JPH0798762A - Image processing method - Google Patents

Image processing method

Info

Publication number
JPH0798762A
JPH0798762A JP6101399A JP10139994A JPH0798762A JP H0798762 A JPH0798762 A JP H0798762A JP 6101399 A JP6101399 A JP 6101399A JP 10139994 A JP10139994 A JP 10139994A JP H0798762 A JPH0798762 A JP H0798762A
Authority
JP
Japan
Prior art keywords
color
extraction
distribution
range
ranges
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP6101399A
Other languages
Japanese (ja)
Inventor
Tatsuo Sumida
達夫 澄田
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Citizen Watch Co Ltd
Original Assignee
Citizen Watch Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Citizen Watch Co Ltd filed Critical Citizen Watch Co Ltd
Priority to JP6101399A priority Critical patent/JPH0798762A/en
Publication of JPH0798762A publication Critical patent/JPH0798762A/en
Pending legal-status Critical Current

Links

Landscapes

  • Color Image Communication Systems (AREA)
  • Image Analysis (AREA)

Abstract

PURPOSE:To provide the image processing method by which a processing object can be extracted as the same processing object, even if uneven brightness or uneven color exists. CONSTITUTION:From a reference image 101 containing a processing object, a color distribution/extraction distribution 106 whose extraction is desired and a color distribution/obstraction distribution 107 whose extraction is not desired are obtained. A contour of the extraction distribution 106 is approximated by combining enough many color extraction ranges. From this contour, plural adjacent color extraction ranges are selected, synthesized to one color extraction range and the number of color extraction ranges is decreased, but if the synthesized color extraction range collides with the obstruction distribution 107, two of them are not synthesized. The combination of the color extraction ranges whose number is decreased to the number used for the extraction processing is a color distribution extraction range for approximating the extraction distribution 106. In a processing image 112, when an area being a color contained in one of the color distribution extraction ranges is extracted, the processing object having unevenness can be extracted.

Description

【発明の詳細な説明】Detailed Description of the Invention

【０００１】[0001]

【産業上の利用分野】本発明は、カラー画像から所望の
色である領域だけを抽出する事で処理対象の画像を処理
画像より抽出し、位置計測や特徴値の計算などを行う画
像処理方法に関わり、特に、所望の色である領域が明る
さムラや色ムラによって広い色分布を持つ場合にも所望
の色である領域だけを抽出する画像処理方法に関する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method for extracting an image to be processed from a processed image by extracting only a region having a desired color from a color image, and performing position measurement and calculation of characteristic values. In particular, the present invention relates to an image processing method for extracting only a region having a desired color even when the region having a desired color has a wide color distribution due to uneven brightness or uneven color.

【０００２】[0002]

【従来の技術】画像処理装置で色を表す場合、画像取り
込み装置、画像記憶装置、画像表示装置などの画像処理
機器類の内部構成上取り扱いやすいことから、多くの場
合光の３原色であるＲ（赤）Ｇ（緑）Ｂ（青）の３要素
が用いられる。また、ＲＧＢの３要素で表現されるデー
タは人間の感性と対応しにくく取り扱いにくいというこ
とから、色をＨ（色合い）Ｓ（色の濃さ）Ｖ（明るさ）
の３要素で表現するようにした画像処理装置も用いられ
ている。2. Description of the Related Art When a color is represented by an image processing device, it is often the three primary colors of light because it is easy to handle because of the internal configuration of image processing devices such as an image capturing device, an image storage device, and an image display device. Three elements of (red) G (green) B (blue) are used. In addition, since the data expressed by the three elements of RGB is difficult to handle and difficult to handle with human sensibilities, the color is H (hue) S (color intensity) V (brightness).
There is also used an image processing device that is represented by three elements.

【０００３】色は色を表す３つの要素を軸とした座標空
間、即ち色空間内の座標値で表現される。ＲＧＢ色空間
の場合はＲ、Ｇ、Ｂのそれぞれが座標軸である直交座標
空間で表現され、ＨＳＶ色空間の場合にはＶを中心軸方
向、Ｓを半径方向、Ｈを角度方向に取った円筒座標形で
表現される。A color is represented by a coordinate space having three color elements as axes, that is, coordinate values in the color space. In the case of the RGB color space, each of R, G, and B is represented by a rectangular coordinate space that is a coordinate axis, and in the case of the HSV color space, V is a central axis direction, S is a radial direction, and H is a cylinder in an angular direction. Expressed in coordinate form.

【０００４】所望の色である領域を抽出する事で処理対
象の画像を処理画像より抽出するとき、カラー画像デー
タのうち３つの要素すべてが所望の色の３要素と一致す
る色である部分を処理対象として抽出すればよいが、感
覚的には同じ色といえる部分であっても実際のカラー画
像データでは要素の値として多少の変化があるので、３
要素の値すべてが等しい領域だけを処理対象として抽出
するのではなく、所望の色を表現する各要素に要素の値
の変化を内包できる適当な幅を持たせた色抽出範囲を考
えて、３要素の値がすべてこの範囲内に入る領域を処理
対象として抽出している。When an image to be processed is extracted from a processed image by extracting a region having a desired color, a portion in which all three elements of the color image data are colors corresponding to three elements of the desired color are extracted. It may be extracted as a processing target, but even if it can be said that the same color is sensuously sensed, there is a slight change in the element value in the actual color image data.
Considering a color extraction range in which each element expressing a desired color has an appropriate width that can include a change in the element value, instead of extracting only an area in which all element values are equal, A region in which all element values fall within this range is extracted for processing.

【０００５】[0005]

【発明が解決しようとする課題】しかしながら、上述し
た従来の方法では、処理対象の色が明るさムラや色ムラ
で色空間内で広い範囲に分布している場合に、この色空
間内に広がった１つの色であるとみなせる分布、即ち色
分布をすべて内包する色抽出範囲で抽出を行うので不要
な部分まで処理対象として抽出する場合がある。However, according to the above-described conventional method, when the color to be processed is distributed in a wide range in the color space due to uneven brightness or uneven color, it spreads in this color space. Since the extraction is performed in a color extraction range that includes all the distributions that can be regarded as one color, that is, the color distribution, unnecessary portions may be extracted as processing targets.

【０００６】例えば、ＲＧＢ要素で色を表現する画像処
理装置で灰色が明るさムラのみで黒〜灰色〜白の範囲に
広がった色分布を１つの色抽出範囲で抽出しようとすれ
ば、ＲＧＢの各要素の範囲は灰色の最小値から最大値を
設定することになる。このためＲＧＢの軸で表される色
空間のほとんどの範囲が色抽出範囲となるため、所望の
色分布である黒〜灰色〜白の領域だけを抽出することが
できない。For example, in an image processing apparatus that expresses colors with RGB elements, if an attempt is made to extract a color distribution in which gray is spread in the range of black to gray to white with only brightness unevenness in one color extraction range, The range of each element will be set from the minimum value of gray to the maximum value. For this reason, most of the range of the color space represented by the axes of RGB is the color extraction range, and it is not possible to extract only the black, gray, and white regions having the desired color distribution.

【０００７】一方、ＨＳＶ要素で色を表現する画像処理
装置ならば、灰色という無彩色の抽出色に対してＶ要
素、即ち明るさの範囲のみを広げればこの色分布の広が
りに対応できる。このように、色分布が色空間の要素の
座標軸に沿った方向にのみ広がっている場合には、その
要素の範囲だけを広げればよいので不要な部分を取り込
んでしまうという問題は表面化しない。On the other hand, an image processing apparatus that expresses colors with HSV elements can cope with the spread of this color distribution by expanding only the V element, that is, the range of brightness, with respect to the achromatic extracted color of gray. In this way, when the color distribution spreads only in the direction along the coordinate axes of the elements of the color space, only the range of the elements needs to be widened, so that the problem of taking in unnecessary parts does not surface.

【０００８】しかし、一般的には明るさが変化すると色
合いや色の濃さも変化して明るさの軸に沿わない色分布
を生ずるし、色ムラは色の濃さと色合いの一方の要素の
みで色分布を広げるものではないので、１つの色抽出範
囲だけで抽出を行うと不要な部分まで処理対象として抽
出してしまう問題は、ＨＳＶ要素で色を表現する画像処
理装置であっても完全に回避する事はできない。However, in general, when the brightness changes, the hue and the density of the color also change, and a color distribution that does not follow the axis of the brightness occurs, and the color unevenness is caused by only one element of the color density and the hue. Since it does not broaden the color distribution, the problem of extracting unnecessary parts as processing targets if extraction is performed with only one color extraction range is completely possible even with an image processing device that expresses colors with HSV elements. There is no avoidance.

【０００９】以上のような問題点があって、従来では、
適当な色抽出範囲を設定や照明条件を整える事、抽出後
の不完全な処理対象の形から画像計測を行う事に熟練を
要するという画像処理の扱いにくさを生じていた。Due to the above-mentioned problems, in the past,
This makes it difficult to handle image processing because it requires skill to set an appropriate color extraction range and adjust lighting conditions, and to perform image measurement from the shape of an incomplete processing target after extraction.

【００１０】本発明の目的は、処理対象の色分布が広い
ものであったり、処理対象とそれ以外の部分の色分布が
近接しているような場合にも不要な部分を抽出せずに所
望の部分だけを抽出できるような画像処理方法を提供す
る事で、色抽出範囲や照明条件を熟練を要せずに設定で
き、抽出後の完全な処理領域の形から画像計測を行える
ようにして画像処理を扱い易くする事にある。The object of the present invention is to obtain a desired object without extracting unnecessary parts even when the object to be processed has a wide color distribution or the object to be processed and other parts have a close color distribution. By providing an image processing method that can extract only the part, it is possible to set the color extraction range and lighting conditions without skill, and to perform image measurement from the shape of the complete processing area after extraction. It is to make image processing easier to handle.

【００１１】[0011]

【課題を解決するための手段】上記目的を達成するため
に、本発明による画像処理方法は、画像処理の対象であ
る処理対象の有する色分布、あるいは処理対象の画像を
含んでいるカラー画像を得て参照画像とし、参照画像よ
り処理対象の有する色分布を含む、所望の領域である、
抽出領域内のカラー画像データから色情報を読み取る事
によって得られた色分布を抽出分布として、抽出分布の
輪郭を色抽出範囲で近似する処理をし、前記色抽出範囲
によって処理対象の画像を処理画像より抽出する画像処
理方法において、抽出分布の輪郭を複数の色抽出範囲で
近似した色分布抽出範囲を得る処理をして、前記色分布
抽出範囲により処理対象の画像を処理画像より抽出す
る。In order to achieve the above object, an image processing method according to the present invention provides a color distribution of a processing target which is an object of image processing, or a color image including an image of the processing target. It is a desired area that includes the color distribution of the processing target from the reference image obtained,
The color distribution obtained by reading the color information from the color image data in the extraction area is used as the extraction distribution, and the contour of the extraction distribution is approximated by the color extraction range, and the image to be processed is processed by the color extraction range. In the image processing method of extracting from an image, a process of obtaining a color distribution extraction range in which the contour of the extraction distribution is approximated by a plurality of color extraction ranges is performed, and an image to be processed is extracted from the processed image by the color distribution extraction range.

【００１２】さらに、本発明による画像処理方法は、色
分布抽出範囲を構成する色抽出範囲を被分割色抽出範囲
として、前記被分割色抽出範囲を、任意の分割位置で複
数の色抽出範囲に分割して分割範囲として、分割範囲で
抽出される、抽出分布のなかの、色分布に外接する色抽
出範囲を外接範囲として、外接範囲の大きさの総和を求
めて、さらに複数の、他の任意の分割位置に関しても、
外接範囲の大きさの総和を求めて、前記、複数の外接範
囲の大きさの総和のなかで、最小の大きさの総和である
外接範囲を被分割色抽出範囲と置き換えることにより、
抽出分布の輪郭を複数の色抽出範囲で近似する。Further, in the image processing method according to the present invention, the color extraction range forming the color distribution extraction range is set as a divided color extraction range, and the divided color extraction range is set to a plurality of color extraction ranges at arbitrary division positions. Divide the divided range, extract the divided range, and use the color extraction range circumscribing the color distribution in the extracted distribution as the circumscribing range to find the sum of the sizes of the circumscribing ranges. For any division position,
Obtaining the sum of the sizes of the circumscribing ranges, among the sum of the sizes of the plurality of circumscribing ranges, by replacing the circumscribing range, which is the sum of the smallest sizes, with the divided color extraction range,
The contour of the extraction distribution is approximated by a plurality of color extraction ranges.

【００１３】また、本発明による画像処理方法は、参照
画像より処理対象の有する色分布を含まない、抽出して
はならない部分の所望の領域である、阻止領域内のカラ
ー画像データから色情報を読み取る事によって得られた
色分布を阻止分布として、抽出分布及び阻止分布の属す
る色情報の表現範囲である色空間から阻止分布を除いた
残りの前記色空間内の領域で、処理対象の画像を処理画
像より抽出する。Further, according to the image processing method of the present invention, the color information is obtained from the color image data in the inhibition area, which is a desired area of the portion which should not be extracted and which does not include the color distribution of the processing object from the reference image. With the color distribution obtained by reading as the stop distribution, the image to be processed is displayed in the remaining area in the color space excluding the stop distribution from the color space that is the expression range of the color information to which the extraction distribution and the stop distribution belong. Extract from the processed image.

【００１４】さらにまた、本発明による画像処理方法
は、参照画像より処理対象の有する色分布を含まない、
抽出してはならない部分の所望の領域である、阻止領域
内のカラー画像データから色情報を読み取る事によって
得られた色分布を阻止分布として、抽出分布の輪郭を十
分に多数の色抽出範囲の組み合わせで近似した色分布抽
出範囲にあっては、色分布抽出範囲を構成する色抽出範
囲の数を減らすために、隣接する複数の色抽出範囲を被
合成色抽出範囲として選択し、前記被合成色抽出範囲を
内包する１つの色抽出範囲である合成色抽出範囲を得
て、前記合成色抽出範囲が、阻止分布を含まないとき
に、被合成色抽出範囲を合成色抽出範囲と置き換えて、
抽出分布の輪郭を複数の色抽出範囲で近似する。Furthermore, the image processing method according to the present invention does not include the color distribution of the processing object as compared with the reference image,
The color distribution obtained by reading the color information from the color image data in the inhibition area, which is the desired area of the portion that should not be extracted, is used as the inhibition distribution, and the contour of the extraction distribution is set to a sufficiently large number of color extraction ranges. In the color distribution extraction range approximated by the combination, in order to reduce the number of color extraction ranges constituting the color distribution extraction range, a plurality of adjacent color extraction ranges are selected as the combined color extraction range, and the combined target color extraction range is selected. A composite color extraction range that is one color extraction range including a color extraction range is obtained, and when the composite color extraction range does not include a rejection distribution, the composite color extraction range is replaced with the composite color extraction range,
The contour of the extraction distribution is approximated by a plurality of color extraction ranges.

【００１５】[0015]

【作用】本発明による画像処理によれば、抽出分布の輪
郭を複数の色抽出範囲で近似した色分布抽出範囲を得る
処理をして、前記色分布抽出範囲により処理対象の画像
を処理画像より抽出する。したがって、処理対象の持つ
色分布である抽出分布の輪郭を複数の色抽出範囲の組み
合わせで近似した色分布抽出範囲を得て、処理画像のう
ち、色分布抽出範囲を構成する複数の色抽出範囲のどれ
かに含まれる色である部分を抽出する事で処理対象を抽
出するので、明るさムラや色ムラなどで抽出分布が広い
範囲に広がっていても抽出分布に含まれる色である領域
を抽出する。According to the image processing of the present invention, processing for obtaining a color distribution extraction range in which the contour of the extraction distribution is approximated by a plurality of color extraction ranges is performed, and the image to be processed is processed from the processed image by the color distribution extraction range. Extract. Therefore, the contour of the extraction distribution, which is the color distribution of the processing target, is approximated by a combination of a plurality of color extraction ranges to obtain a color distribution extraction range, and the plurality of color extraction ranges forming the color distribution extraction range in the processed image are obtained. Since the processing target is extracted by extracting the part that is a color included in any of the above, even if the extraction distribution is spread over a wide range due to uneven brightness or color unevenness, the area that is the color included in the extraction distribution is Extract.

