the image processing apparatus 11 is a CRT (Cathode Ray Tube), LCD (Liquid Crystal Display), etc. of the personal computer 1. Is displayed. When the image processing apparatus 11 is provided in the printer 7 of FIG. 1, an image whose color has been corrected by the image processing apparatus 11 is printed on a predetermined print medium.
[0029]
FIG. 3 is a block diagram illustrating a detailed configuration example of the luminance correction unit 24 of FIG.
[0030]
The correction switching unit 41 corrects the luminance of the pixels of the entire image based on the category of the image notified from the category determining unit 22, the statistic of the memory color region included in the image, or the memory color region ( It is determined whether to correct only the brightness of the peripheral portion as necessary. Hereinafter, the correction method for correcting the luminance of the pixels of the entire image is referred to as global correction, while the correction method for correcting only the luminance of the memory color area is referred to as local correction.
[0031]
Specifically, the correction switching unit 41 is based on a notification from the category determination unit 22 or the like, and the category of the image supplied from the conversion unit 23 is a “portrait image” in which a person is mainly captured as a subject. Alternatively, when it is determined that the image is a “small person image”, global correction is selected as a correction method to be executed by the correction processing unit 44 in the subsequent stage. If only the brightness of the skin color area of a person included in the “portrait image” is corrected, the brightness difference between the skin color area and the surrounding area may become abrupt, and a pseudo contour may be generated. As described above, by selecting the global correction for the “portrait image” and the “small person image”, it is possible to more suitably correct the color without generating a pseudo contour or the like.
[0032]
When the correction switching unit 41 determines that the category of the image supplied from the conversion unit 23 is, for example, a landscape image, the correction switching unit 41 selects local correction as a correction method to be executed by the correction processing unit 44.
[0033]
Information indicating the correction method selected by the correction switching unit 41 is notified to the correction amount determination unit 42 together with the image data supplied from the conversion unit 23.
[0034]
The correction amount determination unit 42 determines the luminance correction amount (correction curve) based on the information notified from the correction switching unit 41, and when it is notified that the image is corrected by local correction, the image data Is output to the memory color area extraction unit 43 together with information representing the correction amount. On the other hand, when it is notified that the correction is to be made by global correction, the image data is output to the correction processing unit 44 together with information representing the correction amount. To do.
[0035]
A range of colors to be extracted is set in advance in the memory color area extraction unit 43. The memory color area extraction unit 43, for example, a color within a predetermined range that approximates the skin color of the memory color, A color within a predetermined range that approximates the sky color of the memory color and an area (pixel) of a color within the predetermined range that approximates the green color of the memory color are extracted, and the memory color The area is notified to the correction processing unit 44. In addition, information indicating the correction amount of the local correction is also output to the correction processing unit 44.
[0036]
When the image data is supplied from the correction amount determination unit 42, the correction processing unit 44 corrects the pixels of the entire image based on the notified correction amount (performs global correction). Further, when the image data is supplied from the memory color area extraction unit 43, the correction processing unit 44 corrects the pixels in the memory color area of the image based on the notified correction amount (performs local correction). The image data that has been subjected to global correction or local correction by the correction processing unit 44 is output to the subsequent saturation correction unit 25.
[0037]
Next, correction processing of the image processing apparatus 11 will be described with reference to the flowchart of FIG.
[0038]
In step S1, the image analysis unit 21 acquires image data, and describes, for example, information about the image represented by the data, such as the model name of the photographed camera, the photographing date and time, the shutter speed, and the photographing mode. Analyze information. The image analysis unit 21 analyzes the header information, and notifies the category determination unit 22 of the image capturing mode described in the header information.
[0039]
In step S <b> 2, the image analysis unit 21 calculates a statistic of the memory color area included in the image and notifies the category determination unit 22 of the calculation result. For example, a color to be extracted as a memory color area is set in advance in the image analysis unit 21, and the image analysis unit 21 has a color that approximates the skin color of the memory color and a color that approximates the sky color of the memory color. The statistic of the memory color area such as the ratio of the pixels designated by the above to the entire image is calculated.
[0040]
In step S <b> 3, the category determination unit 22 determines the category of the image based on the header information notified from the image analysis unit 21 and the statistics of the memory color area. The result of category discrimination by the category discrimination unit 22 is notified to the luminance correction unit 24, and the image data supplied to the category discrimination unit 22 is output to the conversion unit 23. The processing of the category determination unit 22 will be described later with reference to the flowcharts of FIGS.
