KR20140037373A - Apparatus and method for transferring a color of image - Google Patents

Abstract

An apparatus for converting the color of an image and a converting method thereof are disclosed. The present invention converts the color of a target image by using a color comparison with a reference image which is an actual art work of a painter. The present invention is able to: convert the color of the target image automatically to have a similar color as the reference image by converting the color of the target image by using color comparison information of the reference image; and convert the color of the target image to have an artistic feel by converting the color of the target image based on a color expression mechanism of impressionists if an actual art work painted by an impressionist painter such as Van Gogh is used as the reference image.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001]

The present invention relates to an image color conversion apparatus and method, and more particularly to an apparatus and method for converting a color of a target image using a color contrast of a reference image, .

Painterly rendering is a graphics technique that expresses an effect that a person directly draws on a two-dimensional image such as a photograph. Painterly rendering technology is intended to express human-friendly images and is used to produce a variety of contents such as advertisements, movies, and posters. Painterly rendering techniques have been started to be studied in the 1990s, and research is being conducted on factors such as the expression of brushes, reproduction of the drawn process, reproduction of actual pigment, and color reproduction of actual artists. The color reproduction of real artists is recognized as one of the important factors in the aesthetic effect of the image.

The color conversion technique for transferring the color impression appearing in the reference image to the target image is largely divided into a local approach and a global approach. In the global approach, the average value and the deviation value are obtained for the color information of the entire area of the image without discriminating the area of the image, and the color of the image is converted based on the obtained average value and the deviation value. This global approach has the advantage of not needing additional manual work such as user input, but it has a limit to be used for an image having a definite division by region like a real painting work. On the other hand, the local access method divides the reference image and the target image into regions and assigns matching information between the regions to convert the color of the image. This local approach has better results than the global approach, but has the disadvantage that the user has to specify matching information.

Furthermore, there is no study on the color conversion technique using the actual picture as a reference image. It is necessary to study the color conversion considering the painter's style such as the contrast of color and the use of limited color.

Patent Document 1 discloses a multi-scale painting effect filtering method for applying to a color image. In Patent Document 1, a color image is converted into a monochrome image Wavelet transform is performed, edge detection is performed by extracting a gravity vector for each pixel position, landing is added to the length, direction, and color information of the stroke, and rendering is performed in a direction perpendicular to the gravity vector The contents are disclosed. Patent Document 2 discloses a color recommendation apparatus and method in a digital image processor. In Patent Document 2, a color of a part of an image is changed by a user's choice, And a recommendation of a color harmonizing with the selected color at the time of changing the color of the other part is disclosed to induce the change.

An object of the present invention is to provide an image color conversion apparatus and method for converting a color of a target image using a color contrast of a reference image, which is an actual work of an artist.

According to another aspect of the present invention, there is provided a computer program for causing a computer to execute a color conversion method of an image for converting a color of a target image using a color contrast of a reference image, And an object of the present invention is to provide a computer-readable recording medium having recorded thereon a computer program.

According to an aspect of the present invention, there is provided an image color conversion apparatus for converting a reference image into a plurality of cluster regions through a saliency information of a reference image, an image analyzing unit for dividing the plurality of cluster regions into color regions and analyzing color contrast of each of the plurality of cluster regions; And segmenting the target image into a plurality of segment regions through saliency information of the target image and outputting the reference image And an image converting unit for converting a color of the target image in units of the segment region using color contrast information for the plurality of cluster regions obtained from the plurality of cluster regions.

According to an aspect of the present invention, there is provided a method of transforming an image, the method including converting a reference image into a plurality of cluster regions through a saliency information of a reference image, region, and analyzing color contrast for each of the plurality of cluster regions; And segmenting the target image into a plurality of segment regions through saliency information of the target image and extracting the plurality of segment images obtained from the reference image, And converting the color of the target image in units of the segment region using color contrast information for a cluster region of the target region.

According to an aspect of the present invention, there is provided a computer readable medium storing a program for causing a computer to execute any one of the above methods.

According to the image color conversion apparatus and method of the present invention, the color of the target image is converted using the color contrast information of the reference image, You can automatically convert colors to feel similar to image.

In addition, if a real work drawn by an impressionist painter such as Van Gogh is used as a reference image, the target image is transformed into an artistic image by converting the color of the target image based on the color rendering mechanism of the impression wave The feeling can be converted to appear.