【００１６】さらに、本発明による画像処理方法は、色
分布抽出範囲を構成する色抽出範囲を被分割色抽出範囲
として、前記被分割色抽出範囲を、任意の分割位置で複
数の色抽出範囲に分割して分割範囲として、分割範囲で
抽出される、抽出分布のなかの、色分布に外接する色抽
出範囲を外接範囲として、外接範囲の大きさの総和を求
めて、さらに複数の、他の任意の分割位置に関しても、
外接範囲の大きさの総和を求めて、前記、複数の外接範
囲の大きさの総和のなかで、最小の大きさの総和である
外接範囲を被分割色抽出範囲と置き換えることにより、
抽出分布の輪郭を複数の色抽出範囲で近似する。したが
って、限られた数の色抽出範囲で、精度良く抽出分布の
輪郭を近似する。Further, in the image processing method according to the present invention, the color extraction range forming the color distribution extraction range is set as a divided color extraction range, and the divided color extraction range is set to a plurality of color extraction ranges at arbitrary division positions. Divide the divided range, extract the divided range, and use the color extraction range circumscribing the color distribution in the extracted distribution as the circumscribing range to find the sum of the sizes of the circumscribing ranges. For any division position,
Obtaining the sum of the sizes of the circumscribing ranges, among the sum of the sizes of the plurality of circumscribing ranges, by replacing the circumscribing range, which is the sum of the smallest sizes, with the divided color extraction range,
The contour of the extraction distribution is approximated by a plurality of color extraction ranges. Therefore, the contour of the extraction distribution is accurately approximated with a limited number of color extraction ranges.

【００１７】また、本発明による画像処理方法は、参照
画像より処理対象の有する色分布を含まない、抽出して
はならない部分の所望の領域である、阻止領域内のカラ
ー画像データから色情報を読み取る事によって得られた
色分布を阻止分布として、抽出分布及び阻止分布の属す
る色情報の表現範囲である色空間から阻止分布を除いた
残りの前記色空間内の領域で、処理対象の画像を処理画
像より抽出する。したがって、処理対象の有する色分布
を含まず抽出してはならない部分の有する色分布である
阻止分布を除いた残りの色空間の範囲内で、処理対象の
持つ色分布である抽出分布の輪郭を複数の色抽出範囲の
組み合わせで近似した色分布抽出範囲を得て、処理画像
のうち、色分布抽出範囲を構成する複数の色抽出範囲の
どれかに含まれる色である部分を抽出する事で処理対象
を抽出する。Further, according to the image processing method of the present invention, the color information is obtained from the color image data in the inhibition area, which is a desired area of a portion which should not be extracted and which does not include the color distribution of the processing object from the reference image. With the color distribution obtained by reading as the stop distribution, the image to be processed is displayed in the remaining area in the color space excluding the stop distribution from the color space that is the expression range of the color information to which the extraction distribution and the stop distribution belong. Extract from the processed image. Therefore, within the range of the remaining color space excluding the stop distribution, which is the color distribution of the portion that does not include the color distribution of the processing target and that should not be extracted, the contour of the extraction distribution that is the color distribution of the processing target is By obtaining a color distribution extraction range approximated by a combination of a plurality of color extraction ranges, and extracting a portion of the processed image that is a color included in any of the plurality of color extraction ranges that form the color distribution extraction range. Extract the processing target.

【００１８】さらにまた、本発明による画像処理方法
は、参照画像より処理対象の有する色分布を含まない、
抽出してはならない部分の所望の領域である、阻止領域
内のカラー画像データから色情報を読み取る事によって
得られた色分布を阻止分布として、抽出分布の輪郭を十
分に多数の色抽出範囲の組み合わせで近似した色分布抽
出範囲にあっては、色分布抽出範囲を構成する色抽出範
囲の数を減らすために、隣接する複数の色抽出範囲を被
合成色抽出範囲として選択し、前記被合成色抽出範囲を
内包する１つの色抽出範囲である合成色抽出範囲を得
て、前記合成色抽出範囲が、阻止分布を含まないとき
に、被合成色抽出範囲を合成色抽出範囲と置き換えて、
抽出分布の輪郭を複数の色抽出範囲で近似する。したが
って、色分布抽出範囲を構成する色抽出範囲から近接す
るものを選択し、選択した複数の色抽出範囲を全て内包
する合成色抽出範囲を得て、合成色抽出範囲が阻止分布
に掛からなければ選択した複数の色抽出範囲と合成色抽
出範囲とを置き換えて色抽出範囲の総数を減らして行く
ので、阻止分布と抽出分布が接近している部分は細かく
近似され、阻止分布と抽出分布が離れている部分は大ま
かに近似された色分布抽出範囲が得られる。Furthermore, the image processing method according to the present invention does not include the color distribution of the processing object as compared with the reference image,
The color distribution obtained by reading the color information from the color image data in the inhibition area, which is the desired area of the portion that should not be extracted, is used as the inhibition distribution, and the contour of the extraction distribution is set to a sufficiently large number of color extraction ranges. In the color distribution extraction range approximated by the combination, in order to reduce the number of color extraction ranges constituting the color distribution extraction range, a plurality of adjacent color extraction ranges are selected as the combined color extraction range, and the combined target color extraction range is selected. A composite color extraction range that is one color extraction range including a color extraction range is obtained, and when the composite color extraction range does not include a rejection distribution, the composite color extraction range is replaced with the composite color extraction range,
The contour of the extraction distribution is approximated by a plurality of color extraction ranges. Therefore, a color extraction range that is close to the color extraction range is selected, a composite color extraction range that includes all of the selected color extraction ranges is obtained, and if the composite color extraction range does not fall on the inhibition distribution. Since the total number of color extraction ranges is reduced by replacing the selected multiple color extraction ranges with the composite color extraction range, the part where the blocking distribution and the extraction distribution are close to each other is finely approximated, and the blocking distribution and the extraction distribution are separated from each other. In the shaded area, a color distribution extraction range roughly approximated is obtained.

【００１９】[0019]

【実施例】【Example】

＜実施例１＞以下、本発明の実施例１を図面を基に説明
する。図２は、本発明による画像処理方法の実施例１に
おける処理の構成を示す流れ図である。処理対象１０２
の有する色分布、あるいは処理対象の画像を含んでいる
カラー画像を得て参照画像１０１とし、参照画像１０１
の処理対象の有する色分布を含む所望の領域である抽出
領域１０３から色分布取得処理１０５によって得た色分
布を抽出分布１０６とする。色分布抽出範囲取得処理１
０８によって抽出分布１０６の輪郭を複数の色抽出範囲
で近似した色分布抽出範囲１１１を得て、色分布抽出処
理１１３では、処理画像１１２のうち、色分布抽出範囲
１１１を構成する複数の色抽出範囲のどれかに含まれる
色である部分を抽出する事で、処理対象である部分のみ
を抽出した画像データである抽出画像１１４を得る。<First Embodiment> A first embodiment of the present invention will be described below with reference to the drawings. FIG. 2 is a flow chart showing the configuration of processing in the first embodiment of the image processing method according to the present invention. Processing target 102
A color image including the color distribution of the image or the image to be processed is obtained and used as the reference image 101.
The color distribution obtained by the color distribution acquisition processing 105 from the extraction area 103, which is a desired area including the color distribution of the processing target, is set as the extraction distribution 106. Color distribution extraction range acquisition processing 1
08, a color distribution extraction range 111 in which the contour of the extraction distribution 106 is approximated by a plurality of color extraction ranges is obtained, and in the color distribution extraction processing 113, a plurality of color extractions forming the color distribution extraction range 111 are extracted from the processed image 112. By extracting a portion that is a color included in any of the ranges, an extracted image 114 that is image data obtained by extracting only the portion that is the processing target is obtained.

【００２０】図３は、本発明による画像処理方法の実施
例における色分布取得処理１０５の処理内容を説明する
ための図である。参照画像１０１の処理対象の有する色
分布を含む所望の領域である抽出領域１０３から平均色
ではなく色分布を得るために、抽出領域を多数の微小領
域に分割し、個々の微小領域毎に参照画像１０１から色
情報を読み取って記憶する事で抽出分布１０６を得る事
ができる。微小領域の最小単位はカラー画像データの画
素の単位である。FIG. 3 is a diagram for explaining the processing contents of the color distribution acquisition processing 105 in the embodiment of the image processing method according to the present invention. In order to obtain a color distribution, not the average color, from the extraction region 103, which is a desired region including the color distribution of the processing target of the reference image 101, the extraction region is divided into a large number of micro regions, and each micro region is referenced. The extraction distribution 106 can be obtained by reading and storing color information from the image 101. The minimum unit of the minute area is a pixel unit of color image data.

【００２１】図３では長方形である抽出領域１０３を
（Ｍ×Ｎ）分割したｚ個の微小領域を示している。抽出
分布１０６はｚ個の色情報を記憶できるテーブルに記憶
された色情報の集合であり、本実施例では色をＲＧＢの
３つの要素で表現するものとする。色を他の表現方法で
表す画像処理方法の場合は、単に要素の名前が変わるだ
けで、例えばＨＳＶの３要素で色を表す場合は（Ｒ１，
Ｇ１，Ｂ１）は（Ｈ１，Ｓ１，Ｖ１）とすればよい。１
個目の微小領域について参照画像１０１から読み取られ
た色情報（ＲＧＢ）はそれぞれ抽出分布１０６の１個目
の色情報である（Ｒ１，Ｇ１，Ｂ１）に記憶される。同
様に、ｘ個目の微小領域について参照画像１０１から読
み取られた色情報は抽出分布１０６のｘ個目の色情報に
記憶され、ｘについて１個目からｚ個目までの微小領域
について参照画像１０１から色情報を読み取って対応す
る色情報に記憶する事で抽出分布１０６が得られる。図
３では、抽出領域１０３が１つの領域からなる場合を示
しているが、抽出領域１０３が複数の領域からなる場合
であれば、抽出領域１０３を構成する個々の領域につい
て、それぞれの領域を微小領域に分割して個々の微小領
域毎に色情報を読み取れば良い。この時、抽出領域１０
３を構成する全領域の微小領域の数の総和が抽出分布１
０６の要素数となる。FIG. 3 shows z minute regions obtained by dividing the rectangular extraction region 103 into (M × N). The extraction distribution 106 is a set of color information stored in a table capable of storing z pieces of color information, and in this embodiment, colors are represented by three elements of RGB. In the case of an image processing method in which colors are represented by other representation methods, if the names of the elements are simply changed, and if the colors are represented by three HSV elements (R1,
G1, B1) may be (H1, S1, V1). 1
The color information (RGB) read from the reference image 101 for the first minute area is stored in the first color information (R1, G1, B1) of the extraction distribution 106, respectively. Similarly, the color information read from the reference image 101 for the xth micro area is stored in the xth color information of the extraction distribution 106, and the reference image for the first to zth micro areas for x is stored. The extraction distribution 106 is obtained by reading the color information from 101 and storing it in the corresponding color information. Although FIG. 3 shows the case where the extraction region 103 is composed of one region, if the extraction region 103 is composed of a plurality of regions, each region that constitutes the extraction region 103 is divided into small regions. The color information may be read for each minute area by dividing into areas. At this time, the extraction area 10
Extraction distribution 1 is the total sum of the numbers of minute areas of all areas that make up 3
The number of elements is 06.

【００２２】図４（ａ）は本発明による画像処理方法の
実施例における色分布抽出範囲１１１の構成を説明する
ための図である。色分布抽出範囲１１１は、処理画像１
１２から色分布抽出処理１１３によって処理対象を抽出
するとき用いる色抽出範囲の数ｋ個の色抽出範囲１１１
−１〜１１１−ｋを記憶できるテーブルに記憶された色
抽出範囲の集合である。色抽出範囲はその色抽出範囲で
抽出したい色の範囲を示すために、色を表現する３つの
要素毎に最大値と最小値を要素として持つ６つの値の集
合である。本実施例では色をＲＧＢの３要素で表現する
ので、例えば、色分布抽出範囲１１１を構成するｊ番目
の色抽出範囲１１１−ｊは、（Ｒｊｕ，Ｒｊｄ，Ｇｊ
ｕ，Ｇｊｄ，Ｂｊｕ，Ｂｊｄ）で構成される。Ｒｊｕが
Ｒ要素の最大値、ＲｊｄがＲ要素の最小値を示し、他の
要素についても同様である。他の表現方法で色を表現し
た場合は単に要素の名前を変えるだけであり、例えばＨ
ＳＶの３要素で色を表現する画像処理方法の場合、（Ｈ
ｊｕ，Ｈｊｄ，Ｓｊｕ，Ｓｊｄ，Ｖｊｕ，Ｖｊｄ）であ
る。FIG. 4A is a diagram for explaining the configuration of the color distribution extraction range 111 in the embodiment of the image processing method according to the present invention. The color distribution extraction range 111 is the processed image 1
The number k of color extraction ranges used when extracting the processing target from 12 by the color distribution extraction processing 113
It is a set of color extraction ranges stored in a table that can store −1 to 111-k. The color extraction range is a set of six values each having a maximum value and a minimum value as an element for each of the three elements expressing the color in order to indicate the range of the color to be extracted in the color extraction range. In the present embodiment, the color is represented by the three elements of RGB, so for example, the j-th color extraction range 111-j forming the color distribution extraction range 111 is (Rju, Rjd, Gj
u, Gjd, Bju, Bjd). Rju indicates the maximum value of the R element, Rjd indicates the minimum value of the R element, and the same applies to other elements. When colors are expressed by other expression methods, the name of the element is simply changed. For example, H
In the case of the image processing method that expresses colors with the three elements of SV, (H
ju, Hjd, Sju, Sjd, Vju, Vjd).

【００２３】図４（ｂ）はＲＧＢの３要素で表現される
色空間４０１において、色抽出範囲１１１−ｊの抽出範
囲の１例を示す斜視図である。色をＲＧＢの３要素で表
現する場合は、ＲＧＢのそれぞれの要素を３つの直交軸
に取った３次元直交座標空間で表現するのが一般的であ
る。色抽出範囲１１１−ｊの各要素は、ＲＧＢそれぞれ
の軸での最大値と最小値を示すから、それぞれの最大値
と最小値に挟まれた直方体の範囲が色空間４０１内にお
ける色抽出範囲１１１−ｊの抽出範囲である。FIG. 4B is a perspective view showing an example of the extraction range of the color extraction range 111-j in the color space 401 represented by the three elements of RGB. When a color is represented by three elements of RGB, it is general that each element of RGB is represented by a three-dimensional orthogonal coordinate space in which three orthogonal axes are taken. Since each element of the color extraction range 111-j indicates the maximum value and the minimum value on each axis of RGB, the range of the rectangular parallelepiped sandwiched between the maximum value and the minimum value is the color extraction range 111 in the color space 401. It is the extraction range of -j.

【００２４】図５及び図６は、本発明による画像処理方
法の実施例における色分布抽出範囲取得処理１０８の処
理内容を説明するための流れ図である。5 and 6 are flow charts for explaining the processing contents of the color distribution extraction range acquisition processing 108 in the embodiment of the image processing method according to the present invention.

【００２５】（被分割色抽出範囲の初期化：５０１）被
分割色分布抽出範囲５０２は色分布抽出範囲である。ま
ず、（被分割色抽出範囲の初期化：５０１）の処理によ
って被分割色分布抽出範囲５０２を抽出分布１０６の輪
郭全体に外接する色抽出範囲１つからなる色分布抽出範
囲に初期化する。抽出分布１０６の輪郭全体に外接する
色抽出範囲を求めるには、抽出分布１０６を構成する色
情報の全てを調べて、ＲＧＢ各要素の最大値と最小値を
求め、それをそのまま色抽出範囲の要素とする。(Initialization of Divided Color Extraction Range: 501) The divided color distribution extraction range 502 is a color distribution extraction range. First, the divided color distribution extraction range 502 is initialized to a color distribution extraction range including one color extraction range circumscribing the entire contour of the extraction distribution 106 by the process of (initialization of divided color extraction range: 501). In order to obtain the color extraction range circumscribing the entire contour of the extraction distribution 106, all the color information that constitutes the extraction distribution 106 is examined, the maximum value and the minimum value of each RGB element are obtained, and this is directly used as the color extraction range. As an element.

【００２６】（次の色抽出範囲を選択：５０３）被分割
色分布抽出範囲５０２は複数の色抽出範囲の組み合わせ
で抽出分布１０６の輪郭を近似した色分布抽出範囲であ
る。ただし、処理の開始時には（被分割色抽出範囲の初
期化：５０１）によって、１個の色抽出範囲で抽出分布
１０６を近似した状態になっている。（次の色抽出範囲
を選択：５０３）では、被分割色分布抽出範囲５０２を
構成する色抽出範囲の中から１つづつ色分布抽出範囲を
選択して取り出し、被分割色抽出範囲５０４として、後
続する処理に渡す。(Select next color extraction range: 503) The divided color distribution extraction range 502 is a color distribution extraction range in which the outline of the extraction distribution 106 is approximated by a combination of a plurality of color extraction ranges. However, at the start of the process (initialization of the divided color extraction range: 501), the extraction distribution 106 is approximated by one color extraction range. In (Select next color extraction range: 503), the color distribution extraction ranges are selected and extracted one by one from the color extraction ranges forming the divided color distribution extraction range 502. Pass to subsequent processing.

