JP2009296372A - Color conversion method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a color conversion method that can convert color by limiting break of color tone relationship having certain significance to a low level. <P>SOLUTION: When an original image existing in a first color space M1 and composed by a plurality of objects (large areas A, B, C, intermediate areas a1-a3, b1-b3, and so on) is converted in color to an image within a second color space M2, a difference between a color difference between two objects in the first color space and a color difference obtained after color conversion into the second color space between the two objects is used as a distortion, and when using a force exerted in a direction of canceling this distortion as a stress, a stress characteristic (in the figure, a solid line between the objects represents a stress) for distortion caused between one object and the other object is determined for each of the objects constituting the original image to make color conversion in such a manner that a sum of stresses is reduced. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a color conversion method for color-converting an original image existing in a first color space into a second color space.

  In many cases, the color range that can be displayed on the monitor screen (editing color space) does not match the color range that can be output by an output device such as a printer or projector (output color space). For this reason, when an edited image is output by an output device while viewing the monitor screen, it may be output in a color different from the color confirmed on the monitor screen. Thus, various techniques for ensuring color matching between the edit color space and the output color space have been proposed.

  For example, there is an editing apparatus that enables editing while emulating the color space of the output device while checking the color tone at the time of output on a monitor screen (see, for example, Patent Document 1). Further, a color conversion method has been proposed in which color conversion is performed while maintaining saturation as much as possible by applying elastic deformation to a reproducible color range (gamut) (see, for example, Patent Document 2).

  Further, a technique of reducing the number of colors (number of nodes) according to a rule such as integrating colors that are not adjacent to each other and reducing the number of colors and rearranging the colors (see, for example, Patent Document 3). , A technique for adding texture to a grayscale representation of colors that are collided (converted to the same gray level) when converting a color image to black and white (see, for example, Patent Document 4), and when the color difference between adjacent areas is small There is also a technique (for example, refer to Patent Document 5) that secures discrimination by applying decoration such as adding a frame or adding a texture.

JP 2004-139496 A JP 7-123283 A Japanese Patent Laid-Open No. 10-149347 JP 2002-218263 A JP 2007-94585 A

  When emulating the output device color space within the edit color space, accurate color matching can only occur in the common region of the color space. For this reason, when the color space on the output side is small, it is inevitable to compress the color difference. For example, the image P shown in FIG. 3 is divided into elements (objects) of a large area A, a large area B, and a large area C (background), and the large area A is further divided into a medium area a1, a medium area a2, and a medium area a3 ( The large area B is composed of the elements of the middle area b1, the middle area b2, and the middle area b3 (background). The middle area a2 further includes a small area a21, a small area a22, and a small area. It has an object structure composed of a23 (background). FIG. 28A shows the object structure of the large area A, the object structure of the large area B, and the object structure of the large area C individually in relation to the chromaticity thereof. The small area is not shown. The positional relationship between the objects indicates their color relationship. A solid line connecting the objects represents the strength of the relationship between the objects. When the image P composed of these objects is represented in the first color space M1 of the monitor screen of the editing apparatus, it is as shown in FIG. When this image P is handled by a system that emulates an output device having the second color space M2 and is displayed on the monitor screen, as shown in FIG. 28C, the first color space M1 and the second color are displayed. The image P is compressed and stored in the common area of the space M2, and the color difference and saturation range between objects is reduced.

  Further, when the image P is color-converted into the color space M2 of the output device while maintaining the saturation by elastic deformation of the gamut, the color difference in one direction is compressed and the other direction is compressed as shown in FIG. Will be expanded, and unintended color-related distortion will occur. For example, the three objects b1, b2, and b3 constituting the large area B originally have a triangular color arrangement, but when the color is converted to the color space M2, the color arrangement is almost aligned. The color relationship between the objects b1, b2, and b3 constituting the large area B is greatly broken. If the editor gives some meaning to the color relationship between the objects b1, b2, and b3, the output image loses its intention.

  Further, in the technique of Patent Document 3, the number of nodes is reduced according to a rule of integrating colors that are not adjacent to each other regardless of the editor's identification intention. Objects that are to be expressed in different tones may be integrated into the same color, such as being integrated into the same color, and the meaning may be lost if the color is meaningful.

  In addition, a technique for applying decoration such as a frame or texture is effective for ensuring discrimination. However, in color conversion, it is desirable to give priority to ensuring discrimination by color tone and to avoid adding decoration as much as possible.

  An object of the present invention is to provide a color conversion method capable of performing color conversion while minimizing the disruption of a meaningful color tone relationship.

  The gist of the present invention for achieving the object lies in the inventions of the following items.

[1] In a color conversion method for color-converting an original image existing in a first color space and composed of a plurality of objects into an image in a second color space,
The difference between the color difference between two objects in the first color space and the color difference between the two objects after color conversion into the second color space is used as a distortion, and the force acts in a direction to eliminate the distortion. For each of the objects constituting the original image, the characteristic of the stress with respect to the distortion generated between the other objects is determined, and the color conversion is performed so that the total sum of the stresses becomes small. A color conversion method characterized by the above.