1 is a block diagram for explaining an image color conversion apparatus according to a preferred embodiment of the present invention.
2 is a diagram showing an example of saliency information according to a preferred embodiment of the present invention,
3 is a view for explaining a clustering process of a reference image according to a preferred embodiment of the present invention,
4 is a diagram for explaining a color harmony type determination process according to a preferred embodiment of the present invention.
5 is a view for explaining an example of region classification of a target image according to a preferred embodiment of the present invention,
6 is a diagram for explaining a color conversion process of a main segment region of a target image according to a preferred embodiment of the present invention.
7 is a diagram for explaining a color conversion process of the remaining segment region of a target image according to a preferred embodiment of the present invention,
8 is a view for explaining an example of color conversion of a target image according to a preferred embodiment of the present invention,
9 is a flowchart illustrating a color conversion method of an image according to a preferred embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Reference will now be made in detail to the preferred embodiments of the present invention, examples of which are illustrated in the accompanying drawings.

1 is a block diagram for explaining an image color conversion apparatus according to a preferred embodiment of the present invention.

Referring to FIG. 1, an image color conversion apparatus 100 according to an exemplary embodiment of the present invention converts a color of a target image using a color contrast of a reference image, which is a work of an artist . To this end, the image color conversion apparatus 100 includes an image analysis unit 110 and an image conversion unit 130.

The image analyzer 110 analyzes the color contrast of the reference image. That is, the image analyzer 110 divides the reference image into a plurality of cluster regions through the saliency information of the reference image. Here, the saliency region refers to a map obtained from an image through an image saliency technique for extracting a lot of human eyes from an image. In the present invention, Perona's image saliency technique is used.

2 is a diagram showing an example of saliency information according to a preferred embodiment of the present invention.

를 나타내며, (c)는 참조 영상(reference image) S를 나타내고, (d)는 참조 영상(reference image)의 세일리언시 정보(saliency information)

를 나타낸다. 붉은 색으로 표시된 영역이 상대적으로 세일리언시(saliency)가 높은 영역을 의미한다. 세일리언시 정보(saliency information)의 각 픽셀 값은 0~1의 값을 가지며, 1에 가까울수록 세일리언시(saliency)가 높다는 것을 의미한다.2 (a) shows a target image T, (b) shows a saliency information of a target image,

(C) shows the reference image S, (d) shows the saliency information of the reference image,

Indicates. The area marked in red indicates a relatively high saliency area. Each pixel value of saliency information has a value of 0 to 1, and the closer to 1, the higher the saliency.

FIG. 3 is a view for explaining a clustering process of a reference image according to a preferred embodiment of the present invention.

Referring to FIG. 3, k-means clustering is performed on a hue and saturation channel for a reference image. That is, the reference image is subjected to color quantization with K colors. Here, K is obtained experimentally, and K is 3 in the present invention. Then, the hue and saturation (HS) values of the positions where the points are clustered are set as the initial positions of clustering. The reference image is divided into a plurality of cluster regions by performing clustering based on the set initial positions.

The image analysis unit 110 analyzes the color contrast of each of the plurality of cluster regions of the reference image. That is, the image analyzing unit 110 obtains a hue and saturation (HS) value having the highest frequency in each of a plurality of cluster regions.

In more detail, a set of cluster regions R and a representative HS (Hue) of a cluster region for a reference image defined by Equations (1) and (2) and Saturation) values. Here, the representative hue and saturation (HS) value indicates a hue and saturation (HS) value having the highest frequency in the cluster region.

는 i번째 클러스터 영역(cluster region)에 속하는 점들의 집합을 나타낸다.here,

Represents a set of points belonging to the i-th cluster region.

는 i번째 클러스터 영역(cluster region)의 대표 HS(Hue and Saturation) 값을 나타낸다.here,

Represents a representative HS (Hue and Saturation) value of an i-th cluster region.

At this time, the image analysis unit 110 selects a cluster region having the highest average saliency among a plurality of cluster regions as a main cluster region. The image analysis unit 110 obtains an angle value indicating a degree of color distribution in a main cluster region.

을 획득한다.

은 아래의 [수학식 3]과 같이 정의된다.In more detail, after analyzing the saliency distribution of a plurality of cluster regions, a region having the highest saliency is selected as a main cluster region, Color contrast information in the main cluster region

.