【００２７】（次の分割位置に変更：５０５）後続する
処理で被分割色抽出範囲５０４を分割する位置を定める
ために、（次の分割位置に変更：５０５）では、所定の
規則に従って次々と被分割抽出範囲５０４を分割する位
置を変更し、決定した分割位置を後続する処理に渡す。
例えば、分割位置を［Ｒ軸で２等分］［Ｇ軸で２等分］
［Ｂ軸で２等分］の３種類であるとすれば、この３種類
の分割位置が（次の分割位置に変更：５０５）が実行さ
れる度に順番に切り替わって分割位置を決定する。ある
いは、同じ２分でも軸上の分割位置を２等分点に限らず
連続的に変化させても良いし、また、２分割ではなく分
割数を増したｎ分割である場合を考える事もできる。(Change to next division position: 505) In order to determine a position to divide the divided color extraction range 504 in the subsequent processing, in (Change to next division position: 505), one after another according to a predetermined rule. The position where the divided extraction range 504 is divided is changed, and the determined dividing position is passed to the subsequent processing.
For example, divide the position into two equal parts on the R axis and two equal parts on the G axis.
If there are three types of [divided into two equal parts on the B axis], these three types of division positions are switched in order each time (change to next division position: 505) is executed to determine the division position. Alternatively, even in the same two minutes, the division position on the axis may be continuously changed without being limited to the bisector, and it is also possible to consider a case where the number of divisions is n instead of two. .

【００２８】（被分割色抽出範囲の分割：５０６）上述
の（次の色抽出範囲を選択：５０３）で選択された被分
割色抽出範囲５０４を、上述の（次の分割位置に変更：
５０５）で決定した分割位置で分割し、分割範囲５０７
−１から分割範囲５０７−ｎに分割する。例えば、被分
割範囲５０４をＲ軸でｎ分割するとすれば、Ｒｃ１，Ｒ
ｃ２〜Ｒｃｎ−１のｎ−１個の分割点がある。被分割範
囲５０４のＲ軸の要素（Ｒｄ，Ｒｕ）と各分割点の値と
の大小関係は（Ｒｄ＜Ｒｃ１＜Ｒｃ２＜〜＜Ｒｃｎ−１
＜Ｒｕ）とする。ｎ個の分割範囲５０７−１〜５０７−
ｎについて、Ｇ要素とＢ要素についてはそのままで、そ
れぞれのＲ要素について、５０７−１は（Ｒｄ，Ｒｃ
１）、５０７−２は（Ｒｃ１，Ｒｃ２）と続き、５０７
−ｊは（Ｒｃｊ−１，Ｒｃｊ）、最後の５０７−ｎは
（Ｒｃｎ−１，Ｒｕ）である。(Division of divided color extraction range: 506) The divided color extraction range 504 selected in the above (select next color extraction range: 503) is changed to the above (next division position:
Division at the division position determined in 505), and the division range 507
-1 to the division range 507-n. For example, if the divided range 504 is divided into n on the R axis, Rc1, R
There are n-1 division points from c2 to Rcn-1. The magnitude relationship between the R-axis element (Rd, Ru) of the divided range 504 and the value of each division point is (Rd <Rc1 <Rc2 <-<Rcn-1.
<Ru). n division ranges 507-1 to 507-
For n, the G element and the B element remain the same, and for each R element, 507-1 is (Rd, Rc
1) and 507-2 follow (Rc1, Rc2), and 507
-J is (Rcj-1, Rcj), and the last 507-n is (Rcn-1, Ru).

【００２９】（外接範囲の取得：５０８）本実施例にお
ける処理の過程では分割範囲５０７の数ｎだけ外接範囲
の取得処理が行われる。図５ではｎ個の分割範囲５０７
−１〜５０７−ｎに対応した外接範囲の取得である５０
８−１〜５０８−ｎがあり、それぞれの分割範囲に専用
の外接範囲取得処理が用意された形で示している。もち
ろん図示された形に従って実施してもよいが、ｎ回の繰
り返し処理によって処理を進める形で実施するのが一般
的とされる。ｊ番目の分割範囲５０７−ｊについて、抽
出分布１０６の分割範囲５０７−ｊの内部に含まれる部
分だけの色分布に外接する色抽出範囲を求めるには、ま
ず、抽出分布１０６を構成する色情報の全てを調べて、
色情報のＲＧＢの３要素がすべて分割範囲５０７−ｊの
要素の範囲内にある色情報だけを選択する。つまりｉ番
目の色情報ならば（Ｒｄｊ≦Ｒｉ≦Ｒｕｊ，Ｇｄｊ≦Ｇ
ｉ≦Ｇｕｊ，Ｂｄｊ≦Ｂｉ≦Ｂｕｊ）である色情報につ
いてのみ、ＲＧＢ各要素の最大値と最小値を求め、色抽
出範囲の要素とすればよい。(Acquisition of circumscribing range: 508) In the process of the processing in this embodiment, the circumscribing range is acquired by the number n of the divided ranges 507. In FIG. 5, n division ranges 507
It is the acquisition of the circumscribed range corresponding to -1 to 507-n 50
8-1 to 508-n, which are shown in the form in which dedicated circumscribing range acquisition processing is prepared for each divided range. Of course, it may be carried out according to the illustrated form, but it is generally carried out in a form of advancing the process by repeating the process n times. For the j-th divided range 507-j, to obtain the color extraction range circumscribing the color distribution of only the portion included in the inside of the divided range 507-j of the extracted distribution 106, first, the color information forming the extracted distribution 106 is obtained. Check all of
Only the color information in which all three RGB elements of the color information are within the range of the elements of the division range 507-j is selected. That is, if it is the i-th color information (Rdj ≦ Ri ≦ Ruj, Gdj ≦ G
Only for the color information of i ≦ Guj, Bdj ≦ Bi ≦ Buj), the maximum value and the minimum value of each RGB element may be obtained and set as the element of the color extraction range.

【００３０】（最小の色抽出範囲の組み合わせを選択：
５０９）本実施例では、複数の色抽出範囲の組み合わせ
である色分布抽出範囲によって抽出分布１０６の輪郭を
近似している。色分布抽出範囲を構成する色抽出範囲
は、抽出分布１０６に外接する、即ち抽出分布１０６全
体を内包する１つの色抽出範囲から、色抽出範囲の分割
と、分割されたそれぞれの色抽出範囲の内部に含まれる
部分だけの色分布に外接する色抽出範囲を求める事で抽
出分布１０６の輪郭を近似して行く。従って、ある被分
割色抽出範囲５０４について色抽出範囲の分割を行い、
それぞれに外接範囲を求めた色抽出範囲の組み合わせ
の、抽出範囲１０６に対する近似の程度が高いという事
は、外接範囲の取得５０８−１〜５０８−ｎで得られた
ｎ個の色抽出範囲の大きさ、即ちｎ個の色分布抽出範囲
の体積の総和が小さいという事である。１つの色抽出範
囲の体積は図４（ｂ）に示した色抽出範囲１１１−ｊで
ある直方体の体積である。(Select the smallest combination of color extraction ranges:
509) In this embodiment, the contour of the extraction distribution 106 is approximated by the color distribution extraction range that is a combination of a plurality of color extraction ranges. The color extraction range that constitutes the color distribution extraction range is divided into a color extraction range that is circumscribing the extraction distribution 106, that is, one color extraction range that includes the entire extraction distribution 106, and a color extraction range of each of the divided color extraction ranges. The contour of the extraction distribution 106 is approximated by obtaining the color extraction range circumscribing the color distribution of only the portion included inside. Therefore, the color extraction range is divided for a certain divided color extraction range 504,
The high degree of approximation of the combination of the color extraction ranges for which the circumscribing ranges are obtained to the extraction range 106 means that the size of the n color extraction ranges obtained in the acquisition of the circumscribing ranges 508-1 to 508-n is large. That is, the sum of the volumes of the n color distribution extraction ranges is small. The volume of one color extraction range is the volume of a rectangular parallelepiped that is the color extraction range 111-j shown in FIG.

【００３１】本実施例による処理では、ある被分割色抽
出範囲５０４について（次の分割位置に変更：５０５）
によって分割位置を変えて色抽出範囲５０４を分割し、
それぞれの色抽出範囲について（外接範囲の取得：５０
８）で外接範囲を求めたｎ個の色抽出範囲の組み合わせ
が次々と生成される。（最小の色抽出範囲の組み合わせ
を選択：５０９）は、その中から最もｎ個の色分布抽出
範囲の体積の総和が小さい色抽出範囲の組み合わせを選
択し、最小の色抽出範囲の組み合わせ５１０とする。即
ち、最小の色抽出範囲の組み合わせ５１０は、被分割色
抽出範囲５０４を分割して得られる最も近似度の高いｎ
個の色抽出範囲の組み合わせである。最小の色抽出範囲
の組み合わせ５１０を得るために、最小の色抽出範囲の
組み合わせ５１０はｎ個の色抽出範囲とともに、比較用
の情報として、比較用の色抽出範囲の大きさの総和を持
つ。次の色抽出範囲を選択５０３である被分割色抽出範
囲５０４が新たに選択された時点で、前記比較用の色抽
出範囲の大きさの総和は最大値に初期化される。処理の
過程で（外接範囲の取得：５０８）からｎ個の色抽出範
囲の組み合わせが得られる度に、ｎ個の色抽出範囲の大
きさの総和と前記比較用の色抽出範囲の大きさの総和を
比較し、ｎ個の色抽出範囲の大きさの総和の方が小さけ
れば、最小の色抽出範囲の組み合わせ５１０にｎ個の色
抽出範囲を代入し、さらに、前記比較用の色抽出範囲の
大きさにｎ個の色抽出範囲の総和を代入する。In the processing according to the present embodiment, for a certain divided color extraction range 504 (change to the next dividing position: 505).
The color extraction range 504 is divided by changing the division position by
For each color extraction range (acquisition of circumscribed range: 50
The combination of n color extraction ranges for which the circumscribing range is obtained in 8) is generated one after another. (Selecting the smallest combination of color extraction ranges: 509) selects the combination of the color extraction ranges having the smallest total volume of the n color distribution extraction ranges from among them, and the combination 510 of the smallest color extraction ranges is selected. To do. That is, the minimum color extraction range combination 510 is the highest n obtained by dividing the divided color extraction range 504.
This is a combination of individual color extraction ranges. In order to obtain the minimum color extraction range combination 510, the minimum color extraction range combination 510 has the sum of the sizes of the comparison color extraction ranges as information for comparison, together with the n color extraction ranges. When the divided color extraction range 504, which is the selection 503 of the next color extraction range, is newly selected, the total sum of the sizes of the color extraction ranges for comparison is initialized to the maximum value. Every time a combination of n color extraction ranges is obtained from (acquisition of circumscribing range: 508) in the process of processing, the sum of the sizes of the n color extraction ranges and the size of the color extraction range for comparison are calculated. If the sums of the sizes of the n color extraction ranges are smaller, the n color extraction ranges are assigned to the smallest color extraction range combination 510, and the comparison color extraction ranges are further compared. Substitute the sum of n color extraction ranges for the size of.

【００３２】（全ての分割位置で分割？：５１１）ある
被分割色抽出範囲５０４について、（次の分割位置に変
更：５０５）で決定される全ての分割位置で被分割色抽
出範囲５０４を分割し、それぞれの色抽出範囲について
（外接範囲の取得：５０８）で外接範囲を求めたｎ個の
色抽出範囲の組み合わせから（最小の色抽出範囲の組み
合わせを選択：５０９）によって、その中から最もｎ個
の色分布抽出範囲の体積の総和が小さい色抽出範囲の組
み合わせを選択して最小の色抽出範囲の組み合わせ５１
０とする処理を繰り返せば、被分割色抽出範囲５０４を
分割して得られる最も近似度の高いｎ個の色抽出範囲の
組み合わせである。従って、（次の分割位置に変更：５
０５）で決定される全ての分割位置で被分割色抽出範囲
５０４を分割するまでは（次の分割位置に変更：５０
５）に戻って処理を繰り返し、（次の分割位置に変更：
５０５）で決定される全ての分割位置で被分割色抽出範
囲５０４を分割すれば次の処理に移る。(Split at all split positions ?: 511) With respect to a certain split color extraction range 504, split the split color extraction range 504 at all split positions determined by (change to next split position: 505). Then, from the combination of n color extraction ranges for which the circumscribing range was obtained for each color extraction range (acquisition of circumscribing range: 508) (select the smallest combination of color extraction ranges: 509), The minimum color extraction range combination 51 is selected by selecting a combination of color extraction ranges in which the total volume of the n color distribution extraction ranges is small.
If the process of setting 0 is repeated, it is a combination of n color extraction ranges with the highest degree of approximation obtained by dividing the divided color extraction range 504. Therefore, (change to the next division position: 5
Until the divided color extraction range 504 is divided at all the division positions determined in (05) (change to the next division position: 50
Return to 5) and repeat the process, (change to the next division position:
If the divided color extraction range 504 is divided at all the division positions determined in 505), the next process is performed.

【００３３】（生成色分布抽出範囲に追加：５１２）生
成色分布抽出範囲５１３は色分布抽出範囲である。生成
色分布抽出範囲５１３を構成する色抽出範囲として（次
の分割位置に変更：５０５）で決定される全ての分割位
置で被分割色抽出範囲５０４を分割して外接範囲を求め
る処理から選択された最小の色抽出範囲の組み合わせ５
１０は被分割色抽出範囲５０４を分割して得られる最も
近似度の高いｎ個の色抽出範囲の組み合わせを追加する
事で、生成色分布抽出範囲５１３は、被分割色抽出範囲
５０２より抽出分布１０６に対する近似の度合いの高い
色抽出範囲の組み合わせを要素として持つ事になる。被
分割色分布抽出範囲５０２から（次の色抽出範囲を選
択：５０３）によって、すべての色抽出範囲を選択して
被分割色抽出範囲５０４として処理を繰り返せば、生成
色分布抽出範囲５１３は被分割色分布抽出範囲５０２の
ｎ倍の数の色抽出範囲で構成され、もとの被分割色抽出
範囲５０２より抽出分布１０６に対する近似の度合いの
高い色分布抽出範囲となる。(Addition to the generated color distribution extraction range: 512) The generated color distribution extraction range 513 is a color distribution extraction range. It is selected from the process of dividing the to-be-divided color extraction range 504 to obtain the circumscribing range as a color extraction range forming the generated color distribution extraction range 513 (change to next division position: 505). Combination of minimum color extraction range 5
10 is to add a combination of n color extraction ranges with the highest degree of approximation obtained by dividing the divided color extraction range 504, so that the generated color distribution extraction range 513 is extracted from the divided color extraction range 502. A combination of color extraction ranges having a high degree of approximation with respect to 106 is included as an element. If all the color extraction ranges are selected from the divided color distribution extraction range 502 (select next color extraction range: 503) and the process is repeated as the divided color extraction range 504, the generated color distribution extraction range 513 is The color distribution extraction range is composed of n times as many color extraction ranges as the divided color distribution extraction range 502, and has a higher degree of approximation to the extraction distribution 106 than the original divided color extraction range 502.

【００３４】（すべての色抽出範囲を選択？：５１４）
被分割色分布抽出範囲５０２について、（次の色抽出範
囲を選択：５０３）によってすべての色抽出範囲を選択
し、被分割色抽出範囲５０４として処理を繰り返せば、
生成色分布抽出範囲５１３は、もとの被分割色抽出範囲
５０２より抽出分布１０６に対する近似の度合いの高い
色分布抽出範囲となる。従って、（次の抽出範囲を選
択：５０３）で、被分割色分布抽出範囲５０２を構成す
るすべての色抽出範囲について処理するまでは（次の抽
出範囲を選択：５０３）に戻って処理を繰り返し、（次
の抽出範囲を選択：５０３）で、被分割色分布抽出範囲
５０２を構成するすべての色抽出範囲について処理すれ
ば次の処理に移る。(Select all color extraction ranges ?: 514)
For the divided color distribution extraction range 502, if all the color extraction ranges are selected by (select next color extraction range: 503) and the process is repeated as the divided color extraction range 504,
The generated color distribution extraction range 513 is a color distribution extraction range having a higher degree of approximation to the extraction distribution 106 than the original divided color extraction range 502. Therefore, until (select next extraction range: 503) processes all the color extraction ranges forming the divided color distribution extraction range 502, the process returns to (select next extraction range: 503) and repeats the processing. , (Select next extraction range: 503), if all the color extraction ranges forming the divided color distribution extraction range 502 have been processed, the next process is performed.