[0041]
In step S <b> 4, the conversion unit 23 converts the image data in the RGB color space into image data in the YCH color space, and outputs the obtained image data to the luminance correction unit 24.
[0042]
In step S <b> 5, the brightness correction unit 24 corrects the brightness of the image supplied from the conversion unit 23 based on the category notified from the category determination unit 22, and outputs the obtained image to the saturation correction unit 25. To do.
[0043]
Thereafter, in step S6 and step S7, the saturation correction unit 25 and the hue correction unit 26 respectively perform correction, and the obtained image data is output to the conversion unit 27. The image data supplied to the conversion unit 27 is converted into RGB color space data in step S8, and then output to a device subsequent to the image processing device 11 in step S9.
[0044]
Next, with reference to the flowchart of FIG. 5, the process of the category discrimination | determination part 22 performed in step S3 of FIG. 4 is demonstrated.
[0045]
This process represents a process of determining the category of an image shot in the “portrait mode”. That is, in the digital still camera 4 or the like, when the shooting mode is set to “portrait mode”, the information described in the header of the image is analyzed by the image analysis unit 21 and the “portrait mode” is used. The category determination unit 22 is notified that the image has been taken.
[0046]
In step S21, the category discriminating unit 22 determines that the amount of green area and sky blue area (turfgrass sky area) such as grass or grass based on the statistics of the memory color area calculated by the image analysis unit 21 is the threshold TH1. Determine if greater than. That is, a predetermined threshold value is set in the category determination unit 22 in advance, and the category of the image can be determined based on the predetermined threshold value. Note that this threshold value may be set by the user according to his / her preference.
[0047]
When the category determination unit 22 determines in step S21 that the amounts of the green region and the sky blue region are not larger (smaller) than the threshold value TH1, the supplied image is an image in which a person is captured in a relatively wide range. In step S22, the image category is determined as “portrait image”.
[0048]
On the other hand, if the category determination unit 22 determines in step S21 that the amounts of the green region and the sky blue region are larger than the threshold value TH1, the supplied image is an image in which the landscape is reflected in a relatively wide range. In step S23, the image category is determined as “portrait & landscape image”.
[0049]
The category of the image determined by the category determination unit 22 in step S22 and step S23 is notified to the luminance correction unit 24, and thereafter, the processing after step S4 in FIG. 4 is executed.
[0050]
Next, another process of the category determination unit 22 executed in step S3 of FIG. 4 will be described with reference to the flowchart of FIG.
[0051]
This process represents an example of a process for determining the category of an image shot in the “landscape mode”.
[0052]
The processing shown in FIG. 6 is basically the same as that shown in FIG. 5, and the category determination unit 22 is based on the statistic of the memory color area calculated by the image analysis unit 21 in step S31. Thus, it is determined whether or not the skin color area is larger than the threshold value TH11.
[0053]
If the category determination unit 22 determines in step S31 that the skin color area is larger than the threshold TH11, the category determination unit 22 proceeds to step S32 and recognizes that the supplied image is an image in which a person is captured in a relatively wide range. In step S32, the image category is determined as “small person & landscape image”.
[0054]
In step S31, if the category determination unit 22 determines that the skin color area is smaller than the threshold TH11, the supplied image is recognized as an image in which a person is hardly captured, and the process proceeds to step S33. Is determined as “landscape image”.
[0055]
The category of the image determined by the category determination unit 22 in step S32 and step S33 is notified to the luminance correction unit 24, and thereafter, the processing after step S4 in FIG. 4 is executed.
[0056]
Next, with reference to the flowchart of FIG. 7, the process of the category determination part 22 performed in step S3 of FIG. 4 will be described.
[0057]
This process represents a process of determining the category of an image taken without setting any mode.
[0058]
In step S41, the category determination unit 22 determines whether the amount of the skin color area is larger than the threshold value TH31 based on the statistic of the memory color area calculated by the image analysis unit 21.
[0059]
If the category determination unit 22 determines in step S41 that the amount of the skin color area is greater than the threshold value TH31, the process proceeds to step S42. Next, the amount of the green area and the sky blue area such as grass or grass is greater than the threshold value TH32. It is determined whether or not.
[0060]
When the category determination unit 22 determines in step S42 that the amounts of the green region and the sky blue region are not larger (smaller) than the threshold value TH32, the supplied image is an image in which a person is captured in a relatively wide range. In step S43, the image category is determined as “portrait image”.
[0061]
On the other hand, in step S42, when the category determination unit 22 determines that the amount of the green region and the sky blue region is larger than the threshold value TH32, the supplied image is recognized as an image in which a landscape is reflected in a relatively wide range. In step S44, the image category is determined as “portrait & landscape image”.