Is defined as the following equation (3).

는 메인 클러스터 영역(main cluster region)에서 색상의 분포 정도를 나타내는 각도 값을 나타내며 아래의 [수학식 6]과 같이 정의된다.Here, p represents a representative HS (Hue and Saturation) value of the main cluster region and is defined as in Equation (4) below,

Represents an angle value indicating the degree of color distribution in the main cluster region and is defined as shown in Equation (6) below.

는 x에 속하는 픽셀들의 평균 세일리언시(saliency) 값을 의미하며, 아래의 [수학식 5]와 같이 정의된다.here,

Denotes an average saliency value of pixels belonging to x, and is defined as Equation (5) below.

의 범위는 0~360이다.here,

Is in the range of 0 to 360.

In addition, the image analyzing unit 110 may calculate a hue and saturation value based on a representative hue and saturation (HS) value of a main cluster region in each of the remaining cluster regions other than the main cluster region, A color harmony type, an angular value between a representative hue and saturation (HS) value of a main cluster region and a representative hue and saturation (HS) value of a remaining cluster region, The average position value of the other cluster region is obtained.

을 기초로 메인 클러스터 영역(main cluster region)을 제외한 나머지 클러스터 영역(other cluster region)의 색상 대비 정보

를 획득한다.

는 아래의 [수학식 7]과 같이 정의된다.More specifically, the color contrast information of the main cluster region

The color contrast information of the remaining cluster regions (excluding the main cluster region)

.

Is defined as the following equation (7).

은 메인 클러스터 영역(main cluster region)의 대표 HS(Hue and Saturation) 값을 기초로 결정된 i번째 나머지 클러스터 영역(other cluster region)의 색상 하모니(color harmony) 타입을 나타내며, 아래의 [수학식 8]과 같이 정의된다. d는 메인 클러스터 영역(main cluster region)의 대표 HS(Hue and Saturation) 값과 i번째 나머지 클러스터 영역(other cluster region)의 대표 HS(Hue and Saturation) 값 사이의 각도 값을 나타낸다. pos는 i번째 나머지 클러스터 영역(other cluster region)의 평균 위치 값을 나타내며, 아래의 [수학식 9]와 같이 정의된다.here,

Represents a color harmony type of an i-th remaining cluster region determined based on a representative hue and saturation (HS) value of a main cluster region, . d represents an angle value between a representative hue and saturation value of the main cluster region and a representative hue and saturation value of the other cluster region. pos represents the average position value of the i-th remaining cluster region, and is defined as the following Equation (9).

Here, Templete represents eight color harmony types. There are 'i type', 'V type', 'L type', 'I type', 'T type', 'Y type', 'X type' and 'N type' for color harmony type.

에 속한 픽셀의 수를 나타내고, z는 픽셀의 가로와 세로의 위치 값을 나타낸다.Here, n is

And z represents the position value of the pixel in the horizontal and vertical directions.

4 is a diagram for explaining a color harmony type determination process according to a preferred embodiment of the present invention.

은

을 기준으로 보면 색상 하모니(color harmony) 타입이 'I 타입'과 유사하고,

는

을 기준으로 보면 색상 하모니(color harmony) 타입이 'V 타입'과 유사한 것을 확인할 수 있다.Referring to FIG. 4 (b), the color harmony is composed of 8 types in total. Referring to Figure 4 (a)

silver

, The color harmony type is similar to the 'I type'

The

The color harmony type is similar to the 'V type'.

In summary, the image analysis unit 110 obtains the color contrast information of the reference image S defined by Equation (10) below.

The image converting unit 130 converts the color of the target image using the color contrast information of the reference image obtained by the image analyzing unit 110. That is, the image conversion unit 130 divides the target image into a plurality of segment regions through the saliency information of the target image. More specifically, the target image T is divided into a plurality of segment regions. Here, the number of segment regions is equal to the number of cluster regions of the reference image S. FIG.

5 is a view for explaining an example of region classification of a target image according to a preferred embodiment of the present invention. 5 (a) shows a target image, and FIG. 5 (b) shows a segment region of a target image. Referring to FIG. 5, a target image T is divided into a total of three segment regions.