【００３５】（まだ色抽出範囲を増やせる？：５１５）
色分布抽出範囲１１１は、処理画像１１２から色分布抽
出処理１１３によって処理対象を抽出するとき用いる色
抽出範囲の数ｋ個の色抽出範囲１１１−１〜１１１−ｋ
を記憶できるテーブルに記憶された色抽出範囲の集合で
ある。生成色分布抽出範囲５１３は被分割色分布抽出範
囲５０２のｎ倍の数の色抽出範囲で構成され、もとの被
分割色抽出範囲５０２より抽出分布１０６に対する近似
の度合いの高い色分布抽出範囲であるが、生成色分布抽
出範囲５１３を構成する色抽出範囲の数がｋ個に満たな
ければ、被分割色分布抽出範囲５０２に生成色分布抽出
範囲５０３を代入して処理を繰り返す必要がある。生成
色分布抽出範囲５１３を構成する色抽出範囲の数がｋ個
になれば生成色分布抽出範囲５１３を色分布抽出範囲取
得処理１０８の処理結果である色分布抽出範囲１１１と
して処理を終了する。 例えば、処理画像１１２から色
分布抽出処理１１３によって処理対象を抽出するとき用
いる色抽出範囲の数ｋを８とし、（被分割色抽出範囲の
分割：５０６）での被分割色抽出範囲５０４の分割数ｎ
を２とすれば、生成色分布抽出範囲５１３を構成する色
抽出範囲の総数は、１回目の処理で２、２回目の処理で
４、３回目の処理で８となって、３回目の処理で色分布
抽出範囲取得処理１０８の処理結果である色分布抽出範
囲１１１として処理を終了する。(Can the color extraction range still be increased ?: 515)
The color distribution extraction range 111 is the number k of color extraction ranges used when extracting a processing target from the processed image 112 by the color distribution extraction processing 113.
Is a set of color extraction ranges stored in a table that can store The generated color distribution extraction range 513 is composed of n times as many color extraction ranges as the divided color distribution extraction range 502, and has a higher degree of approximation to the extraction distribution 106 than the original divided color extraction range 502. However, if the number of color extraction ranges forming the generated color distribution extraction range 513 is less than k, it is necessary to substitute the generated color distribution extraction range 503 in the divided color distribution extraction range 502 and repeat the process. . When the number of color extraction ranges forming the generated color distribution extraction range 513 reaches k, the generated color distribution extraction range 513 is set as the color distribution extraction range 111 which is the processing result of the color distribution extraction range acquisition processing 108, and the processing is ended. For example, the number k of color extraction ranges used when extracting the processing target from the processed image 112 by the color distribution extraction process 113 is 8, and the division of the divided color extraction range 504 in (division of divided color extraction range: 506) is performed. Number n
2 is 2, the total number of color extraction ranges forming the generated color distribution extraction range 513 is 2 in the first process, 4 in the second process, 8 in the third process, and 8 in the third process. Then, the processing is ended as the color distribution extraction range 111 which is the processing result of the color distribution extraction range acquisition processing 108.

【００３６】（生成色分布抽出範囲を被分割色分布抽出
範囲とする：５１６）色分布抽出範囲１１１は、処理画
像１１２から色分布抽出処理１１３によって処理対象を
抽出するとき用いる色抽出範囲の数ｋ個の色抽出範囲２
０７−１〜２０７−ｋを記憶できるテーブルに記憶され
た色抽出範囲の集合であるから、生成色分布抽出範囲５
１３を構成する色抽出範囲の数がｋ個に満たなければ、
被分割色分布抽出範囲５０２に生成色分布抽出範囲５０
３を代入して処理を繰り返す。(The generated color distribution extraction range is the divided color distribution extraction range: 516) The color distribution extraction range 111 is the number of color extraction ranges used when extracting the processing target from the processed image 112 by the color distribution extraction processing 113. k color extraction range 2
Since it is a set of color extraction ranges stored in a table capable of storing 07-1 to 207-k, the generated color distribution extraction range 5
If the number of color extraction ranges forming 13 is less than k,
Generated color distribution extraction range 50 in divided color distribution extraction range 502
Substitute 3 and repeat the process.

【００３７】（生成色分布抽出範囲を色分布抽出範囲と
する：５１７）色分布抽出範囲１１１は、処理画像１１
２から色分布抽出処理１１３によって処理対象を抽出す
るとき用いる色抽出範囲の数ｋ個の色抽出範囲２０７−
１〜２０７−ｋを記憶できるテーブルに記憶された色抽
出範囲の集合であるから、生成色分布抽出範囲５１３を
構成する色抽出範囲の数がｋ個になれば生成色分布抽出
範囲５１３を色分布抽出範囲取得処理２０６の処理結果
である色分布抽出範囲１１１として処理を終了する。(The generated color distribution extraction range is the color distribution extraction range: 517) The color distribution extraction range 111 is the processed image 11
The number k of color extraction ranges used when extracting the processing target from 2 through the color distribution extraction process 113.
1 to 207-k is a set of color extraction ranges stored in a table, the generated color distribution extraction range 513 is colored when the number of color extraction ranges forming the generation color distribution extraction range 513 becomes k. The process ends as the color distribution extraction range 111, which is the processing result of the distribution extraction range acquisition process 206.

【００３８】図７及び図８は、本発明による画像処理方
法の実施例における色分布抽出処理１１３の処理内容を
説明するための流れ図である。色分布抽出処理１１３
は、処理対象１０２の有する色分布である抽出分布１０
６の輪郭を複数の色抽出範囲で近似した色分布抽出範囲
１１１を得て、処理画像１１２のうち、色分布抽出範囲
１１１を構成する複数の色抽出範囲のどれかに含まれる
色である部分を抽出する事で、処理対象である部分のみ
を抽出した画像データである抽出画像１１４を得る。7 and 8 are flow charts for explaining the processing contents of the color distribution extraction processing 113 in the embodiment of the image processing method according to the present invention. Color distribution extraction processing 113
Is the extraction distribution 10 that is the color distribution of the processing target 102.
A color distribution extraction range 111 obtained by approximating the contour of No. 6 with a plurality of color extraction ranges is obtained, and a portion of the processed image 112 that is a color included in any of the plurality of color extraction ranges forming the color distribution extraction range 111. Is extracted to obtain an extracted image 114 that is image data obtained by extracting only the portion to be processed.

【００３９】（ｙをＹ軸の開始位置：６０１） （ｘをＸ軸の開始位置：６０２）まず画像データ上の処
理座標である（ｘ，ｙ）をＸ軸及びＹ軸の開始位置に初
期化する。本実施例では全体の処理の構成上（ｙをＹ軸
の開始位置：６０１）と（ｘをＸ軸の開始位置：６０
２）の２つに分けている。例えば処理画像全体で色分布
抽出処理を行うなら（ｘ，ｙ）を（０，０）に初期化す
る。(Y is the starting position of the Y axis: 601) (x is the starting position of the X axis: 602) First, the processing coordinates (x, y) on the image data are initially set at the starting positions of the X axis and the Y axis. Turn into. In the present embodiment, in terms of the overall processing configuration (y is Y-axis start position: 601) and (x is X-axis start position: 60).
It is divided into 2) of 2). For example, if color distribution extraction processing is performed on the entire processed image, (x, y) is initialized to (0, 0).

【００４０】（色情報の読み取り：６０３）処理画像１
１２より前記処理座標（ｘ，ｙ）の画素の色情報を読み
取る。本実施例ではＲＧＢの３要素で色情報を表現して
いるので、読み取った色情報６０４は（６０４−Ｒ，６
０４−Ｇ，６０４−Ｂ）である。(Reading of color information: 603) Processed image 1
From 12, the color information of the pixel at the processing coordinate (x, y) is read. In this embodiment, since the color information is expressed by the three elements of RGB, the read color information 604 is (604-R, 6).
04-G, 604-B).

【００４１】（色抽出範囲との比較：６０５）色情報６
０４が色分布抽出範囲１１１を構成する色抽出範囲の内
部にあるかどうかを判定する。色分布抽出範囲１１１
は、処理画像１１２から色分布抽出処理１１３によって
処理対象を抽出するとき用いる色抽出範囲の数ｋ個の色
抽出範囲１１−１〜１１−ｋを記憶できるテーブルに記
憶された色抽出範囲の集合であるから、色抽出範囲の数
ｋだけ色情報６０４と色分布抽出範囲１１１を構成する
色抽出範囲との比較が行われる。図７ではｋ個の色抽出
範囲１１１−１〜１１１−ｋに対応した色抽出範囲との
比較である６０５−１〜６０５−ｋがあり、それぞれの
色抽出範囲に専用の色抽出範囲との比較が用意された形
で示している。もちろん図示された形に従って実施して
もよいが、ｋ回の繰り返し処理によって処理を進める形
で実施するのが一般的とされる。読み取った色情報６０
４は（６０４−Ｒ，６０４−Ｇ，６０４−Ｂ）であり、
ｊ番目の色抽出範囲は（Ｒｊｕ，Ｒｊｄ，Ｇｊｕ，Ｇｊ
ｄ，Ｂｊｕ，Ｂｊｄ）であるから、色情報６０４がｊ番
目の色抽出範囲の内部にあるという事は、それぞれの要
素毎に３つの条件式があって、３つとも成立するという
事である。つまり、（Ｒｊｄ≦６０４−Ｒ≦Ｒｊｕ）か
つ（Ｇｊｄ≦６０４−Ｇ≦Ｇｊｕ）かつ（Ｂｊｄ≦６０
４−Ｂ≦Ｂｊｕ）であればｊ番目の色抽出範囲との比較
結果は［真］であり、１つでも成立していなければｊ番
目の色抽出範囲との比較結果は［偽］である。(Comparison with the color extraction range: 605) Color information 6
It is determined whether or not 04 is inside the color extraction range forming the color distribution extraction range 111. Color distribution extraction range 111
Is a set of color extraction ranges stored in a table capable of storing the number k of color extraction ranges 11-1 to 11-k used when extracting a processing target from the processed image 112 by the color distribution extraction process 113. Therefore, the number k of color extraction ranges is compared with the color information 604 and the color extraction ranges forming the color distribution extraction range 111. In FIG. 7, there are 605-1 to 605-k, which are comparisons with the color extraction ranges corresponding to the k color extraction ranges 111-1 to 111-k, and each color extraction range has a dedicated color extraction range. The comparison is shown in a prepared form. Of course, the process may be performed according to the illustrated form, but it is general to perform the process by repeating the process k times. Read color information 60
4 is (604-R, 604-G, 604-B),
The jth color extraction range is (Rju, Rjd, Gju, Gj
d, Bju, Bjd), the fact that the color information 604 is inside the jth color extraction range means that there are three conditional expressions for each element and all three are satisfied. . That is, (Rjd ≦ 604−R ≦ Rju) and (Gjd ≦ 604−G ≦ Gju) and (Bjd ≦ 60)
If 4-B ≦ Bju), the comparison result with the jth color extraction range is [true], and if even one is not established, the comparison result with the jth color extraction range is [false]. .

【００４２】（どれかの抽出範囲に入る？：６０７）比
較結果６０６は色情報６０４が色分布抽出範囲１１１を
構成する色抽出範囲１１１−１〜１１１−ｋの内部にあ
るかどうかのそれぞれの比較結果６０６−１〜６０６−
ｋの集合である。色情報６０４がどれかの色抽出範囲の
内部に入っているという事は、すべての色抽出範囲毎の
比較結果６０６−１〜６０６−ｋの論理和を求めた結果
が［真］であるという事である。色情報６０４がどれか
の色抽出範囲の内部に入っていれば、色情報６０４は色
分布抽出範囲１１１の内部に含まれる事を意味する。(Are any of the extraction ranges ?: 607) The comparison result 606 indicates whether the color information 604 is within the color extraction ranges 111-1 to 111-k constituting the color distribution extraction range 111. Comparison results 606-1 to 606-
It is a set of k. The fact that the color information 604 is inside any of the color extraction ranges means that the result of obtaining the logical sum of the comparison results 606-1 to 606-k for all the color extraction ranges is [true]. It is a thing. If the color information 604 is inside any of the color extraction ranges, it means that the color information 604 is included inside the color distribution extraction range 111.

【００４３】（画素の値は０：６０８） （画素の値は１：６０９）本実施例では、色情報６０４
がどれかの色抽出範囲の内部に入っていれば、色情報６
０４は色分布抽出範囲１１１の内部に含まれる事を意味
する画素の値［１は色分布抽出範囲１１１の内部に含ま
れない事を意味する画素の値［０］を準備する。画素の
値が［１］である事は色情報６０４は色分布抽出範囲２
０７の内部に含まれ、その画素を抽出した事を意味す
る。もちろん［１］と［０］の替わりに別の値を用いて
もかまわない。(Pixel value is 0: 608) (Pixel value is 1: 609) In this embodiment, color information 604
Is within any color extraction range, color information 6
A pixel value 04 is a pixel value [1] which means that the pixel value is included in the color distribution extraction range 111 [1 is a pixel value [0] which is not included in the color distribution extraction range 111. The fact that the pixel value is [1] means that the color information 604 indicates the color distribution extraction range 2
It is included in 07 and means that the pixel is extracted. Of course, another value may be used instead of [1] and [0].

【００４４】（抽出画像に書き込み：６１０）色情報６
０４である画素が色分布抽出範囲の内部に含まれ、抽出
したか否かを示す画素の値を抽出画像１１４の前記処理
座標（ｘ，ｙ）に書き込む。前記処理座標（ｘ，ｙ）に
ついて、Ｘ軸の開始位置及びＹ軸の開始位置からＸ軸の
終了位置及びＹ軸の終了位置まで上述の処理を行えば、
抽出画像１１４は抽出した画素の値である［１］と抽出
しなかった画素の値である［０］とからなる画像データ
となる。画素の値が［１］である領域が、処理画像１１
２のうち抽出分布１０６に含まれる色である領域であ
る。(Write in extracted image: 610) Color information 6
The pixel of 04 is included in the color distribution extraction range, and the value of the pixel indicating whether or not the pixel is extracted is written in the processing coordinate (x, y) of the extracted image 114. For the processing coordinates (x, y), if the above processing is performed from the X-axis start position and the Y-axis start position to the X-axis end position and the Y-axis end position,
The extracted image 114 is image data composed of the extracted pixel value [1] and the unextracted pixel value [0]. The area where the pixel value is [1] is the processed image 11
It is a region that is a color included in the extraction distribution 106 out of 2.

【００４５】（ｘはＸ軸終了位置？：６１１） （次のｘ：６１２）前記処理座標（ｘ，ｙ）のｘが終了
位置に達するまでｘを次のｘに変更し、（色情報の読み
取り：６０３）に戻って処理を繰り返し、Ｘ軸方向につ
いての処理を行う。ｘが終了位置に達したらＹ軸方向の
変更にうつる。(X is the X-axis end position ?: 611) (Next x: 612) Change x to the next x until x of the processing coordinate (x, y) reaches the end position, and The process returns to reading: 603) and the process is repeated to perform the process in the X-axis direction. When x reaches the end position, change in the Y-axis direction is started.

【００４６】（ｙはＹ軸終了位置？：６１３） （次のｙ：６１４）前記処理座標（ｘ，ｙ）のｙが終了
位置に達するまでｙを次のｙに変更し、（ｘをＸ軸の開
始位置：６０２）に戻って処理を繰り返し、Ｙ軸方向に
ついての処理を行う。ｙを次のｙに変更するときは、変
更前のｙについてＸ軸方向の処理は終了しているので、
ｙを次のｙに変更した後はｘをＸ軸の開始位置に初期化
する。従って、Ｙ軸方向についての処理の繰り返しは
（ｘをＸ軸の開始位置：６０２）に戻っての処理の繰り
返しとなる。(Y is the Y-axis end position ?: 613) (Next y: 614) y is changed to the next y until y of the processing coordinate (x, y) reaches the end position, and (x is X The process is repeated by returning to the axis start position: 602), and the process in the Y-axis direction is performed. When y is changed to the next y, the processing in the X-axis direction for y before the change is completed,
After changing y to the next y, x is initialized to the start position of the X axis. Therefore, the process is repeated in the Y-axis direction after returning to (x is the start position of the X-axis: 602).

【００４７】図９から図１５は、上述の色分布抽出範囲
取得処理１０８の実施例による色分布抽出範囲取得処理
の過程を説明するための説明図である。実際の色分布は
色情報が３つの要素で表現されるので３次元立体的な形
状であるが、ここでは、説明を簡単にするために平面的
な形状であるとする。また、説明を簡単にするために、
（被分割色抽出範囲の分割：５０６）の分割数を２に限
定し、さらに（次の分割位置に変更：５０５）が定める
分割位置は被分割色抽出範囲５０４を２等分する位置で
あって、分割する方向を図面垂直方向及び水平方向の２
種類に限定する。また、色分布抽出範囲１１１は４つの
色抽出範囲で抽出分布１０６を近似するものとする。FIGS. 9 to 15 are explanatory diagrams for explaining the process of the color distribution extraction range acquisition process according to the embodiment of the color distribution extraction range acquisition process 108 described above. The actual color distribution has a three-dimensional three-dimensional shape because color information is represented by three elements, but here it is assumed that it has a two-dimensional shape to simplify the description. Also, to simplify the explanation,
The number of divisions of (divided color extraction range division: 506) is limited to 2, and the division position determined by (change to next division position: 505) is a position that divides the divided color extraction range 504 into two equal parts. The vertical direction and horizontal direction in the drawing.
Limited to types. Further, the color distribution extraction range 111 approximates the extraction distribution 106 with four color extraction ranges.