[0062]
If the category determination unit 22 determines in step S41 that the amount of the skin color region is smaller than the threshold value TH31, the process proceeds to step S45, and then whether the amount of the skin color region is larger than the threshold value TH33 that is a value smaller than the threshold value TH31 Determine whether or not.
[0063]
In step S45, if the category determination unit 22 determines that the amount of the skin color area is greater than the threshold TH33, the process proceeds to step S46, and then determines whether the amount of the green area and the sky blue area is greater than the threshold TH34. .
[0064]
If the category determination unit 22 determines in step S46 that the amount of the green region and the sky blue region is smaller than the threshold value TH34, the category determination unit 22 recognizes that the image includes a certain amount of people, and proceeds to step S47 to determine the category of the image. Determined as “small person image”.
[0065]
On the other hand, if the category determination unit 22 determines in step S46 that the amounts of the green region and sky blue region are larger than the threshold TH34, the category determination unit 22 recognizes that the image includes a certain amount of people and scenery as subjects, and the process proceeds to step S47. Proceed to determine the category of the image as “small person & landscape image”.
[0066]
In step S45, if the category determination unit 22 determines that the amount of the skin color area is smaller than the threshold TH33, the process proceeds to step S49, and then determines whether the amounts of the green area and the sky blue area are larger than the threshold TH35. .
[0067]
If the category determination unit 22 determines in step S49 that the amounts of the green region and the sky blue region are larger than the threshold TH35, the category determination unit 22 proceeds to step S50 and determines the category of the image as “landscape image”.
[0068]
On the other hand, if the category determination unit 22 determines in step S49 that the amount of the green area and the sky blue area is smaller than the threshold value TH35, the process proceeds to step S51, and the supplied image does not include the memory color area. Recognize that An image recognized by the category determination unit 22 as an image that does not include a memory color area is output to a subsequent processing unit without being corrected by the luminance correction unit 24 or the like.
[0069]
Thus, even for an image taken without setting a mode, the category of the image can be determined based on the amount of the memory color area included in the image. Of course, when detailed information about the category of the image is described in the header, the category of the image may be determined based only on the information.
[0070]
Next, the process of the luminance correction unit 24 executed in step S5 of FIG. 4 will be described with reference to the flowchart of FIG.
[0071]
In step S <b> 61, the correction switching unit 41 of the luminance correction unit 24 determines whether the category of the supplied image is “portrait image” or “small person image” based on the notification from the category determination unit 22. judge.
[0072]
For example, when the process shown in FIG. 7 is executed by the category determination unit 22 and it is notified that the category of the image is “portrait image” or “small person image”, the brightness correction unit 24 In step S62, global correction is selected as a correction method for the image supplied from the conversion unit 23.
[0073]
The correction switching unit 41 outputs the image data supplied from the conversion unit 23 to the correction amount determination unit 42 and notifies the correction amount determination unit 42 that correction is performed by global correction.
[0074]
In step S63, the correction amount determination unit 42 acquires the luminance value of the image before correction and the luminance value of the image after correction, proceeds to step S64, and determines a correction curve (correction amount). For example, the correction amount determination unit 42 acquires, as the luminance value of the image before correction, the average value of the luminance values of the entire image before correction, the average value of the luminance values of the memory color area included in the image, and the like. As the luminance value of the subsequent image, an average value of the luminance values of the memory color area is acquired.
[0075]
FIG. 9 is a diagram illustrating an example of a correction curve for global correction determined by the correction amount determination unit 42. In FIG. 9, the horizontal axis represents the luminance value before correction (input luminance value), and the vertical axis represents the luminance value after correction (output luminance value).
[0076]
In FIG. 9, a correction curve is created by combining two binomial distribution functions. According to this correction curve, pixels of an input image having a luminance value near “0.2 to 0.8” are On the other hand, the pixel of the input image having a luminance value near “0.0 to 0.2” is corrected to a pixel value lower than that in order to maintain the contrast. For example, in the image before correction, a pixel designated by a luminance value of “0.4” is corrected to a pixel designated by a luminance value of “0.48”.
[0077]
Naturally, the correction curve created by the correction amount determination unit 42 is not limited to that shown in FIG. 9 and is a smooth curve (a linear differential function and a secondary differential function are continuous without gradation skipping). Curve), various curves can be used as the correction curve.
[0078]
Returning to FIG. 8, the correction processing unit 44 is notified of the correction curve determined by the correction amount determination unit 42 in step S64.