The image conversion unit 130 acquires information used for image conversion in each of a plurality of segment regions of a target image. That is, the image converting unit 130 obtains a representative HS (Hue and Saturation) value having the highest frequency in each of a plurality of segment regions. At this time, the image converting unit 130 selects a segment region having the highest average saliency among a plurality of segment regions as a main segment region.

을 구하는 과정과 동일한 방식에 의해, 복수의 세그먼트 영역(segment region)의 세일리언시(saliency) 분포를 분석한 후, 가장 세일리언시(saliency)가 높은 영역을 메인 세그먼트 영역(main segment region)

으로 선정하고, 메인 세그먼트 영역(main segment region)

의 대표 HS(Hue and Saturation) 값을

으로 설정한다. 또한, 메인 세그먼트 영역(main segment region)을 제외한 나머지 세그먼트 영역(other segment region)

의 대표 HS(Hue and Saturation) 값을

로 설정한다.More specifically,

A saliency distribution of a plurality of segment regions is analyzed in the same manner as the process of obtaining a main segment region and a region having a high saliency is divided into a main segment region,

And a main segment region,

(Hue and Saturation) value of the representative HS

. In addition, in the other segment region excluding the main segment region,

(Hue and Saturation) value of the representative HS

.

The image transforming unit 130 transforms the representative HS (Hue and Saturation) values of the main segment region and the remaining HS (Hue and Saturation) values of each of the remaining segment regions except for the main segment region, (Hue and Saturation) value of the other segment region, and an average position value of the remaining segment region.

를 획득한다.More specifically, in the other segment region, the following equation (11) is defined as: " (11) "

.

Here, d represents an angle value between a representative hue and saturation value of the main segment region and a representative hue and saturation value of the i-th remaining segment region, (12). &Quot; (12) " and pos represents the average position value of the i-th other segment region.

Thereafter, the image transforming unit 130 transforms the color information of the cluster region into a segment region unit (image region) by using the color contrast information of a plurality of cluster regions obtained from the reference image by the image analyzing unit 110, To convert the color of the target image.

That is, the image transform unit 130 selects a plurality of points in the order of higher saliency values in the main segment region, and calculates a hue and saturation (HS) value of a plurality of selected points (Hue and Saturation) values obtained from the main cluster region, and the points belonging to the main segment region are calculated using the acquired average HS (Hue and Saturation) values and the angle values of the main cluster region Converts colors.

More specifically, a set H of n pixels is extracted in the order of higher saliency values in the main segment region. The set H is defined as: " (13) "

을 획득한다.

은 아래의 [수학식 14]와 같이 정의된다.Then, an average HS (Hue and Saturation) value from the HS (Hue and Saturation) values of the pixels belonging to the set H

.

Is defined as the following equation (14).

과 메인 클러스터 영역(main cluster region)의 각도 값을 이용하여 메인 세그먼트 영역(main segment region)에 속하는 점들을 색상을 변환한다.Then, the obtained average HS (Hue and Saturation) value

And an angle value of a main cluster region to convert colors of points belonging to a main segment region.

6 is a diagram for explaining a color conversion process of a main segment region of a target image according to a preferred embodiment of the present invention.

을 획득한 평균 HS(Hue and Saturation) 값

으로 변경하고, 메인 클러스터 영역(main cluster region)의

에 설정된 각도 값

의 범위 안에서 메인 클러스터 영역(main cluster region)에 속하는 나머지 픽셀들의 색상을 쉬프트(shift)한다.Referring to FIG. 6, a representative HS (Hue and Saturation) value of an existing main segment region

Average hue and saturation (HS) value obtained

, And in the main cluster region (main cluster region)

Angle value set in

The color of the remaining pixels belonging to the main cluster region is shifted within a range of " 0 "

In addition, the image transforming unit 130 transforms the other segment region using the angle value and the average position value of the remaining segment region and the angle value and the average position value of the remaining cluster region (other cluster region) ) And the remaining cluster region (one cluster region).

More specifically, a set V1 of the remaining segment region defined by the following equation (15) and a set V2 of the remaining cluster region defined by the following equation (16) The other segment region and the other cluster region are matched one-to-one using Equation (17).

는 아래의 [수학식 18]과 같이 정의된다.here,

Is defined as the following equation (18).

는 아래의 [수학식 19]와 같이 정의된다.here,

Is defined as the following equation (19).