【００４８】図９について説明する。色空間４０１は色
情報の表現範囲であり、抽出分布１０６は色空間４０１
の内部に存在する。（被分割色抽出範囲の初期化：５０
１）で、抽出分布１０６に外接する１つの色抽出範囲で
被分割色分布抽出範囲５０２が初期化される。この初期
化された被分割色分布抽出範囲５０２は抽出分布１０６
の輪郭を１つの色抽出範囲で近似した色分布抽出範囲と
考えることができる。FIG. 9 will be described. The color space 401 is the expression range of color information, and the extraction distribution 106 is the color space 401.
Exists inside. (Initialization of divided color extraction range: 50
In 1), the divided color distribution extraction range 502 is initialized with one color extraction range circumscribing the extraction distribution 106. The initialized divided color distribution extraction range 502 is the extraction distribution 106.
Can be considered as a color distribution extraction range that is approximated by one color extraction range.

【００４９】図１０について説明する。（次の色抽出範
囲を選択：５０３）で被分割色分布抽出範囲５０２から
初期値である（抽出分布１０６に外接する１つの色抽出
範囲）を被分割色抽出範囲５０４として取り出す。分割
位置を（次の分割位置に変更：５０５）で水平方向２分
割と定めれば、被分割色抽出範囲５０２は（被分割色抽
出範囲の分割：５０６）で分割範囲５０７−１及び５０
７−２に分割される。色抽出範囲１００１は（外接範囲
の取得：５０８−１）によって取得される、抽出分布１
０６の分割範囲５０７−１に含まれる部分のみに外接す
る色抽出範囲である。同様に、色抽出範囲１００２は
（外接範囲の取得：５０８−２）によって取得される、
抽出分布１０６の分割範囲５０７−２に含まれる部分の
みに外接する色抽出範囲である。（最小の色抽出範囲の
組み合わせを選択：５０９）では、色抽出範囲の大きさ
の最大値に初期化された状態から始まるので、色抽出範
囲１００１及び１００２の組み合わせを最小の色抽出範
囲の組み合わせ５１０とする。ここまででは、水平方向
２分割について処理しただけなので、（全ての分割位置
で分割？：５１１）では、次の垂直方向２分割で処理を
繰り返すために（次の分割位置に変更：５０５）に戻っ
て処理を繰り返す。Referring to FIG. In (Select next color extraction range: 503), an initial value (one color extraction range circumscribing the extraction distribution 106) is extracted from the divided color distribution extraction range 502 as the divided color extraction range 504. If the dividing position is set to (change to the next dividing position: 505) to be two-division in the horizontal direction, the divided color extraction range 502 is (divided color extraction range division: 506) and the divided ranges 507-1 and 50.
It is divided into 7-2. The color extraction range 1001 is the extraction distribution 1 acquired by (acquisition of circumscribing range: 508-1).
This is a color extraction range circumscribing only the part included in the division range 507-1 of 06. Similarly, the color extraction range 1002 is acquired by (acquisition of circumscribing range: 508-2),
This is a color extraction range circumscribing only the portion included in the division range 507-2 of the extraction distribution 106. In (Select minimum color extraction range combination: 509), since the state is initialized to the maximum value of the size of the color extraction range, the combination of the color extraction ranges 1001 and 1002 is set to the minimum color extraction range combination. It is set to 510. Up to this point, processing has been performed only for the horizontal two-division, so in (division at all divisions ?: 511), the processing is repeated for the next two vertical divisions (change to the next division position: 505). Return and repeat the process.

【００５０】図１１について説明する。図１０について
説明した処理の過程と同様に、（次の色抽出範囲を選
択：５０３）で被分割色分布抽出範囲５０２から初期値
である（抽出分布１０６に外接する１つの色抽出範囲）
を被分割色抽出範囲５０４として取り出し、分割位置を
（次の分割位置に変更：５０５）で垂直方向２分割と定
め、被分割色抽出範囲５０２は（被分割色抽出範囲の分
割：５０６）で分割範囲５０７−１及び５０７−２に分
割される。色抽出範囲１１０１は（外接範囲の取得：５
０８−１）によって取得される、抽出分布１０６の分割
範囲５０７−１に含まれる部分のみに外接する色抽出範
囲である。同様に、色抽出範囲１１０２は（外接範囲の
取得：５０８−２）によって取得される、抽出分布１０
６の分割範囲５０７−２に含まれる部分のみに外接する
色抽出範囲である。FIG. 11 will be described. Similar to the process described with reference to FIG. 10, an initial value is set from the divided color distribution extraction range 502 in (select next color extraction range: 503) (one color extraction range circumscribing the extraction distribution 106).
As the divided color extraction range 504, the division position is defined as (division into the next division position: 505) in the vertical direction, and the divided color extraction range 502 is (divided color extraction range division: 506). It is divided into division ranges 507-1 and 507-2. The color extraction range 1101 is (acquisition of circumscribed range: 5
08-1) is the color extraction range circumscribing only the portion included in the division range 507-1 of the extraction distribution 106. Similarly, the color extraction range 1102 is acquired by (acquisition of circumscribing range: 508-2).
6 is a color extraction range circumscribing only a portion included in the division range 507-2.

【００５１】図１２について説明する。図１０に示した
水平方向２分割の場合と図１１に示した垂直方向２分割
の場合とを比較すると、色抽出範囲１００１と１００２
の面積の総和の方が、色抽出範囲１１０１と１１０２の
面積の総和よりも小さい。（最小の色抽出範囲の組み合
わせを選択：５０９）では、色抽出範囲１００１と１０
０２の組み合わせの方を選択し、最小の色抽出範囲の組
み合わせ５１０は色抽出範囲１００１と１００２の組み
合わせとなる。水平方向と垂直方向の両方の分割位置に
ついて処理を行ったので、（全ての分割位置で分割？：
５１１）では、次の処理である（生成色分布抽出範囲に
追加：５１２）に移る。（生成色分布抽出範囲に追加：
５１２）では、最小の色抽出範囲の組み合わせ５１０を
生成色分布抽出範囲５１３に追加する。生成色分布抽出
範囲５１３を構成する色抽出範囲数は２となる。ここま
での処理では、被分割色分布抽出範囲５０２は抽出分布
１０６に外接する１つの色抽出範囲のみで構成されてい
たので、１つの色抽出範囲を被分割色抽出範囲５０４と
して処理を進めればすべての色抽出範囲を選択した事に
なる。従って、（すべての色抽出範囲を選択？：５１
４）では、次の処理である（まだ色抽出範囲を増やせる
？：５１５）に移る。ここまでの処理で、生成色分布抽
出範囲５１３を構成する色抽出範囲の総数は２である。
ここでは、色分布抽出範囲１１１は４つの色抽出範囲で
抽出分布１０６を近似するとしているので、生成色分布
抽出範囲５１３は色抽出範囲の数をまだ増やす事ができ
る。従って、（まだ色抽出範囲を増やせる？：５１５）
では、（生成色分布抽出範囲を被分割色抽出範囲とす
る：５１６）で生成色分布抽出範囲５１３の内容を被分
割色分布抽出範囲５０２に複写して、（次の色抽出範囲
を選択：５０３）に戻って処理を繰り返す。FIG. 12 will be described. Comparing the case of the horizontal two divisions shown in FIG. 10 with the case of the two vertical divisions shown in FIG. 11, the color extraction ranges 1001 and 1002 are shown.
Is smaller than the total area of the color extraction ranges 1101 and 1102. In (select the smallest combination of color extraction ranges: 509), the color extraction ranges 1001 and 10
The combination of the color extraction ranges 1001 and 1002 is selected as the minimum color extraction range combination 510. Since the processing has been performed for both the horizontal and vertical division positions, (the division at all division positions ?:
In 511), the process proceeds to the next process (add to generated color distribution extraction range: 512). (Add to generated color distribution extraction range:
In 512), the minimum color extraction range combination 510 is added to the generated color distribution extraction range 513. The number of color extraction ranges forming the generated color distribution extraction range 513 is 2. In the processing up to this point, the divided color distribution extraction range 502 is composed of only one color extraction range circumscribing the extraction distribution 106, and therefore one color extraction range is set as the divided color extraction range 504. For example, all color extraction ranges are selected. Therefore, (select all color extraction ranges ?: 51
In step 4), the process proceeds to the next process (can the color extraction range still be increased ?: 515). By the processing up to this point, the total number of color extraction ranges forming the generated color distribution extraction range 513 is 2.
Here, since the color distribution extraction range 111 approximates the extraction distribution 106 with four color extraction ranges, the generated color distribution extraction range 513 can still increase the number of color extraction ranges. Therefore, (can you still increase the color extraction range ?: 515)
Then, the content of the generated color distribution extraction range 513 is copied to the divided color distribution extraction range 502 in (the generated color distribution extraction range is set as the divided color extraction range: 516), and (the next color extraction range is selected: The process is repeated by returning to step 503).

【００５２】図１３について説明する。生成色分布抽出
範囲５１３を複写した被分割色分布抽出範囲５０２から
（次の色抽出範囲を選択：５０３）で色抽出範囲を一つ
選んで被分割色抽出範囲５０４とすれば、図１３の色抽
出範囲１３０１と１３０２は水平方向２分割の場合の、
２つの分割範囲５０７−１及び５０７−２に含まれる部
分のみの抽出分布１０６に外接する色抽出範囲である。FIG. 13 will be described. From the divided color distribution extraction range 502 obtained by copying the generated color distribution extraction range 513 (selecting the next color extraction range: 503), one color extraction range is selected as the divided color extraction range 504. When the color extraction ranges 1301 and 1302 are divided into two in the horizontal direction,
It is a color extraction range circumscribing the extraction distribution 106 of only the part included in the two divided ranges 507-1 and 507-2.

【００５３】図１４について説明する。図１４の色抽出
範囲１４０１と１４０２は、図１３の場合と同じ被分割
色抽出範囲５０４について、垂直方向２分割の場合で
の、２つの分割範囲５０７−１及び５０７−２に含まれ
る部分のみの抽出分布１０６に外接する色抽出範囲であ
る。図１３及び図１４の場合、（最小の色抽出範囲の組
み合わせを選択：５０９）で最小の色抽出範囲の組み合
わせ５１０に保持される色抽出範囲の組み合わせは、図
１３の場合の色抽出範囲１３０１と１３０２の組み合わ
せである。FIG. 14 will be described. The color extraction ranges 1401 and 1402 in FIG. 14 are only the portions included in the two division ranges 507-1 and 507-2 in the case where the same divided color extraction range 504 as in FIG. 13 is divided into two in the vertical direction. Is a color extraction range circumscribing the extraction distribution 106 of. In the case of FIG. 13 and FIG. 14, the combination of the color extraction ranges held in the minimum color extraction range combination 510 in (select the minimum color extraction range combination: 509) is the color extraction range 1301 in the case of FIG. And 1302.

【００５４】図１５について説明する。被分割色分布抽
出範囲５０２には、まだ被分割色抽出範囲５０４として
選択されてない色抽出範囲が残されている。被分割色分
布抽出範囲５０２を構成するもう一つの色抽出範囲につ
いても、図１３及び図１４と同様に２つの分割位置で２
つに分割し、２つの分割範囲５０７−１及び５０７−２
に含まれる部分のみの抽出分布１０６に外接する色抽出
範囲を求め、最小となる組み合わせを選択すれば、生成
色分布抽出範囲５１３は、図１５に示した５１３−１か
ら５１３−４の４つの色抽出範囲の組み合わせとなる。
このとき、生成色分布抽出範囲５１３を構成する色抽出
範囲の総数は４である。 ここでは、色分布抽出範囲１
１１は４つの色抽出範囲で抽出分布１１１を近似すると
しているので、生成色分布抽出範囲５１３は色抽出範囲
の数をもう増やす事ができない。従って、（まだ色抽出
範囲を増やせる？：５１５）では、（生成色分布抽出範
囲を色分布抽出範囲とする：５１７）で生成色分布抽出
範囲５１３の内容を色分布抽出範囲１１１に複写して、
処理を終了する。FIG. 15 will be described. In the divided color distribution extraction range 502, there remains a color extraction range that has not been selected as the divided color extraction range 504. Similarly to FIGS. 13 and 14, another color extraction range forming the divided color distribution extraction range 502 is 2 at two division positions.
Divided into two, and two divided ranges 507-1 and 507-2
If the color extraction range circumscribing the extraction distribution 106 of only the portion included in the area is obtained and the smallest combination is selected, the generated color distribution extraction range 513 is the four of 513-1 to 513-4 shown in FIG. It is a combination of color extraction ranges.
At this time, the total number of color extraction ranges forming the generated color distribution extraction range 513 is four. Here, the color distribution extraction range 1
Since 11 indicates that the extraction distribution 111 is approximated by four color extraction ranges, the generated color distribution extraction range 513 cannot further increase the number of color extraction ranges. Therefore, in (Can the color extraction range be increased ?: 515), the contents of the generated color distribution extraction range 513 are copied to the color distribution extraction range 111 in (Set generated color distribution extraction range as color distribution extraction range: 517). ,
The process ends.

【００５５】＜実施例２＞図１は、本発明による画像処
理方法の実施例２における処理の構成を示す流れ図であ
る。処理対象１０２の有する色分布、あるいは処理対象
の画像を含んでいるカラー画像を得て参照画像１０１と
し、参照画像１０１の処理対象の有する色分布を含む所
望の領域である抽出領域１０３から色分布取得処理１０
５によって得た色分布を抽出分布１０６とする。また、
参照画像１０１のうち処理対象１０２ではない背景の部
分など、色空間から抽出してはならない部分の色分布を
含む所望の領域である阻止領域１０４から色分布取得処
理１０５によって得た色分布を阻止分布１０７とする。
色分布抽出範囲取得処理１０８によって抽出分布１０６
の輪郭を十分に多数のの色抽出範囲で近似した色分布抽
出範囲である抽出分布近似範囲１０９を得る。近似範囲
合成処理１１０は、抽出分布近似範囲１０９を構成する
色抽出範囲から選択した複数の色抽出範囲について、阻
止分布１０７に掛からなければそれを１つの色抽出範囲
に合成した色抽出範囲に置き換える事で、色空間から阻
止分布１０７を除いた残りの色空間の範囲内で抽出分布
１０６を内包しながら、抽出分布近似範囲１０９を構成
する色抽出範囲の総数を減じる。抽出分布近似範囲１０
９を構成する色抽出範囲の総数を、色分布抽出処理１１
３で用いる色分布抽出範囲１１１を構成する色抽出範囲
の総数まで減じれば、抽出分布近似範囲１０９を色分布
抽出範囲１１１とする。色分布抽出処理１１３では、処
理画像１１２のうち、色分布抽出範囲１１１を構成する
複数の色抽出範囲のどれかに含まれる色である部分を抽
出する事で、処理対象である部分のみを抽出した画像デ
ータである抽出画像１１４を得る。抽出領域１０３より
色分布取得処理１０５によって抽出分布１０６を得る過
程は、図３を用いた色分布取得処理１０５の実施例の説
明で述べられており、本実施例においても同じである。<Embodiment 2> FIG. 1 is a flow chart showing the configuration of processing in Embodiment 2 of the image processing method according to the present invention. The color distribution of the processing target 102 or the color image including the processing target image is obtained as the reference image 101, and the color distribution is extracted from the extraction region 103 that is a desired region including the color distribution of the processing target of the reference image 101. Acquisition process 10
The color distribution obtained in 5 is set as the extraction distribution 106. Also,
The color distribution obtained by the color distribution acquisition processing 105 is blocked from the blocking area 104 that is a desired area including the color distribution of the portion that should not be extracted from the color space, such as the background portion that is not the processing target 102 in the reference image 101. The distribution is 107.
The extraction distribution 106 is acquired by the color distribution extraction range acquisition processing 108.
An extraction distribution approximation range 109, which is a color distribution extraction range, is obtained by approximating the contour of the above with a sufficiently large number of color extraction ranges. The approximate range synthesis processing 110 replaces a plurality of color extraction ranges selected from the color extraction ranges forming the extraction distribution approximate range 109 with a color extraction range combined into one color extraction range if the inhibition distribution 107 is not reached. By doing so, the total number of color extraction ranges forming the extraction distribution approximation range 109 is reduced while including the extraction distribution 106 within the range of the remaining color space excluding the inhibition distribution 107 from the color space. Extraction distribution approximation range 10
The color distribution extraction processing 11
When the total number of color extraction ranges forming the color distribution extraction range 111 used in 3 is reduced, the extraction distribution approximation range 109 is set as the color distribution extraction range 111. In the color distribution extraction processing 113, by extracting a portion of the processed image 112 that is a color included in any of the plurality of color extraction areas that form the color distribution extraction area 111, only the portion that is the processing target is extracted. An extracted image 114, which is the image data of the extracted image, is obtained. The process of obtaining the extraction distribution 106 from the extraction region 103 by the color distribution acquisition process 105 has been described in the description of the embodiment of the color distribution acquisition process 105 using FIG. 3, and is the same in this embodiment.