[0079]
When the image to be subjected to global correction and the determined correction curve are notified from the correction amount determining unit 42, the correction processing unit 44 uses the notified correction curve in step S65 to Each pixel is corrected. Therefore, when the correction curve shown in FIG. 9 is notified, as described above, for example, a pixel having the luminance value “0.4” is corrected to a pixel having the luminance value “0.48”.
[0080]
In this way, all the pixels of the image are corrected, and the obtained image data is output to the saturation correction unit 25 in the subsequent stage. And the process after step S6 of FIG. 4 is performed.
[0081]
On the other hand, if the correction switching unit 41 determines in step S61 that the category of the supplied image is not “portrait image” or “small person image” based on the notification from the category determination unit 22, the process proceeds to step S66. Then, the local correction is selected as a correction method for the image supplied from the conversion unit 23. The correction switching unit 41 outputs the image data supplied from the conversion unit 23 to the correction amount determination unit 42 and notifies the correction amount determination unit 42 that correction is performed by local correction.
[0082]
In step S67, the correction amount determination unit 42 acquires the luminance value of the image before correction and the luminance value of the image after correction, proceeds to step S68, and calculates the Mahalanobis distance for determining the correction curve.
[0083]
For example, if there is a color that is close to a predetermined red and is shifting in the yellow direction, and a color that is shifting in the blue direction, the Euclidean distance between the pixel value of those colors and the red pixel value of the memory color is the same. Even so, in general, the color shifted in the yellow direction will be felt closer to red by humans. The Mahalanobis distance is a distance calculated in consideration of such a human sense shift.
[0084]
Specifically, assuming that the difference between the luminance value and chromaticity of the pixel before correction and the luminance value and chromaticity of the pixel after correction are (ΔY, ΔCb, ΔCr), respectively, the correction amount determination unit 42 represents the following equation: The Mahalanobis distance is calculated by (1) and equation (2), and the process proceeds to step S69 to determine a correction curve as shown in FIG.
[0085]
[Expression 1]

[Expression 2]

[0086]
In FIG. 10, the horizontal axis represents the Mahalanobis distance from the center of the memory color area, and the vertical axis represents the correction amount (the luminance value after correction minus the luminance value before correction (ΔY)). .
[0087]
As shown in the figure, the correction amount determination unit 42 creates a smooth curve in which the correction amount gradually decreases as the Mahalanobis distance increases from the center of the memory color area, for example. In this example, normalization is performed so that the area to be corrected is within “1”, and areas that are not memory color areas are also corrected as necessary. For example, in the case where the memory color area is within the range of 0.6 distance on the Mahalanobis distance from the center of the memory color area, “0.6” to “1.0” from the center of the memory color area. An area that is not a memory color area at a distance is also corrected.
[0088]
In this way, by using a correction curve whose correction amount gradually decreases as the Mahalanobis distance increases, the periphery of the memory color area is also corrected, compared to the case where correction is performed only on the memory color area. Thus, the change in luminance can be smoothed, and the occurrence of a pseudo contour can be suppressed. That is, when the correction is performed only on the memory color area, the difference in luminance between the memory color area and the non-memory color area increases, and a pseudo contour may occur.
[0089]
Naturally, the correction curve created by the correction amount determination unit 42 is not limited to that shown in FIG. 10, and various curves are used as correction curves as long as they are smooth curves that do not cause gradation skipping. be able to.
[0090]
The correction curve created by the correction amount determination unit 42 is supplied to the memory color region extraction unit 43 together with the image data.
[0091]
In the memory color area extraction unit 43, the hue value of the memory color is set in advance, and in step S70, the memory color area extraction unit 43 confirms the hue value for each pixel, proceeds to step S71, and is set. Based on the hue value, the pixels that need to be corrected (the pixels in the memory color area and the surrounding pixels) are extracted. The memory color region extraction unit 43 outputs information on the pixel that needs correction to the correction processing unit 44 together with the correction curve determined by the correction amount determination unit 42 and the image data.
[0092]
In step S72, the correction processing unit 44 corrects the pixels in the memory color area extracted by the memory color area extraction unit 43 and, if necessary, the pixels around the memory color area based on the supplied correction curve. To do. For example, when a correction curve as shown in FIG. 10 has been supplied, the correction processing unit 44 corrects the luminance value of the pixel corresponding to the center of the memory color area by about “0.8” to be centered. The luminance value is corrected so that the correction amount gradually decreases as the Mahalanobis distance from the corresponding pixel increases.