는 유클리디안 디스턴스(Euclidean distance)를 나타내며,

는 영역 X의 각도 값을 나타내고,

는 영역 X의 평균 위치 값을 나타낸다.here,

Represents an Euclidean distance,

Represents the angular value of the area X,

Represents the average position value of the area X.

The image converting unit 130 converts the color of the remaining segment region into a color of the other cluster region matching the color of the other segment region by the same process as the color conversion of the main segment region Use the color harmony type to convert.

More specifically, the color of the other segment region is matched with the color harmony of the other cluster region, which is the same as the operation of converting the color of the main segment region, harmony) type.

FIG. 7 is a diagram for explaining the color conversion process of the remaining segment region of the target image according to the preferred embodiment of the present invention.

은 제2나머지 클러스터 영역(other cluster region)

의 색상 하모니(color harmony) 타입을 이용하여 제1나머지 세그먼트 영역(other segment region)에 속하는 픽셀들의 색상을 쉬프트(shift)한다. 그리고, 제2나머지 세그먼트 영역(other segment region)

는 제1나머지 클러스터 영역(other cluster region)

의 색상 하모니(color harmony) 타입을 이용하여 제2나머지 세그먼트 영역(other segment region)에 속하는 픽셀들의 색상을 쉬프트(shift)한다.Referring to FIG. 7, a first remaining segment region,

A second cluster region < RTI ID = 0.0 >

The color of the pixels belonging to the first remaining segment region is shifted by using the color harmony type of the first color. In addition, the second remaining segment region

A first cluster region < RTI ID = 0.0 >

The color of the pixels belonging to the second remaining segment region is shifted by using the color harmony type of the color harmony.

8 is a diagram for explaining an example of color conversion of a target image according to a preferred embodiment of the present invention. 8A shows a color distribution of an original image of a target image, FIG. 8B shows a color distribution of a result image obtained by changing a color of a target image, .

Referring to FIG. 8, it can be seen that the color distribution before and after the color conversion is different. That is, the color contrast of the target image is changed so as to be similar to the color contrast of the reference image.

In this way, by converting the color of the target image using the color contrast information of the reference image, the target image is automatically converted to the color Lt; / RTI >

In addition, if a real work drawn by an impressionist painter such as Van Gogh is used as a reference image, the target image is transformed into an artistic image by converting the color of the target image based on the color rendering mechanism of the impression wave The feeling can be converted to appear.

9 is a flowchart illustrating a color conversion method of an image according to a preferred embodiment of the present invention.

Referring to FIG. 9, the image color conversion apparatus 100 analyzes a color contrast of a reference image (S910). That is, the apparatus 100 for converting color of an image divides the reference image into a plurality of cluster regions through the saliency information of the reference image. The image color conversion apparatus 100 analyzes the color contrast of each of a plurality of cluster regions of a reference image. In other words, the image color conversion apparatus 100 may include a representative hue and saturation value in each of a plurality of cluster regions, an angle value of a main cluster region, The color harmony type of the other cluster region, the angle value of the other cluster region, and the average position value of the remaining cluster region.

Then, the image color conversion apparatus 100 converts the color of the target image using the color contrast information of the reference image (S930). That is, the image color conversion apparatus 100 divides the target image into a plurality of segment regions through the saliency information of the target image. The image color conversion apparatus 100 acquires information used for image conversion in each of a plurality of segment regions of a target image. In other words, the color conversion apparatus 100 of the image has a representative hue and saturation (HS) value in each of a plurality of segment regions, an angular value of the remaining segment region, region is obtained. Then, the image color conversion apparatus 100 converts the color information of the target image (image region) into a plurality of segment regions by using the color contrast information for a plurality of cluster regions obtained from the reference image, ).

The present invention can also be embodied as computer-readable codes on a computer-readable recording medium. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer apparatus is stored. Examples of the computer-readable recording medium include a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like in the form of a carrier wave . In addition, the computer-readable recording medium may be distributed to computer devices connected to a wired / wireless communication network, and a computer-readable code may be stored and executed in a distributed manner.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It will be understood by those skilled in the art that various changes may be made and equivalents may be substituted without departing from the scope of the appended claims.