【００５６】阻止領域１０４から色分布取得処理１０５
によって阻止分布１０７を得る過程は、（抽出領域１０
３より色分布取得処理１０５によって抽出分布１０６を
得る過程）に対して、色分布を取得する領域を阻止領域
１０４とし、得られた色分布を阻止分布１０４とした色
分布取得処理１０５による色分布取得処理に他ならな
い。従って、図３を用いた色分布取得処理１０５の実施
例の説明において、抽出領域１０３を阻止領域１０４と
読み換え、抽出分布１０６を阻止分布１０７と読み換え
れば阻止分布１０７を得る過程を説明できる。色分布抽
出範囲取得処理１０８によって抽出分布１０６の輪郭を
十分に多数の色抽出範囲で近似した抽出分布近似範囲１
０９を得る過程は、（抽出分布１０６から色分布抽出範
囲取得処理１０８によって色分布抽出範囲１１１を得る
過程）に対して、処理結果として得られる色分布抽出範
囲を色分布抽出処理で用いる色分布抽出範囲１１１に換
えて（抽出分布１０６の輪郭を十分に多数の色抽出範囲
で近似した抽出分布近似範囲１０９）とした色分布抽出
範囲取得処理１０８による色分布抽出範囲取得処理に他
ならない。Color distribution acquisition processing 105 from the blocking area 104
The process of obtaining the rejection distribution 107 by
3 is a process of obtaining the extraction distribution 106 by the color distribution acquisition process 105), the region for which the color distribution is acquired is the blocking region 104, and the obtained color distribution is the blocking distribution 104. It is nothing but an acquisition process. Therefore, in the description of the embodiment of the color distribution acquisition processing 105 using FIG. 3, the process of obtaining the inhibition distribution 107 by reading the extraction region 103 as the inhibition region 104 and the extraction distribution 106 as the inhibition distribution 107 can be described. . Extraction distribution approximation range 1 in which the contour of the extraction distribution 106 is approximated by a sufficiently large number of color extraction ranges by the color distribution extraction range acquisition processing 108
In the process of obtaining 09, the color distribution extraction range obtained as a processing result is used for the color distribution extraction process with respect to (the process of obtaining the color distribution extraction range 111 from the extraction distribution 106 by the color distribution extraction range acquisition process 108). It is nothing but the color distribution extraction range acquisition process by the color distribution extraction range acquisition process 108 in which the extraction range 111 is replaced (the extraction distribution approximation range 109 in which the contour of the extraction distribution 106 is approximated by a sufficiently large number of color extraction ranges).

【００５７】従って、図５及び図６を用いた色分布抽出
範囲取得処理１０８の実施例の説明において、色分布抽
出範囲１１１を抽出分布近似範囲１０９と読み換え、
（処理画像１１２から色分布抽出処理１１３によって処
理対象を抽出するとき用いる色抽出範囲の数ｋ個）を
（処理画像１１２から色分布抽出処理１１３によって処
理対象を抽出するとき用いる色抽出範囲の数に比べて十
分大きい数ｋ個）と読み換えれば、抽出分布１０６から
色分布抽出範囲取得処理１０８によって抽出分布近似範
囲１０９を得る過程を説明できる。また図９から図１５
を用いて説明されている色分布抽出範囲取得処理１０８
の実施例による色分布抽出範囲取得処理の過程も同様に
色分布抽出範囲１１１を抽出分布近似範囲１０９と読み
換えれば、抽出分布１０６から色分布抽出範囲取得処理
１０８によって抽出分布近似範囲１０９を得る過程を説
明できる。Therefore, in the description of the embodiment of the color distribution extraction range acquisition processing 108 using FIGS. 5 and 6, the color distribution extraction range 111 is read as the extraction distribution approximation range 109,
(The number k of color extraction ranges used when extracting a processing target from the processed image 112 by the color distribution extraction process 113) is calculated as (the number of color extraction ranges used when extracting a processing target from the processed image 112 by the color distribution extraction process 113). If it is read as a number k which is sufficiently larger than the above, the process of obtaining the extraction distribution approximate range 109 from the extraction distribution 106 by the color distribution extraction range acquisition processing 108 can be explained. 9 to 15
Color distribution extraction range acquisition processing 108 described using FIG.
Similarly, in the process of color distribution extraction range acquisition processing according to this embodiment, if the color distribution extraction range 111 is read as the extraction distribution approximation range 109, the extraction distribution approximation range 109 is obtained from the extraction distribution 106 by the color distribution extraction range acquisition processing 108. Explain the process.

【００５８】色分布抽出処理１１３では、処理画像１１
２のうち、色分布抽出範囲１１１を構成する複数の色抽
出範囲のどれかに含まれる色である部分を抽出する事
で、処理対象である部分のみを抽出した画像データであ
る抽出画像１１４を得る処理の過程は、図７及び図８を
用いた色分布抽出処理１１３の実施例の説明で述べられ
ており、本実施例においても同じである。In the color distribution extraction processing 113, the processed image 11
By extracting a part that is a color included in any of a plurality of color extraction ranges that configure the color distribution extraction range 111 out of 2, an extracted image 114 that is image data obtained by extracting only a part to be processed is obtained. The process of the process to be obtained is described in the description of the embodiment of the color distribution extraction process 113 using FIGS. 7 and 8, and is the same in this embodiment.

【００５９】図１６及び図１７は、本発明による画像処
理方法の実施例２における近似範囲合成処理１１０の処
理内容を説明するための流れ図である。近似範囲合成処
理１１０は、抽出分布近似範囲１０９を構成する色抽出
範囲から選択した複数の色抽出範囲を被合成色抽出範囲
１６０２とし、被合成色抽出範囲１６０２を全て内包す
る合成色抽出範囲１６０４を得る。合成色抽出範囲１６
０４が阻止分布１０７に掛からなければ被合成色抽出範
囲１６０２を合成色抽出範囲１６０４に置き換える事
で、色空間から阻止分布１０７を除いた残りの色空間の
範囲内で抽出分布１０６を内包しながら、抽出分布近似
範囲１０９を構成する色抽出範囲の総数を減じる。抽出
分布近似範囲１０９を構成する色抽出範囲の総数を、色
分布抽出処理１１３で用いる色分布抽出範囲１１１を構
成する色抽出範囲の総数まで減じれば、抽出分布近似範
囲１０９を色分布抽出範囲１１１として処理を終了す
る。16 and 17 are flow charts for explaining the processing contents of the approximate range synthesis processing 110 in the second embodiment of the image processing method according to the present invention. In the approximate range synthesizing process 110, a plurality of color extraction ranges selected from the color extraction ranges forming the extraction distribution approximate range 109 are set as the combined color extraction range 1602, and the combined color extraction range 1604 including all the combined color extraction ranges 1602. To get Composite color extraction range 16
If 04 does not fall on the inhibition distribution 107, the composite color extraction range 1602 is replaced with the composite color extraction range 1604, so that the extraction distribution 106 is included within the range of the remaining color space excluding the inhibition distribution 107 from the color space. , The total number of color extraction ranges forming the extraction distribution approximation range 109 is reduced. If the total number of color extraction ranges forming the extraction distribution approximation range 109 is reduced to the total number of color extraction ranges forming the color distribution extraction range 111 used in the color distribution extraction processing 113, the extraction distribution approximation range 109 is converted into the color distribution extraction range. The processing ends as 111.

【００６０】（次の色抽出範囲を選択：１６０１）本実
施例では被合成色抽出範囲１６０２として１度の処理で
選択して取り出す色抽出範囲の数をｐ個とする。（次の
色抽出範囲を選択：１６０１）は抽出分布近似範囲１０
９を構成する色抽出範囲から互いに近接するｐ個の色抽
出範囲を選択して取り出し被合成色抽出範囲１６０２−
１〜１６０２−ｐとする。(Select next color extraction range: 1601) In this embodiment, the number of color extraction ranges to be selected and extracted in one process as the combined color extraction range 1602 is p. (Select next color extraction range: 1601) is the extraction distribution approximation range 10
9. The p color extraction ranges that are close to each other are selected and extracted from the color extraction ranges that form 9 and the combined color extraction range 1602-
1 to 1602-p.

【００６１】（色抽出範囲の合成：１６０３）近似範囲
合成処理１１０では抽出分布近似範囲１０９をから近接
する複数の色抽出範囲を選択し、それを１つの色抽出範
囲に合成したものに置き換える事で抽出分布近似範囲１
０９を構成する色抽出範囲の総数を減らす処理を行う。
（色抽出範囲の合成：１６０３）は、ｐ個の被合成色抽
出範囲１６０２−１〜１６０２−ｐのそれぞれの色抽出
範囲の全てを内包する１個の色抽出範囲である合成色抽
出範囲１６０４を得る処理である。ｐ個の被合成色抽出
範囲１６０２−１〜１６０２−ｐのそれぞれの色抽出範
囲の全てを内包する１個の色抽出範囲を得るには、被合
成色抽出範囲１６０２−１〜１６０２−ｐの要素を（Ｒ
１ｕ，Ｒ１ｄ，Ｇ１ｕ，Ｇ１ｄ，Ｂ１ｕ，Ｂ１ｄ）〜
（Ｒｐｕ，Ｒｐｄ，Ｇｐｕ，Ｇｐｄ，Ｂｐｕ，Ｂｐｄ）
で表し、合成色抽出範囲１６０４の要素を（Ｒｕ，Ｒ
ｄ，Ｇｕ，Ｇｄ，Ｂｕ，Ｂｄ）で表せば、ＲｕはＲ１ｕ
〜Ｒｐｕの最大値、ＲｄはＲ１ｄ〜Ｒｐｄの最小値、Ｇ
ｕはＧ１ｕ〜Ｇｐｕの最大値、ＧｄはＧ１ｄ〜Ｇｐｄの
最小値、ＢｕはＢ１ｕ〜Ｂｐｕの最大値、ＢｄはＢ１ｄ
〜Ｂｐｄの最小値である。(Combination of Color Extraction Range: 1603) In the approximation range synthesis process 110, a plurality of adjacent color extraction ranges are selected from the extraction distribution approximation range 109, and they are replaced with one that is combined into one color extraction range. Extracted distribution approximation range 1
Processing for reducing the total number of color extraction ranges that make up 09 is performed.
(Composition of color extraction ranges: 1603) is a composite color extraction range 1604 which is one color extraction range including all of the respective color extraction ranges of the p combined color extraction ranges 1602-1 to 1602-p. Is a process to obtain. In order to obtain one color extraction range that includes all the respective color extraction ranges of the p combined color extraction ranges 1602-1 to 1602-p, the combined color extraction ranges 1602-1 to 1602-p Element (R
1u, R1d, G1u, G1d, B1u, B1d) to
(Rpu, Rpd, Gpu, Gpd, Bpu, Bpd)
The elements of the composite color extraction range 1604 are represented by (Ru, R
d, Gu, Gd, Bu, Bd), Ru is R1u
~ Rpu maximum value, Rd is R1d ~ Rpd minimum value, G
u is the maximum value of G1u to Gpu, Gd is the minimum value of G1d to Gpd, Bu is the maximum value of B1u to Bpu, and Bd is B1d.
Is the minimum value of Bpd.

【００６２】（合成色抽出範囲は阻止分布に掛かる？：
１６０５） （被合成色抽出範囲を追加：１６０６） （合成色抽出範囲を追加：１６０７）近似範囲合成処理
１１０では合成色抽出範囲１６０４が阻止分布１０７に
掛からなければそれを被合成色抽出範囲１６０２と置き
換える処理を行う。本実施例では両者を直接置き換える
処理の替わりに、合成近似範囲１６０８という抽出分布
近似範囲１０９と同型のデータを用意して、合成色抽出
範囲１６０４が阻止分布１０７に掛かっていれば、（被
合成色抽出範囲を追加：１６０６）で被合成色抽出範囲
１６０２を合成近似範囲１６０８の要素として追加し、
合成色抽出範囲１６０４が阻止分布１０７に掛かってい
なければ、（合成色抽出範囲を追加：１６０７）で合成
色抽出範囲１６０４を合成近似範囲１６０８の要素とし
て追加する。抽出分布近似範囲１０９を構成するすべて
の色抽出範囲を被合成色抽出範囲として処理を終えれ
ば、合成色抽出範囲１６０４が阻止分布１０７に掛から
なければそれを被合成色抽出範囲１６０２と置き換えた
抽出分布近似範囲が合成近似範囲１６０８にできあが
る。(Does the composite color extraction range cover the rejection distribution ?:
1605) (Add a combined color extraction range: 1606) (Add a combined color extraction range: 1607) In the approximate range combining process 110, if the combined color extraction range 1604 does not fall on the inhibition distribution 107, it is set as the combined color extraction range 1602. Replace with. In the present embodiment, instead of the process of directly replacing the two, data of the same type as the extraction distribution approximation range 109, which is the synthesis approximation range 1608, is prepared. Add a color extraction range: 1606) to add a combined color extraction range 1602 as an element of a synthetic approximation range 1608,
If the composite color extraction range 1604 does not overlap the inhibition distribution 107, the composite color extraction range 1604 is added as an element of the composite approximate range 1608 in (added composite color extraction range: 1607). When all the color extraction ranges forming the approximate extraction distribution range 109 are set as the combined color extraction range, if the combined color extraction range 1604 does not overlap the inhibition distribution 107, it is replaced with the combined color extraction range 1602. The extraction distribution approximation range is created as the synthesis approximation range 1608.

【００６３】合成色抽出範囲１６０４が阻止分布１０７
に掛かるかどうかは、阻止分布１０７を構成する色情報
が１つでも合成色抽出範囲の範囲内に入るか否かで判断
する。合成色抽出範囲１６０４を（Ｒｄ，Ｒｕ，Ｇｄ，
Ｇｕ，Ｂｄ，Ｂｕ）、阻止分布１０７を構成するｉ番目
の色情報を（Ｒｉ，Ｇｉ，Ｂｉ）とすれば、（Ｒｄ≦Ｒ
ｉ≦ＲｕかつＧｄ≦Ｇｉ≦ＧｕかつＢｄ≦Ｂｉ≦Ｂｕ）
である色情報が１つでもあれば合成色抽出範囲１６０４
は阻止分布に掛かっていると判断し、（Ｒｄ≦Ｒｉ≦Ｒ
ｕかつＧｄ≦Ｇｉ≦ＧｕかつＢｄ≦Ｂｉ≦Ｂｕ）である
色情報が１つも無ければ合成色抽出範囲１６０４は阻止
分布に掛かっていないと判断される。The composite color extraction range 1604 is the inhibition distribution 107.
Whether or not any of the color information forming the inhibition distribution 107 falls within the composite color extraction range is determined. The composite color extraction range 1604 is set to (Rd, Ru, Gd,
Gu, Bd, Bu) and the i-th color information forming the rejection distribution 107 is (Ri, Gi, Bi), (Rd ≦ R
i ≦ Ru and Gd ≦ Gi ≦ Gu and Bd ≦ Bi ≦ Bu)
If there is even one piece of color information, the composite color extraction range 1604
Is determined to be on the rejection distribution, and (Rd ≦ Ri ≦ R
If there is no color information that is u and Gd ≦ Gi ≦ Gu and Bd ≦ Bi ≦ Bu), it is determined that the composite color extraction range 1604 does not fall on the inhibition distribution.

【００６４】（全ての色抽出範囲を選択？：１６０９）
近似範囲合成処理１１０では、合成色抽出範囲１６０４
が阻止分布１０７に掛からなければそれを被合成色抽出
範囲１６０２と置き換える処理を行う。本実施例では、
抽出分布近似範囲１０９を構成する全ての色抽出範囲に
ついて被合成色抽出範囲１６０２として選択し処理を行
う事で、合成色抽出範囲１６０４が阻止分布１０７に掛
からなければそれを被合成色抽出範囲１６０２と置き換
えた抽出分布近似範囲が合成近似範囲１６０８にできあ
がる。従って、全ての色抽出範囲を被合成色抽出範囲１
６０２として選択し処理を行うまでは、（次の色抽出範
囲を選択：１６０１）に戻って処理を繰り返す。(Select all color extraction ranges ?: 1609)
In the approximate range synthesis processing 110, the synthetic color extraction range 1604
Does not fall on the inhibition distribution 107, it is replaced with the composite color extraction range 1602. In this embodiment,
By selecting all the color extraction ranges constituting the extraction distribution approximate range 109 as the combined color extraction range 1602 and performing the processing, if the combined color extraction range 1604 does not fall on the inhibition distribution 107, it is set as the combined color extraction range 1602. The extracted distribution approximate range replaced by is formed into a synthetic approximate range 1608. Therefore, all color extraction ranges are set to the combined color extraction range 1
Until it is selected as 602 and the process is performed, the process returns to (select next color extraction range: 1601) and repeats the process.