  In the above invention, the stress characteristics are set between the objects, and the color conversion is performed so that the total sum of the stresses becomes small. Color conversion can be performed from the space to the second color space. An object is an element in an image on which uniform color conversion processing is performed or a related set of the elements. For example, a single area filled with solid or similar texture, a background, characters, a graph, a photograph, and the like.

[2] The color conversion method according to [1], wherein the stress characteristic is determined based on an object structure of the original image.

  In the said invention, the stress characteristic between each object is set based on an object structure. The object structure is an arrangement relationship or connection method of objects constituting an image. Attributes of the object itself (character, figure, photo type, font type, size, color, monochrome, etc.) may be included in the elements of the object structure.

[3] The color conversion method according to [1] or [2], wherein the color conversion is performed so that the sum is minimized.

  In the above invention, optimization is performed so that the total sum of stresses is minimized.

[4] As a result of the color conversion, if the amount of change in the color tone of an arbitrary object or the distortion between two arbitrary objects exceeds a predetermined threshold value, the user can change whether or not the color conversion is possible or change the conditions regarding the color conversion. The color conversion method according to any one of [1] to [3], wherein

  In the above invention, if the change in color tone of the object due to color conversion or the change amount (distortion) of the color difference between the objects is larger than the threshold value, there is a possibility that a color tone change that is not intended by the operator or exceeds an allowable range may occur. The operator is inquired about whether to change the color conversion conditions or whether the color conversion is appropriate. Conditions include, for example, restrictions on setting stress characteristics.

[5] The color conversion method according to any one of [1] to [4], wherein when the color conversion is performed, a change in color tone for a predetermined object is limited to a predetermined color region.

  In the above invention, color conversion is prevented such that the object to be restricted is out of the predetermined color region. For example, a color area corresponding to a color name such as orange or red is defined, and an object belonging to a predetermined color name is restricted from coming out of the color area corresponding to the color name by color conversion.

[6] The color conversion method according to any one of [1] to [5], wherein the stress characteristic is non-linear.

  In the above-described invention, it is possible to limit and restrict color tone change in color conversion with various degrees of freedom by making it non-linear. In addition, fine color control is possible, contributing to the realization of a technique for minimizing the total sum of stresses.

[7] The stress is divided into a stress relating to the luminance or brightness component and a stress relating to the other component, and a characteristic of each stress is set so that the stress relating to the brightness or brightness component is greater than the stress relating to the other component. The color conversion method according to any one of [1] to [6], wherein the color conversion method is set.

  In the above invention, the change (distortion) of the color difference caused by the color conversion is divided into a luminance or brightness component and a component not containing the component (for example, hue), and the stress characteristics are set for each. Since the luminance and brightness components have a greater visual impact than other components such as hue and saturation, the stress on the distortion of the luminance and brightness components should be greater than the stress on the distortion of the other components. If the stress characteristics are individually set, color conversion can be performed while maintaining the “easy to see” of the material better.

[8] The object according to any one of [1] to [7], wherein a decoration corresponding to the residual stress is applied to an object whose residual stress after the color conversion is a predetermined value or more. The color conversion method described.

  In the above invention, even when the conversion destination color space is very small and residual stress is inevitably left between objects, the distinctiveness due to the residual stress is reduced by applying decorations such as adding contours and textures. Thus, a material that better utilizes the object setting intention in the editing environment (first color space) can be created in the second color space.

[9] The color conversion method according to any one of [1] to [8], wherein setting of a constraint condition related to color conversion of a specific object is received.

  In the above invention, color conversion is performed under the set constraint conditions. Thereby, for example, it is possible to place importance on the color tone of an object that is meaningful in the absolute color tone itself, and it is possible to create a color tone balance material more suitable for the drawing intention in the second color space.

[10] The color conversion method according to any one of [1] to [9], wherein an object belonging to the specific pattern is detected, and a predetermined restriction condition is set for the color conversion of the detected object. .

  In the above invention, for example, an object having a specific pattern such as “person's face” is detected, and color conversion is performed by automatically setting constraint conditions so that the object does not exit a predetermined color area defined as “person's face”. Done. Accordingly, it is possible to place importance on the color tone of an object having a memory color such as “person's face” so as not to deviate from the memory color, and to create a material with a more preferable color tone in the second color space. Can do.

[11] When the color conversion of the first image and the second image is performed, the object constituting the first image and the object constituting the second image are more than a predetermined relationship with each other. The color according to any one of [1] to [10], wherein when there is an object having strength, a predetermined constraint condition is set with respect to color conversion of an object having a related strength greater than or equal to the predetermined strength. Conversion method.