100: Image color conversion apparatus, 110: Image analysis unit,
130:

Claims (11)

A reference image is divided into a plurality of cluster regions through salience information of a reference image and a color of each of the plurality of cluster regions An image analysis unit for analyzing contrast; And
The target image is divided into a plurality of segment regions through salary information of a target image, and the reference image is analyzed by the image analyzer. And an image converter configured to convert colors of the target image in units of segment regions by using color contrast information of the plurality of cluster regions obtained from the plurality of cluster regions. Color converting apparatus for images. The method of claim 1,
The image analyzer obtains a representative Hue (Hue and Saturation) value having the highest frequency in each of the plurality of cluster regions, and averages the saliency of the plurality of cluster regions. This highest cluster region is chosen as the main cluster region.
The image converting unit obtains a representative Hue (Hue and Saturation) value having the highest frequency in each of the plurality of segment regions, and averages the saliency of the plurality of segment regions. And the highest segment region is selected as the main segment region. 3. The method of claim 2,
Wherein the image analyzing unit obtains an angle value indicating a degree of color distribution in the main cluster region and determines an angle value of the main cluster region in each of the remaining cluster regions other than the main cluster region, A color harmony type based on a representative hue and saturation value of a main cluster region, a representative hue and saturation value of the main cluster region, (hue and saturation) value of the other cluster region and an average position value of the remaining cluster region,
Wherein the image transformer transforms the representative HS (Hue and Saturation) values of the main segment region and the other segment region (other) of each of the other segment regions except for the main segment region, (Hue and Saturation) value of a segment region of the current segment region, and an average position value of the remaining segment region. The method of claim 3, wherein
Wherein the image transformation unit selects a plurality of points in order of higher saliency values in the main segment region and calculates average HS values from the HS values of the selected plurality of points, Obtaining a hue and saturation value of the main cluster region and using the average hue and saturation value and an angle value of the main cluster region to color the points belonging to the main segment region Wherein the color conversion unit converts the color of the image into a color signal. 5. The method of claim 4,
Wherein the image transformation unit transforms the other segment region and the other segment region using an angle value and an average position value of the other segment region and an angle value and an average position value of the remaining cluster region, The other cluster regions are matched one-to-one and the color of the other segment region is converted using the color harmony type of the other cluster region to be matched A color conversion device for the image. A reference image is divided into a plurality of cluster regions through salience information of a reference image and a color of each of the plurality of cluster regions Analyzing contrast; And
The target image is divided into a plurality of segment regions through salary information of a target image, and the plurality of target images are obtained from the reference image. And converting a color of the target image in units of segment regions by using color contrast information of a cluster region. The method according to claim 6,
In the image analysis step, a representative HS (Hue and Saturation) value having the highest frequency in each of the plurality of cluster regions is obtained, and a saliency of the plurality of cluster regions is obtained. The cluster region having the highest average value of is selected as the main cluster region.
In the image conversion step, a representative HS (Hue and Saturation) value having the highest frequency in each of the plurality of segment regions is obtained, and a saliency of the plurality of segment regions is obtained. And selecting a segment region having the highest average value as a main segment region. 8. The method of claim 7,
The method of claim 1, wherein, in the image analysis step, an angle value indicating a degree of color distribution is obtained in the main cluster region, and in each of the remaining cluster regions other than the main cluster region, A color harmony type based on a representative hue and saturation value of a main cluster region, a representative hue and saturation value of the main cluster region, An angle value between a representative hue and saturation (HS) value of the other cluster region and an average position value of the remaining cluster region,
In the image conversion step, a representative HS (Hue and Saturation) value of the main segment region in each of the other segment regions except for the main segment region, an angle value between a representative hue and saturation (HS) value of the other segment region, and an average position value of the other segment region. The method of claim 8,
Wherein the image transformation step comprises the steps of: selecting a plurality of points in the order of higher saliency values in the main segment region and selecting a plurality of points from the hue and saturation values of the selected plurality of points; (Hue and Saturation) value and an angle value of the main cluster region to obtain points that belong to the main segment region, And converting the color of the image. The method of claim 9,
The method of claim 1, wherein, in the image transformation step, an angle value and an average position value of the other segment region and an angle value and an average position value of the remaining cluster region are used to calculate the other segment region ) And the other cluster region in a one-to-one correspondence, and the color of the remaining segment region is transformed by using a color harmony type of the remaining cluster region to be matched And converting the color of the image into a color signal. A computer-readable recording medium having recorded thereon a program for causing a computer to execute a color conversion method of an image according to any one of claims 6 to 10.

Images