[0093]
In step S73, the correction processing unit 44 determines whether or not all pixels that need to be corrected have been corrected. If it is determined that correction has not been performed, the correction processing unit 44 returns to step S70 and repeats the subsequent processing. On the other hand, if the correction processing unit 44 determines that all the pixels that need correction have been corrected, the correction processing unit 44 outputs the corrected image to the saturation correction unit 25 and ends the processing.
[0094]
As described above, since the correction method is switched according to the image category, it is possible to suppress the occurrence of a pseudo contour near the boundary of the memory color area. That is, when a portrait image in which a person is captured in a relatively wide range is supplied, and only the memory color area (skin color area) is extracted and the face color of the subject is corrected, the area other than the face area and face There is a case where a large difference occurs in the luminance value of the region, and a pseudo contour is generated at the boundary portion.
[0095]
Even when the entire image is corrected, as shown in FIG. 9, for pixels corresponding to the vicinity of the center of the luminance value, the luminance value is corrected so as to be higher and lower than the vicinity of the center. Since the correction is made so that the luminance value is further lowered with respect to the pixel for which the luminance value is set, the contrast of the entire image can be maintained.
[0096]
Furthermore, as shown in FIG. 10, when the memory color area is extracted and subjected to local correction, correction is performed on the pixels around the memory color area, so that only the memory color area is corrected. Compared to the case, it is possible to suppress a sudden change in the luminance value in the memory color area and the area other than the memory color area.
[0097]
Next, the processing of the saturation correction unit 25 executed in step S6 of FIG. 4 will be described with reference to the flowchart of FIG.
[0098]
In step S <b> 81, the saturation correction unit 25 acquires the saturation value of the storage color area of the image supplied from the luminance correction unit 24 and the corrected saturation value (saturation value of the storage color). For example, the saturation correction unit 25 is preset with a skin color saturation value, a sky blue saturation value, a green saturation value, and the like as memory colors.
[0099]
In step S82, the saturation correction unit 25 determines a correction curve for correcting the saturation value of the memory color area included in the image to the saturation value of the memory color itself.
[0100]
FIG. 12 is a diagram illustrating an example of a correction curve determined by the saturation correction unit 25. In FIG. 12, the horizontal axis represents the saturation value (input saturation value) before correction, and the vertical axis represents the correction amount (a value obtained by subtracting the input saturation value from the output saturation value (ΔC)). Yes.
[0101]
In this example, a saturation value of about “0.06” is added to a pixel having an input saturation value of “0.2”, and the input saturation value is approximately centered on the input saturation value. The correction curve is such that the correction amount gradually decreases in the large and small directions.
[0102]
FIG. 13 is a diagram illustrating an example of a correction curve determined by the saturation correction unit 25. In FIG. 13, the horizontal axis represents the input saturation value, and the vertical axis represents the output saturation value.
[0103]
The correction curve shown in FIG. 13 is similar to the correction curve shown in FIG. 12 except that a saturation value of “0.06” is added to a pixel with a saturation value of “0.2”. ”Is output. Further, the correction curve is such that the correction amount gradually decreases in the magnitude direction of the input saturation value, with the input saturation value “0.2” substantially as the center.
[0104]
FIG. 14 is a diagram illustrating an example of a correction curve determined by the saturation correction unit 25. In FIG. 14, the horizontal axis represents the input saturation value, and the vertical axis represents the gain of the saturation value (output saturation value / input saturation value).
[0105]
In FIG. 14, as in the case of FIGS. 12 and 13, the gain “1.3” is set so that the pixel with the input saturation value “0.2” is output as the saturation value “0.26”. Is shown.
[0106]
Naturally, the correction curve created by the saturation correction unit 25 is not limited to that shown in FIGS. 12 to 14, and various curves can be corrected as long as they are smooth curves with no gradation skipping. Can be used as
[0107]
Returning to the description of FIG. 11, the saturation correction unit 25 confirms the hue value of each pixel in step S83, and determines in step S84 whether the confirmed pixel is a pixel in the memory color area.
[0108]
If the saturation correction unit 25 determines in step S84 that the confirmed pixel is a pixel in the memory color area, the saturation correction unit 25 proceeds to step S85 and corrects the pixel based on the correction curve determined in step S83. Therefore, when the correction curves as shown in FIG. 12 and FIG. 13 are determined, the pixels in the memory color area with the saturation value “0.2” are the pixels with the saturation value “0.26”. It is corrected to.
[0109]
On the other hand, if it is determined in step S84 that the confirmed pixel is not a pixel in the memory color area, the process in step S85 is skipped.