【００６５】（まだ色抽出範囲数が多い：１６１０） （合成近似範囲を抽出分布近似範囲とする：１６１１） （合成近似範囲を色分布抽出範囲とする：１６１２）近
似範囲合成処理１１０では、合成色抽出範囲１６０４が
阻止分布１０７に掛からなければそれを被合成色抽出範
囲１６０２と置き換える処理を行う。本実施例では、抽
出分布近似範囲１０９を構成する全ての色抽出範囲につ
いて被合成色抽出範囲１６０２として選択し処理を行う
事で、合成色抽出範囲１６０４が阻止分布１０７に掛か
らなければそれを被合成色抽出範囲１６０２と置き換え
た抽出分布近似範囲が合成近似範囲１６０８にできあが
る。色分布抽出処理１１３で用いる色分布抽出範囲１１
１を構成する色抽出範囲の総数まで、合成近似範囲１６
０８を構成する色抽出範囲の総数が減じられていれば
（合成近似範囲を色分布抽出範囲とする：１６１２）で
合成近似範囲１６０８を構成する色抽出範囲の内容を色
分布抽出範囲１１１の要素として複写して近似範囲合成
処理１１０を終了する。色分布抽出処理１１３で用いる
色分布抽出範囲１１１を構成する色抽出範囲の総数よ
り、合成近似範囲１６０８を構成する色抽出範囲の総数
が多い間は（合成近似範囲を抽出分布近似範囲とする：
１６１１）で合成近似範囲１６０８の内容を抽出分布近
似範囲１０９に複写し、（次の色抽出範囲を選択：１６
０１）に戻って処理を繰り返す。(The number of color extraction ranges is still large: 1610) (The composite approximation range is the extraction distribution approximation range: 1611) (The synthesis approximation range is the color distribution extraction range: 1612) If the color extraction range 1604 does not overlap the inhibition distribution 107, a process of replacing it with the combined color extraction range 1602 is performed. In this embodiment, all the color extraction ranges forming the approximate extraction distribution range 109 are selected as the combined color extraction range 1602 and processed, and if the combined color extraction range 1604 does not fall on the inhibition distribution 107, it is covered. An extraction distribution approximation range replaced with the synthesis color extraction range 1602 is created as a synthesis approximation range 1608. Color distribution extraction range 11 used in color distribution extraction processing 113
Up to the total number of color extraction ranges that make up 1
If the total number of color extraction ranges that make up 08 is reduced (the composite approximation range is the color distribution extraction range: 1612), the contents of the color extraction ranges that make up the synthesis approximate range 1608 are the elements of the color distribution extraction range 111. Then, the approximate range synthesizing process 110 is ended. While the total number of color extraction ranges forming the composite approximation range 1608 is larger than the total number of color extraction ranges forming the color distribution extraction range 111 used in the color distribution extraction process 113 (the composite approximation range is defined as an extraction distribution approximation range:
1611), copy the contents of the synthetic approximation range 1608 to the extraction distribution approximation range 109, and select (select the next color extraction range: 16
Return to 01) and repeat the process.

【００６６】図１８から図２１は上述の近似範囲合成処
理１１０の実施例による抽出分布近似範囲１０９から色
分布抽出範囲１１１を得る処理の過程を説明するための
説明図である。実際の色分布は色情報が３つの要素で表
現されるので３次元立体的な形状であるが、ここでは、
説明を簡単にするために平面的な形状であるとする。ま
た、説明を簡単にするために、（次の色抽出範囲を選
択：１６０１）で選択される被合成色抽出範囲１６０２
の数を２とし、抽出分布近似範囲１０９を構成する色抽
出範囲の数を８、色分布抽出範囲１１１を構成する色抽
出範囲の数を４とする。18 to 21 are explanatory diagrams for explaining the process of obtaining the color distribution extraction range 111 from the extraction distribution approximation range 109 according to the embodiment of the above-described approximation range synthesis process 110. The actual color distribution has a three-dimensional shape because color information is represented by three elements, but here,
For simplicity of explanation, it is assumed that the shape is planar. In order to simplify the description, the combined color extraction range 1602 selected in (select next color extraction range: 1601)
Is 2, the number of color extraction ranges forming the extraction distribution approximation range 109 is 8, and the number of color extraction ranges forming the color distribution extraction range 111 is 4.

【００６７】図１８について説明する。阻止分布１０７
は２つの分布１０７−１と１０７−２に分かれており、
それに挟まれる形で８つの色抽出範囲から構成される近
似分布範囲１０９−１〜１０９−８がある。１度の処理
で１組の被合成色抽出範囲１６０２−１と１６０２−２
が選択され、合成色抽出範囲１６０４が生成される。抽
出分布近似範囲１０９−１〜１０９−８より近接する２
つの色抽出範囲を順次選択し、合成色抽出範囲を生成す
れば、４つの合成色抽出範囲１６０４−１〜１６０４−
４が得られる。FIG. 18 will be described. Stop distribution 107
Is divided into two distributions 107-1 and 107-2,
There are approximate distribution ranges 109-1 to 109-8 composed of eight color extraction ranges sandwiched between them. A set of combined color extraction ranges 1602-1 and 1602-2 can be processed once.
Is selected, and a composite color extraction range 1604 is generated. Extraction distribution Approximate range 109-1 to 109-8 closer to 2
If one color extraction range is sequentially selected and a combined color extraction range is generated, four combined color extraction ranges 1604-1 to 1604-
4 is obtained.

【００６８】図１９について説明する。合成色抽出範囲
１６０４が阻止分布１０７に掛からなければ合成色抽出
範囲１６０４を合成近似範囲１６０８に追加し、合成色
抽出範囲１６０４が阻止分布１０７に掛かっていれば被
合成色抽出範囲１６０２を合成近似範囲１６０８に追加
する。図１８において、合成色抽出範囲１６０４−１と
１６０４−４は阻止分布１０７に掛からないが、合成色
抽出範囲１６０４−２は阻止分布１０７−１に掛かり、
合成色抽出範囲１６０４−３は阻止分布１０７−２に掛
かる。従って、合成近似範囲１６０８には、合成色抽出
範囲１６０４−１と１６０４−２及び、被合成色抽出範
囲として選択された１０９−３、１０９−４、１０９−
５、１０９−６との６つの色抽出範囲が追加され、結果
として、合成近似範囲１６０８−１〜１６０８−６が得
られる。色分布抽出範囲１１１を構成する色抽出範囲の
総数は４である。今、合成近似範囲１６０８を構成する
色抽出範囲の総数は６であり、まだ多いので合成近似範
囲１６０８の内容を抽出分布近似範囲１０９に複写して
（次の色抽出範囲を選択：１６０１）に戻って処理を繰
り返す。FIG. 19 will be described. If the composite color extraction range 1604 does not fall on the inhibition distribution 107, the composite color extraction range 1604 is added to the synthesis approximation range 1608. If the composite color extraction range 1604 falls on the inhibition distribution 107, the composite color extraction range 1602 is synthesized approximation. Add to range 1608. In FIG. 18, the composite color extraction ranges 1604-1 and 1604-4 do not cover the stop distribution 107, but the composite color extraction range 1604-2 covers the stop distribution 107-1.
The composite color extraction range 1604-3 covers the inhibition distribution 107-2. Therefore, in the synthetic approximation range 1608, the synthetic color extraction ranges 1604-1 and 1604-2, and 109-3, 109-4, and 109- selected as the combined color extraction range.
Six color extraction ranges of 5 and 109-6 are added, and as a result, synthetic approximate ranges 1608-1 to 1608-6 are obtained. The total number of color extraction ranges forming the color distribution extraction range 111 is four. Now, the total number of color extraction ranges forming the synthetic approximate range 1608 is 6, and since there are still many, the contents of the synthetic approximate range 1608 are copied to the extracted distribution approximate range 109 (select the next color extraction range: 1601). Return and repeat the process.

【００６９】図２０について説明する。抽出分布近似範
囲１０９−１〜１０９−６より近接する２つの色抽出範
囲を順次選択し、合成色抽出範囲を生成すれば、３つの
合成色抽出範囲１６０４−１〜１６０４−３が得られ
る。FIG. 20 will be described. If two color extraction ranges closer to the extraction distribution approximation range 109-1 to 109-6 are sequentially selected to generate a composite color extraction range, three composite color extraction ranges 1604-1 to 1604-3 are obtained.

【００７０】図２１について説明する。図２０におい
て、合成色抽出範囲１６０４−１と１６０４−２は阻止
分布１０７に掛からないが、合成色抽出範囲１６０４−
３は阻止分布１０７−２に掛かる。従って、合成近似範
囲１６０８には、合成色抽出範囲１６０４−１と１６０
４−２及び、被合成色抽出範囲として選択された１０９
−５、１０９−６との４つの色抽出範囲が追加され、結
果として、合成近似範囲１６０８−１〜１６０８−４が
得られる。色分布抽出範囲１１１を構成する色抽出範囲
の総数は４である。今、合成近似範囲１６０８を構成す
る色抽出範囲の総数は４であり等しくなったので、合成
近似範囲１６０８の内容を色分布抽出範囲１１１に複写
して近似範囲合成処理１１０を終了する。得られた色分
布抽出範囲１１１は、図２１の合成近似範囲を構成する
４つの色抽出範囲１６０８−１〜１６０８−４をそのま
ま複写したものである。阻止分布から遠い部分は大まか
に近似され大きな色抽出範囲となるが、接近している部
分は小さい色抽出範囲で細かく近似される。また、図２
１において、合成近似範囲１６０８を構成する色抽出範
囲の内、２つの色抽出範囲１６０８−２と１６０８−３
が重なりあっているが、これは色分布抽出処理１１３で
の抽出処理において問題にはならない。色分布抽出処理
１１３では、色分布抽出範囲１１１を構成する色抽出範
囲の内、どれか１つの色抽出範囲に含まれる色である部
分を抽出するので、ある色である部分が複数の色抽出範
囲に含まれる事は、単にその色の部分が抽出される事に
他ならず、処理上の障害にはならない。FIG. 21 will be described. In FIG. 20, although the composite color extraction ranges 1604-1 and 1604-2 do not overlap the inhibition distribution 107, the composite color extraction range 1604-
3 is on the stop distribution 107-2. Therefore, the synthetic approximation range 1608 includes the synthetic color extraction ranges 1604-1 and 1604-1.
4-2 and 109 selected as the combined color extraction range
Four color extraction ranges -5 and 109-6 are added, and as a result, synthetic approximate ranges 1608-1 to 1608-4 are obtained. The total number of color extraction ranges forming the color distribution extraction range 111 is four. Since the total number of color extraction ranges forming the synthetic approximate range 1608 is 4, which is equal, the content of the synthetic approximate range 1608 is copied to the color distribution extraction range 111, and the approximate range synthesizing process 110 ends. The obtained color distribution extraction range 111 is a copy of the four color extraction ranges 1608-1 to 1608-4 forming the synthetic approximation range of FIG. 21 as they are. The part far from the rejection distribution is roughly approximated to have a large color extraction range, while the close part is finely approximated with a small color extraction range. Also, FIG.
1, two color extraction ranges 1608-2 and 1608-3 out of the color extraction ranges forming the synthetic approximate range 1608.
Overlap, but this does not pose a problem in the extraction processing in the color distribution extraction processing 113. In the color distribution extraction processing 113, a part that is a color included in any one of the color extraction ranges that form the color distribution extraction range 111 is extracted. Therefore, a part that is a certain color is extracted by a plurality of colors. What is included in the range is nothing but the extraction of the part of the color, and does not hinder the processing.

【００７１】[0071]

【発明の効果】以上のように、本発明による画像処理に
よれば、抽出分布の輪郭を複数の色抽出範囲で近似した
色分布抽出範囲を得る処理をして、前記色分布抽出範囲
により処理対象の画像を処理画像より抽出する。したが
って、処理対象の持つ色分布である抽出分布の輪郭を複
数の色抽出範囲の組み合わせで近似した色分布抽出範囲
を得て、処理画像のうち、色分布抽出範囲を構成する複
数の色抽出範囲のどれかに含まれる色である部分を抽出
する事で処理対象を抽出するので、明るさムラや色ムラ
などで抽出分布が広い範囲に広がっていても抽出分布に
含まれる色である領域を抽出できるとともに、従来の単
一の色抽出範囲で処理対象を抽出する画像処理方法のよ
うに抽出分布に含まれない色の領域を多量に抽出してし
まう事はない。As described above, according to the image processing of the present invention, a process for obtaining a color distribution extraction range in which the contour of the extraction distribution is approximated by a plurality of color extraction ranges is performed, and the color distribution extraction range is processed. The target image is extracted from the processed image. Therefore, the contour of the extraction distribution, which is the color distribution of the processing target, is approximated by a combination of a plurality of color extraction ranges to obtain a color distribution extraction range, and the plurality of color extraction ranges forming the color distribution extraction range in the processed image are obtained. Since the processing target is extracted by extracting the part that is a color included in any of the above, even if the extraction distribution is spread over a wide range due to uneven brightness or color unevenness, the area that is the color included in the extraction distribution is In addition to being able to perform extraction, a large amount of color regions not included in the extraction distribution, unlike the conventional image processing method of extracting a processing target with a single color extraction range, is not extracted.

【００７２】さらに、本発明による画像処理方法は、色
分布抽出範囲を構成する色抽出範囲を被分割色抽出範囲
として、前記被分割色抽出範囲を、任意の分割位置で複
数の色抽出範囲に分割して分割範囲として、分割範囲で
抽出される、抽出分布のなかの、色分布に外接する色抽
出範囲を外接範囲として、外接範囲の大きさの総和を求
めて、さらに複数の、他の任意の分割位置に関しても、
外接範囲の大きさの総和を求めて、前記、複数の外接範
囲の大きさの総和のなかで、最小の大きさの総和である
外接範囲を被分割色抽出範囲と置き換えることにより、
抽出分布の輪郭を複数の色抽出範囲で近似する。したが
って、限られた数の色抽出範囲で、精度良く抽出分布の
輪郭を近似できるので、より精度良く処理対象を抽出で
きる。Further, in the image processing method according to the present invention, the color extraction range forming the color distribution extraction range is set as the divided color extraction range, and the divided color extraction range is set to a plurality of color extraction ranges at arbitrary division positions. Divide the divided range, extract the divided range, and use the color extraction range circumscribing the color distribution in the extracted distribution as the circumscribing range to find the sum of the sizes of the circumscribing ranges. For any division position,
Obtaining the sum of the sizes of the circumscribing ranges, among the sum of the sizes of the plurality of circumscribing ranges, by replacing the circumscribing range, which is the sum of the smallest sizes, with the divided color extraction range,
The contour of the extraction distribution is approximated by a plurality of color extraction ranges. Therefore, the contour of the extraction distribution can be accurately approximated with a limited number of color extraction ranges, and the processing target can be more accurately extracted.

【００７３】また、本発明による画像処理方法は、参照
画像より処理対象の有する色分布を含まない、抽出して
はならない部分の所望の領域である、阻止領域内のカラ
ー画像データから色情報を読み取る事によって得られた
色分布を阻止分布として、抽出分布及び阻止分布の属す
る色情報の表現範囲である色空間から阻止分布を除いた
残りの前記色空間内の領域で、処理対象の画像を処理画
像より抽出する。したがって、処理対象の有する色分布
を含まず抽出してはならない部分の有する色分布である
阻止分布を除いた残りの色空間の範囲内で、処理対象の
持つ色分布である抽出分布の輪郭を複数の色抽出範囲の
組み合わせで近似した色分布抽出範囲を得て、処理画像
のうち、色分布抽出範囲を構成する複数の色抽出範囲の
どれかに含まれる色である部分を抽出する事で処理対象
を抽出するので、抽出分布と阻止分布とが接近していて
も阻止分布に含まれる色である領域を抽出せずに、処理
対象を抽出できる。In the image processing method according to the present invention, the color information is obtained from the color image data in the inhibition area, which is a desired area of the portion which should not be extracted and which does not include the color distribution of the processing object from the reference image. With the color distribution obtained by reading as the stop distribution, the image to be processed is displayed in the remaining area in the color space excluding the stop distribution from the color space that is the expression range of the color information to which the extraction distribution and the stop distribution belong. Extract from the processed image. Therefore, within the range of the remaining color space excluding the stop distribution, which is the color distribution of the portion that does not include the color distribution of the processing target and that should not be extracted, the contour of the extraction distribution that is the color distribution of the processing target is By obtaining a color distribution extraction range approximated by a combination of a plurality of color extraction ranges, and extracting a portion of the processed image that is a color included in any of the plurality of color extraction ranges that form the color distribution extraction range. Since the processing target is extracted, even if the extraction distribution and the blocking distribution are close to each other, the processing target can be extracted without extracting the region having the color included in the blocking distribution.