[0110]
In step S86, the saturation correction unit 25 determines whether or not all the pixels have been confirmed. If it is determined that all the pixels have not been confirmed, the process returns to step S83, and the subsequent processing is repeatedly executed. On the other hand, if it is determined that all the pixels have been confirmed, the image data with the saturation corrected is output to the hue correction unit 26. And the process after step S7 of FIG. 4 is performed.
[0111]
Next, processing of the hue correction unit 26 executed in step S7 of FIG. 4 will be described with reference to the flowchart of FIG.
[0112]
In step S <b> 91, the hue correction unit 26 acquires the hue value of the memory color area of the image supplied from the saturation correction unit 25 and the corrected hue value (hue value of the memory color). For example, the hue correction unit 26 is preset with a skin color hue value, a sky blue hue value, a green hue value, and the like, which are memory colors.
[0113]
In step S92, the hue correction unit 26 determines a correction curve for correcting the hue value of the memory color area included in the image to the hue value of the memory color.
[0114]
FIG. 16 is a diagram illustrating an example of a correction curve for the skin color region determined by the hue correction unit 26. In FIG. 16, the horizontal axis represents the hue value before correction (input hue value), and the vertical axis represents the correction amount (amount (ΔH) obtained by subtracting the input hue value from the output hue value).
[0115]
In this example, it is assumed that correction is performed on the pixel specified by the hue value in the range of “80 to 150”, and the correction curve is created by combining two power functions. According to this correction curve, for example, a flesh color pixel having an input hue value of “120” is corrected by adding a hue value of about “10”.
[0116]
FIG. 17 is a diagram showing an example of a correction curve for the skin color area determined by the hue correction unit 26. The horizontal axis represents the input hue value with respect to the correction curve shown in FIG. Represents an output hue value.
[0117]
Accordingly, the correction curve shown in FIG. 17 is, like FIG. 16, corrected for flesh-colored pixels having a hue value in the range of “80 to 150”. For example, a pixel having an input hue value of “120” is corrected by adding a hue value of about “10”.
[0118]
The correction curves shown in FIGS. 18 and 19 are obtained by changing the exponents of the power functions with respect to the correction curves shown in FIGS. 16 and 17, respectively. It can be used to correct the hue value.
[0119]
FIG. 20 is a diagram illustrating an example of a correction curve for the green region determined by the hue correction unit 26. In FIG. 20, the horizontal axis represents a hue value (input hue value) before correction, and the vertical axis represents a correction amount (ΔH).
[0120]
In this example, correction is performed on pixels having hue values in the range of “165 to 255”, and the correction curve is created by combining two power functions. According to this correction curve, for example, a pixel having an input hue value of “205” is corrected by subtracting a hue value of about “4”.
[0121]
FIG. 21 is a diagram showing an example of a correction curve for the green region determined by the hue correction unit 26. The horizontal axis represents the input hue value with respect to the correction curve shown in FIG. Represents an output hue value.
[0122]
FIG. 22 is a diagram illustrating an example of a correction curve for the sky blue region determined by the hue correction unit 26. In FIG. 22, the horizontal axis represents the hue value (input hue value) before correction, and the vertical axis represents the correction amount (ΔH).
[0123]
In this example, correction is performed on pixels specified by hue values in the range of “280 to 340”, and the correction curve is created by combining two power functions. According to this correction curve, for example, a pixel having an input hue value of “290” is corrected by adding a hue value of about “5”.
[0124]
FIG. 23 is a diagram showing an example of a correction curve for the sky blue region determined by the hue correction unit 26. The horizontal axis represents the input hue value with respect to the correction curve shown in FIG. Represents an output hue value.
[0125]
Naturally, the correction curve created by the hue correction unit 26 is not limited to that shown in FIGS. 16 to 23, and various curves can be used as the correction curves as long as they are smooth curves that do not cause gradation skipping. Can be used.
[0126]
Returning to the description of FIG. 15, the hue correction unit 26 confirms the hue value of each pixel in step S <b> 93, and determines in step S <b> 94 whether or not the confirmed pixel is a memory color pixel.
[0127]
If it is determined in step S94 that the confirmed pixel is a memory color pixel, the hue correction unit 26 proceeds to step S95 and corrects the pixel based on the correction curve determined in step S93. Accordingly, when the correction curves as shown in FIGS. 16 and 17 are determined, the pixel of the memory color having the hue value “120” is corrected to the pixel having the hue value of “130”.
[0128]
On the other hand, if it is determined in step S94 that the confirmed pixel is not a memory color pixel, the process of step S95 is skipped.
[0129]
In step S96, the hue correction unit 26 determines whether or not all the pixels of the image have been confirmed. If it is determined that all the pixels have not been confirmed, the process returns to step S93, and the subsequent processing is repeatedly executed. On the other hand, if it is determined that all the pixels have been confirmed, the image data is output to the conversion unit 27. Then, the processing after step S8 in FIG. 4 is executed, and the image in which the luminance value, the saturation value, and the hue value are corrected is converted into the RGB color space by the conversion unit 27 and output.
[0130]
In the above, basically, the category of the image is determined based on the shooting mode of the image, and based on this, either global correction or local correction is selected and the correction amount is determined, but the image is observed. Since the method of recognizing the memory color varies depending on the user's age and sex, for example, when various user information such as the user's age and sex is set in the digital still camera 4 and included in the header information The correction method and the correction amount may be determined based on such various information.
[0131]
The series of processes described above can be executed by hardware, but can also be executed by software. In this case, the apparatus for executing the software is constituted by a personal computer as shown in FIG. 24, for example.
[0132]
In FIG. 24, a CPU (Central Processing Unit) 51 performs various processes according to a program stored in a ROM (Read Only Memory) 52 or a program loaded from a storage unit 58 to a RAM (Random Access Memory) 53. Execute. The RAM 53 also appropriately stores data necessary for the CPU 51 to execute various processes.
[0133]
The CPU 51, ROM 52, and RAM 53 are connected to each other via a bus 54. An input / output interface 55 is also connected to the bus 54.
[0134]
The input / output interface 55 includes an input unit 56 including a keyboard and a mouse, a display including a CRT and an LCD, an output unit 57 including a speaker, a storage unit 58 including a hard disk, a modem, a terminal adapter, and the like. A configured communication unit 59 is connected. The communication unit 59 performs communication processing via a network.
[0135]
A drive 50 is connected to the input / output interface 55 as required, and a magnetic disk 61, an optical disk 62, a magneto-optical disk 63, a semiconductor memory 64, or the like is appropriately mounted, and a computer program read from the disk is required. Is installed in the storage unit 58 accordingly.
[0136]
When a series of processing is executed by software, a program constituting the software may execute various functions by installing a computer incorporated in dedicated hardware or various programs. For example, it is installed from a network or a recording medium into a general-purpose personal computer or the like.
[0137]
As shown in FIG. 24, this recording medium is distributed to provide a program to the user separately from the main body of the apparatus, and includes a magnetic disk 61 (including a floppy disk) on which the program is recorded, an optical disk 62 ( Package media including CD-ROM (compact disk-read only memory), DVD (digital versatile disk), magneto-optical disk 63 (including MD (registered trademark) (mini-disk)), or semiconductor memory 64 And a ROM 52 that stores a program provided to the user in a state of being preinstalled in the apparatus main body, a hard disk included in the storage unit 58, and the like.
[0138]
In the present specification, the step of describing the program recorded in the recording medium is not limited to the processing performed in time series according to the described order, but is not necessarily performed in time series, either in parallel or individually. The process to be executed is also included.
[0139]
【The invention's effect】
As described above, according to the image processing apparatus and method, the recording medium, and the program of the present invention, it is possible to more suitably correct the color of an image without generating a pseudo contour or the like.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating an example of a device to which an image processing apparatus of the present invention is applied.
FIG. 2 is a block diagram illustrating a configuration example of an image processing apparatus to which the present invention is applied.
3 is a block diagram illustrating a configuration example of a luminance correction unit in FIG. 2;
4 is a flowchart illustrating processing of the image processing apparatus in FIG.
FIG. 5 is a flowchart for explaining processing of a category determination unit in FIG. 2;
FIG. 6 is a flowchart illustrating another process of the category determination unit in FIG. 2;
FIG. 7 is a flowchart illustrating still another process of the category determination unit in FIG. 2;
FIG. 8 is a flowchart for explaining processing of a brightness correction unit in FIG. 2;
9 is a diagram illustrating an example of a correction curve created by the process of FIG.
10 is a diagram showing another example of a correction curve created by the process of FIG.
11 is a flowchart illustrating processing of a saturation correction unit in FIG.
12 is a diagram showing an example of a correction curve created by the process of FIG.
13 is a diagram showing another example of a correction curve created by the process of FIG.
14 is a diagram showing still another example of a correction curve created by the process of FIG.
15 is a flowchart for explaining processing of a hue correction unit in FIG. 2;
16 is a diagram illustrating an example of a correction curve created by the process of FIG.