【００７４】さらにまた、本発明による画像処理方法
は、参照画像より処理対象の有する色分布を含まない、
抽出してはならない部分の所望の領域である、阻止領域
内のカラー画像データから色情報を読み取る事によって
得られた色分布を阻止分布として、抽出分布の輪郭を十
分に多数の色抽出範囲の組み合わせで近似した色分布抽
出範囲にあっては、色分布抽出範囲を構成する色抽出範
囲の数を減らすために、隣接する複数の色抽出範囲を被
合成色抽出範囲として選択し、前記被合成色抽出範囲を
内包する１つの色抽出範囲である合成色抽出範囲を得
て、前記合成色抽出範囲が、阻止分布を含まないとき
に、被合成色抽出範囲を合成色抽出範囲と置き換えて、
抽出分布の輪郭を複数の色抽出範囲で近似する。したが
って、色分布抽出範囲を構成する色抽出範囲から近接す
るものを選択し、選択した複数の色抽出範囲を全て内包
する合成色抽出範囲を得て、合成色抽出範囲が阻止分布
に掛からなければ選択した複数の色抽出範囲と合成色抽
出範囲とを置き換えて色抽出範囲の総数を減らして行く
ので、阻止分布と抽出分布が接近している部分は細かく
近似され、阻止分布と抽出分布が離れている部分は大ま
かに近似された色分布抽出範囲が得られる。よって限ら
れた数の色抽出範囲で、精度良く抽出分布の輪郭を近似
できるので、より精度良く処理対象を抽出できる。Furthermore, the image processing method according to the present invention does not include the color distribution of the processing object as compared with the reference image,
The color distribution obtained by reading the color information from the color image data in the inhibition area, which is the desired area of the portion that should not be extracted, is used as the inhibition distribution, and the contour of the extraction distribution is set to a sufficiently large number of color extraction ranges. In the color distribution extraction range approximated by the combination, in order to reduce the number of color extraction ranges constituting the color distribution extraction range, a plurality of adjacent color extraction ranges are selected as the combined color extraction range, and the combined target color extraction range is selected. A composite color extraction range that is one color extraction range including a color extraction range is obtained, and when the composite color extraction range does not include a rejection distribution, the composite color extraction range is replaced with the composite color extraction range,
The contour of the extraction distribution is approximated by a plurality of color extraction ranges. Therefore, a color extraction range that is close to the color extraction range is selected, a composite color extraction range that includes all of the selected color extraction ranges is obtained, and if the composite color extraction range does not fall on the inhibition distribution. Since the total number of color extraction ranges is reduced by replacing the selected multiple color extraction ranges with the composite color extraction range, the part where the blocking distribution and the extraction distribution are close to each other is finely approximated, and the blocking distribution and the extraction distribution are separated from each other. In the shaded area, a color distribution extraction range roughly approximated is obtained. Therefore, the contour of the extraction distribution can be accurately approximated with a limited number of color extraction ranges, so that the processing target can be more accurately extracted.

【００７５】そして、本発明による画像処理方法におい
ては、処理対象の色分布が広いものであったり、処理対
象とそれ以外の部分の色分布が近接しているような場合
にも不要な部分を抽出せずに所望の部分だけを抽出でき
るような画像処理方法を提供する事で、色抽出範囲や照
明条件を熟練を要せずに設定でき、抽出後の完全な処理
領域の形から画像計測を行えるようにして画像処理を扱
い易くするという本発明の目的を達成する事ができる。In the image processing method according to the present invention, even if the color distribution of the processing target is wide or the color distributions of the processing target and other parts are close to each other, unnecessary portions are not removed. By providing an image processing method that allows you to extract only the desired part without extracting, you can set the color extraction range and lighting conditions without skill, and measure the image from the shape of the complete processing area after extraction. It is possible to achieve the object of the present invention to make the image processing easy to handle by performing the above.

【図面の簡単な説明】[Brief description of drawings]

【図１】本発明による画像処理方法の実施例２における
処理の構成を示す流れ図である。FIG. 1 is a flowchart showing a configuration of processing in a second embodiment of an image processing method according to the present invention.

【図２】本発明による画像処理方法の実施例１における
処理の構成を示す流れ図である。FIG. 2 is a flowchart showing a configuration of processing in the first embodiment of the image processing method according to the present invention.

【図３】本発明の色分布取得処理１０５の一実施例を説
明するための説明図である。FIG. 3 is an explanatory diagram illustrating an example of a color distribution acquisition process 105 of the present invention.

【図４】本発明の一実施例における色分布抽出範囲１１
１を説明するための説明図である。FIG. 4 is a color distribution extraction range 11 according to an embodiment of the present invention.
It is an explanatory view for explaining 1.

【図５】本発明の色分布抽出範囲取得処理１０８の一実
施例を示し、処理の構成を示す流れ図の一部である。FIG. 5 shows an embodiment of the color distribution extraction range acquisition processing 108 of the present invention, and is a part of a flow chart showing the configuration of the processing.

【図６】本発明の色分布抽出範囲取得処理１０８の一実
施例を示し、処理の構成を示す流れ図の一部である。FIG. 6 shows an embodiment of a color distribution extraction range acquisition processing 108 of the present invention, and is a part of a flowchart showing a processing configuration.

【図７】本発明の色分布抽出処理１１３の一実施例を示
し、処理の構成を示す流れ図の一部である。FIG. 7 shows an embodiment of the color distribution extraction processing 113 of the present invention, and is a part of a flowchart showing the configuration of the processing.

【図８】本発明の色分布抽出処理１１３の一実施例を示
し、処理の構成を示す流れ図の一部である。FIG. 8 shows an embodiment of the color distribution extraction processing 113 of the present invention, and is a part of a flowchart showing the configuration of the processing.

【図９】本発明による色分布抽出範囲取得処理１０８に
おける処理の過程を説明するための説明図である。FIG. 9 is an explanatory diagram illustrating a process of a color distribution extraction range acquisition process 108 according to the present invention.

【図１０】本発明による色分布抽出範囲取得処理１０８
における処理の過程を説明するための説明図である。FIG. 10: Color distribution extraction range acquisition processing 108 according to the present invention
4 is an explanatory diagram for explaining a process of processing in FIG.

【図１１】本発明による色分布抽出範囲取得処理１０８
における処理の過程を説明するための説明図である。FIG. 11: Color distribution extraction range acquisition processing 108 according to the present invention
4 is an explanatory diagram for explaining a process of processing in FIG.

【図１２】本発明による色分布抽出範囲取得処理１０８
における処理の過程を説明するための説明図である。FIG. 12: Color distribution extraction range acquisition processing 108 according to the present invention
4 is an explanatory diagram for explaining a process of processing in FIG.

【図１３】本発明による色分布抽出範囲取得処理１０８
における処理の過程を説明するための説明図である。FIG. 13: Color distribution extraction range acquisition processing 108 according to the present invention
4 is an explanatory diagram for explaining a process of processing in FIG.

【図１４】本発明による色分布抽出範囲取得処理１０８
における処理の過程を説明するための説明図である。FIG. 14: Color distribution extraction range acquisition processing 108 according to the present invention
4 is an explanatory diagram for explaining a process of processing in FIG.

【図１５】本発明による色分布抽出範囲取得処理１０８
における処理の過程を説明するための説明図である。FIG. 15: Color distribution extraction range acquisition processing 108 according to the present invention
4 is an explanatory diagram for explaining a process of processing in FIG.

【図１６】本発明の近似範囲合成処理１１０の一実施例
を示し、処理の構成を示す流れ図の一部である。FIG. 16 is a part of a flowchart showing an embodiment of an approximate range synthesizing process 110 of the present invention and showing the configuration of the process.

【図１７】本発明の近似範囲合成処理１１０の一実施例
を示し、処理の構成を示す流れ図の一部である。FIG. 17 is a part of a flowchart showing an embodiment of an approximate range synthesizing process 110 of the present invention and showing the configuration of the process.

【図１８】本発明による近似範囲合成処理１１０におけ
る処理の過程を説明するための説明図である。FIG. 18 is an explanatory diagram illustrating a process of the approximate range synthesizing process 110 according to the present invention.

【図１９】本発明による近似範囲合成処理１１０におけ
る処理の過程を説明するための説明図である。FIG. 19 is an explanatory diagram illustrating a process of an approximate range synthesis process 110 according to the present invention.

【図２０】本発明による近似範囲合成処理１１０におけ
る処理の過程を説明するための説明図である。FIG. 20 is an explanatory diagram illustrating a process of an approximate range synthesizing process 110 according to the present invention.

【図２１】本発明による近似範囲合成処理１１０におけ
る処理の過程を説明するための説明図である。FIG. 21 is an explanatory diagram illustrating a process of an approximate range synthesis process 110 according to the present invention.

【符号の説明】[Explanation of symbols]

１０１ 参照画像 １０２ 処理対象 １０３ 抽出領域 １０４ 阻止領域 １０５ 色分布取得処理 １０６ 抽出分布 １０７ 阻止分布 １０８ 色分布抽出範囲取得処理 １０９ 抽出分布近似範囲 １１０ 近似範囲合成処理 １１１ 色分布抽出範囲 １１２ 処理画像 １１３ 色分布抽出処理 １１４ 抽出画像 101 Reference Image 102 Processing Target 103 Extraction Area 104 Blocking Area 105 Color Distribution Acquisition Processing 106 Extraction Distribution 107 Blocking Distribution 108 Color Distribution Extraction Range Acquisition Processing 109 Extraction Distribution Approximate Range 110 Approximation Range Synthesis Processing 111 Color Distribution Extraction Range 112 Processed Image 113 Colors Distribution extraction process 114 Extracted image

Claims (4)

【特許請求の範囲】[Claims] 【請求項１】 画像処理の対象である処理対象の有する
色分布、あるいは処理対象の画像を含んでいるカラー画
像を得て参照画像とし、参照画像より処理対象の有する
色分布を含む、所望の領域である、抽出領域内のカラー
画像データから色情報を読み取る事によって得られた色
分布を抽出分布として、抽出分布の輪郭を色抽出範囲で
近似する処理をし、前記色抽出範囲によって処理対象の
画像を処理画像より抽出する画像処理方法において、 抽出分布の輪郭を複数の色抽出範囲で近似した色分布抽
出範囲を得る処理をして、前記色分布抽出範囲により処
理対象の画像を処理画像より抽出する事を特徴とする画
像処理方法。1. A desired distribution that includes a color distribution of a processing target which is a target of image processing or a color image including an image of the processing target is used as a reference image, and includes a color distribution of the processing target from the reference image. The color distribution obtained by reading the color information from the color image data in the extraction area, which is the area, is used as the extraction distribution, and the contour of the extraction distribution is approximated by the color extraction range, and the processing target is based on the color extraction range. In the image processing method for extracting the image from the processed image, the process of obtaining the color distribution extraction range in which the contour of the extraction distribution is approximated by a plurality of color extraction ranges is performed, and the image to be processed is processed by the color distribution extraction range. An image processing method characterized by further extraction. 【請求項２】 色分布抽出範囲を構成する色抽出範囲を
被分割色抽出範囲として、前記被分割色抽出範囲を、任
意の分割位置で複数の色抽出範囲に分割して分割範囲と
して、分割範囲で抽出される、抽出分布のなかの、色分
布に外接する色抽出範囲を外接範囲として、外接範囲の
大きさの総和を求めて、 さらに複数の、他の任意の分割位置に関しても、外接範
囲の大きさの総和を求めて、 前記、複数の外接範囲の大きさの総和のなかで、最小の
大きさの総和である外接範囲を被分割色抽出範囲と置き
換えることにより、抽出分布の輪郭を複数の色抽出範囲
で近似することを特徴とする、請求項１記載で画像処理
方法。2. A color extraction range constituting a color distribution extraction range is set as a divided color extraction range, and the divided color extraction range is divided into a plurality of color extraction ranges at arbitrary division positions to be divided as a division range. The color extraction range that circumscribes the color distribution in the extracted distribution is calculated as the circumscribing range, and the sum of the sizes of the circumscribing ranges is calculated. Obtaining the sum of the size of the range, in the sum of the sizes of the plurality of circumscribing ranges, by replacing the circumscribing range which is the sum of the smallest sizes with the divided color extraction range, the contour of the extraction distribution The image processing method according to claim 1, wherein is approximated by a plurality of color extraction ranges. 【請求項３】 参照画像より処理対象の有する色分布を
含まない、抽出してはならない部分の所望の領域であ
る、阻止領域内のカラー画像データから色情報を読み取
る事によって得られた色分布を阻止分布として、 抽出分布及び阻止分布の属する色情報の表現範囲である
色空間から阻止分布を除いた残りの前記色空間内の領域
で、処理対象の画像を処理画像より抽出する事を特徴と
する、請求項１記載の画像処理方法。3. A color distribution obtained by reading color information from color image data in a blocking area, which is a desired area of a portion that should not be extracted and that does not include the color distribution of the processing target, from the reference image. Is a stop distribution, and the image to be processed is extracted from the processed image in the remaining area in the color space excluding the stop distribution from the color space that is the expression range of the extraction distribution and the color information to which the stop distribution belongs. The image processing method according to claim 1, wherein 【請求項４】 参照画像より処理対象の有する色分布を
含まない、抽出してはならない部分の所望の領域であ
る、阻止領域内のカラー画像データから色情報を読み取
る事によって得られた色分布を阻止分布として、 抽出
分布の輪郭を十分に多数の色抽出範囲の組み合わせで近
似した色分布抽出範囲にあっては、 色分布抽出範囲を構成する色抽出範囲の数を減らすため
に、隣接する複数の色抽出範囲を被合成色抽出範囲とし
て選択し、前記被合成色抽出範囲を内包する１つの色抽
出範囲である合成色抽出範囲を得て、前記合成色抽出範
囲が、阻止分布を含まないときに、被合成色抽出範囲を
合成色抽出範囲と置き換えて、抽出分布の輪郭を複数の
色抽出範囲で近似することを特徴とする、請求項１記載
で画像処理方法。4. A color distribution obtained by reading color information from color image data in a blocking area, which is a desired area of a portion which should not be extracted and which does not include the color distribution of the processing target from the reference image. In the case of a color distribution extraction range in which the contour of the extraction distribution is approximated by a sufficiently large number of combinations of color extraction ranges, the adjacent distributions are reduced in order to reduce the number of color extraction ranges that make up the color distribution extraction range. A plurality of color extraction ranges are selected as the combined color extraction range, a combined color extraction range that is one color extraction range that includes the combined color extraction range is obtained, and the combined color extraction range includes a rejection distribution. The image processing method according to claim 1, wherein when there is no color, the combined color extraction range is replaced with the combined color extraction range to approximate the contour of the extraction distribution with a plurality of color extraction ranges.
JP6101399A 1993-07-27 1994-05-16 Image processing method Pending JPH0798762A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP6101399A JPH0798762A (en) 1993-07-27 1994-05-16 Image processing method

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
JP5-204659 1993-07-27
JP20465993 1993-07-27
JP6101399A JPH0798762A (en) 1993-07-27 1994-05-16 Image processing method

Publications (1)

Publication Number Publication Date
JPH0798762A true JPH0798762A (en) 1995-04-11

Family

ID=26442280

Family Applications (1)

Application Number Title Priority Date Filing Date
JP6101399A Pending JPH0798762A (en) 1993-07-27 1994-05-16 Image processing method

Country Status (1)

Country Link
JP (1) JPH0798762A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010003124A (en) * 2008-06-20 2010-01-07 Hitachi Computer Peripherals Co Ltd Dropout color processor and dropout color processing method using the same
JP2011018359A (en) * 2010-09-16 2011-01-27 Hitachi Computer Peripherals Co Ltd Dropout color processing apparatus and dropout color processing method using the same
US8326028B2 (en) 2007-12-26 2012-12-04 Hitachi Computer Peripherals Co., Ltd. Dropout color processing method and processing apparatus using same

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8326028B2 (en) 2007-12-26 2012-12-04 Hitachi Computer Peripherals Co., Ltd. Dropout color processing method and processing apparatus using same
JP2010003124A (en) * 2008-06-20 2010-01-07 Hitachi Computer Peripherals Co Ltd Dropout color processor and dropout color processing method using the same
JP2011018359A (en) * 2010-09-16 2011-01-27 Hitachi Computer Peripherals Co Ltd Dropout color processing apparatus and dropout color processing method using the same

Similar Documents

Publication Publication Date Title
US4488245A (en) Method and means for color detection and modification
US5317678A (en) Method for changing color of displayed images by use of color components
US5103407A (en) Apparatus and method for color selection
US5473736A (en) Method and apparatus for ordering and remapping colors in images of real two- and three-dimensional objects
US6081276A (en) Method and apparatus for creating a color name dictionary and for querying an image by color name
US4225861A (en) Method and means for texture display in raster scanned color graphic
CN100414447C (en) Color adjusting method. color adjusting apparatus, color converting definiting compiling apparatus, image treatment apparatus, prgram and storage medium
US6021221A (en) Image processing apparatus
US5471535A (en) Method for detecting a contour of a given subject to be separated from images and apparatus for separating a given subject from images
US5105469A (en) Control data array generation apparatus and method
EP0558054B1 (en) Image filing method
US5835086A (en) Method and apparatus for digital painting
US20120007879A1 (en) Color selection interface
EP0693738A2 (en) Method and apparatus for generating color image mask
EP0369702A2 (en) Image processing apparatus and method
WO1994008422A1 (en) Method of color enhancing monochrome images
CN105096286A (en) Fusion method and device for remote sensing images
CN101751904B (en) Method for color enhancement
JPH0798762A (en) Image processing method
JP2695868B2 (en) Color stereo display
Shen et al. A holistic image segmentation framework for cloud detection and extraction
WO2021096520A1 (en) Three-dimensional color reference objects
JPH0467670B2 (en)
EP0380090A2 (en) Image processing system
JP6580220B1 (en) Coloring program and coloring method