FIG. 17 is a diagram illustrating another example of a correction curve created by the process of FIG.
18 is a diagram showing still another example of the correction curve created by the process of FIG.
FIG. 19 is a diagram illustrating an example of a correction curve created by the process of FIG.
20 is a diagram illustrating another example of the correction curve created by the process of FIG.
FIG. 21 is a diagram showing still another example of the correction curve created by the process of FIG.
22 is a diagram showing an example of a correction curve created by the process of FIG.
FIG. 23 is a diagram illustrating another example of the correction curve created by the process of FIG.
FIG. 24 is a block diagram illustrating a configuration example of a personal computer.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 11 Image processing apparatus, 21 Image analysis part, 22 Category discrimination | determination part, 23 Conversion part, 24 Luminance correction part, 25 Saturation correction part, 26 Hue correction part, 27 Conversion part, 28 Output part, 41 Correction switching part, 42 Correction Quantity determining unit, 43 Memory color area extracting unit, 44 Correction processing unit, 61 Magnetic disk, 62 Optical disk, 63 Magneto-optical disk, 64 Semiconductor memory

Claims (12)

Discrimination means for discriminating a category for the entire image data based on the size of a plurality of predetermined memory color areas in the entire image data ;
According to the category for the entire image data determined by the determining means, either correction of only a memory color area that approximates a predetermined memory color or correction of the entire image data is selected. A selection means;
Correction means for correcting the image data in accordance with the selection by the selection means;
An image processing apparatus comprising: The determination means determines the category of the image data as a portrait image when the size of the skin color area as the memory color area is larger than a first threshold and the sizes of the green area and the sky blue area are smaller than a second threshold. Discriminate
  The image processing apparatus according to claim 1. When the size of the skin color area as the memory color area is larger than the first threshold value and the size of the green color area and the sky blue area is larger than the second threshold value, the determination means determines the category of the image data Discriminate from landscape image
  The image processing apparatus according to claim 1. The discriminating unit is configured such that the size of the skin color area as the memory color area is smaller than a first threshold value and larger than a second threshold value that is smaller than the first threshold value, and the sizes of the green color area and the sky blue color area are larger. When the value is smaller than the third threshold, the category of the image data is determined as a small person image
  The image processing apparatus according to claim 1. The discriminating unit is configured such that the size of the skin color area as the memory color area is smaller than a first threshold value and larger than a second threshold value that is smaller than the first threshold value, and the sizes of the green color area and the sky blue color area are larger. When the value is larger than a third threshold, the category of the image data is determined as a small person & landscape image
  The image processing apparatus according to claim 1. The determination means sets the category of the image data as a landscape image when the size of the skin color area as the memory color area is smaller than a first threshold and the sizes of the green area and the sky blue area are larger than a second threshold. Discriminate
  The image processing apparatus according to claim 1. It said selection means, as the correction of the luminance of the image data, correction of the memory color area only, or select one of the overall correction of the image data,
The correction unit corrects the luminance of the image data according to the selection by the selection unit.
The image processing apparatus according to claim 1 . Further comprising extraction means for extracting the memory color region;
The correction unit corrects the saturation and hue of the memory color area extracted by the extraction unit.
The image processing apparatus according to claim 1 . The discriminating unit further discriminates a category for the entire image data based on information described in a header of the image data.
The image processing apparatus according to claim 1 . A determining step of determining a category for the entire image data based on the size of a plurality of predetermined memory color areas in the entire image data ;
According to the category for the entire image data determined by the processing of the determination step, either correction of only a memory color area that approximates a predetermined memory color or correction of the entire image data is performed. A selection step to select;
A correction step for correcting the image data in accordance with the selection made by the selection step;
An image processing method including : A determining step of determining a category for the entire image data based on the size of a plurality of predetermined memory color areas in the entire image data ;
According to the category for the entire image data determined by the processing of the determination step, either correction of only a memory color area that approximates a predetermined memory color or correction of the entire image data is performed. A selection step to select;
A correction step for correcting the image data in accordance with the selection made by the selection step;
A computer-readable recording medium on which a computer-readable program is recorded. A determining step of determining a category for the entire image data based on the size of a plurality of predetermined memory color areas in the entire image data ;
According to the category for the entire image data determined by the processing of the determination step, either correction of only a memory color area that approximates a predetermined memory color or correction of the entire image data is performed. A selection step to select;
A program that causes a computer to execute a process including a correction step of correcting the image data in accordance with selection by the process of the selection step.

Images