  In the above-described invention, it is possible to consider that the color tone of a pattern or the like that is repeatedly cited in a material covering a plurality of pages can be considered, and a material having a more preferable color balance can be created in the second color space. For example, if there is a similar object with high related strength such as a company logo mark on multiple pages of material, it is possible to convert the color so that the color of the logo mark is unified in the multiple pages of material. Become.

[12] The constraint conditions are: (1) restricting the color change of the object within a predetermined color area; (2) the color of the object in the first color space and the color converted in the second color space; (3) set the stress for the distortion as a difference from the color, and (3) use the difference between the color determined based on the predetermined definition and the color in the second color space after the color conversion of the object as the distortion. The color conversion method according to any one of [9] to [11], wherein at least one of setting stress for strain is set.

  In the above invention, in the case of (1), the color after color conversion is restricted so as not to go out of the set color area. In the case of (2), it is restricted so that the color tone does not greatly deviate from the original color. In the case of (3), for example, the color area corresponding to the color name is restricted from coming out. By using the exemplified method as the constraint condition, it is possible to easily add a condition.

[13] In a color conversion method for color-converting an original image that exists in a first color space and includes a plurality of objects into an image in a second color space,
A color conversion method characterized in that a degree of association is set between each object constituting the original image, and the color conversion is performed such that a change in color difference between objects becomes smaller between objects having a higher degree of association.

  In the above invention, the degree of association is set between the objects, and the color conversion is performed so that the change in color difference between the objects with the strong degree of association is reduced. For example, if the relevance level is set higher as the degree of deterioration in distinguishability between objects due to a change in color tone between objects, color conversion can be performed so that the distinguishability does not decrease.

[14] The color conversion method according to [13], wherein the related strength is determined based on an object structure of the original image.

[15] The color conversion method according to any one of [1] to [14], wherein the color conversion is executed only when the number of objects constituting the original image is three or more.

In the above invention, it is possible to define two or more types of stress characteristics or relevance by having three or more objects.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 shows a configuration of an image processing system 10 that performs color conversion using a color conversion method according to an embodiment of the present invention. The image processing system 10 includes an image input unit 11 including one or more arbitrary image input devices such as a communication unit 11a, an imaging unit 11b, and a scanner unit 11c, and an image acquired by the image input unit 11 in various formats. An image storage unit 12 that stores the image as an image file, an editing processing unit 13 that performs an editing function of an image stored in the image storage unit 12, a function for setting conditions for color conversion, and the like, and during editing and before and after editing A monitor device 14 serving as a first image output device for displaying an image and the like, an instruction input unit 15 for receiving various operations related to image editing, input operations of conditions for color conversion, and the like, and a set The image processing unit 16 executes color conversion according to conditions, and a second image output device 17 that visualizes and outputs the color-converted image.

  For the image storage unit 12, for example, a large-capacity storage device such as a hard disk device is used. The editing processing unit 13 and the image processing unit 16 are mainly configured by a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and the like. The functions of the editing processing unit 13 and the image processing unit 16 are realized by the CPU executing various processes according to the program stored in the ROM. The monitor device 14 is composed of a color display device. The instruction input unit 15 includes a keyboard, a mouse, and the like. The second image output device 17 may be one or more arbitrary devices that visualize an image, such as a printer 17a or a projection display 17b.

  A color area that can be displayed on the monitor device 14 is a first color space, and a color area that can be visualized and output by the second image output device 17 is a second color space. The first and second color spaces have different reproducible color areas.

  An image file designated as a processing target by the operator through the instruction input unit 15 is read from the image storage unit 12, captured in the editing processing unit 13, and displayed on the monitor device 14. The operator looks at the image on the monitor device 14 and performs color confirmation and various editing operations. When an editing operation is input from the instruction input unit 15, the image processing unit 16 edits and processes an image (image file) displayed on the monitor device 14 according to the input editing operation. Further, the image processing unit 16 performs color conversion on the image file (original image existing in the first color space) that has been edited and processed into an image in the second color space according to the conditions input from the instruction input unit 15. To the second image output device 17, and the second image output device 17 visualizes and outputs the image.

  FIG. 2 shows an outline of processing related to color conversion performed by the image processing unit 16. First, image information (image file) of an original image that is a target of color conversion is acquired from the image storage unit 12 (step S101). Next, the original image is divided into objects, and the object structure is defined.

  Here, the object is one element in the image on which uniform color conversion processing is performed. There is typically a single area filled with solid or similar textures, or a series of sentences written in similar fonts, but one area consisting of several components, such as skin-colored solids and dots, lines A face pattern or the like represented by combining elements may be set as one object, or several objects may be grouped and defined as a larger area (or higher hierarchy) object.

  Note that an object is also defined when processing an image having a vast image gradation in a region such as a natural image. The object in this case is not an area for performing a single process, but represents a representative image gradation in the natural image. After the color conversion condition is determined using the representative value, the image gradation conversion condition is determined. The natural image object will be described in detail later.

  The original image P shown in FIG. 3 is largely divided into three objects, a large area A, a large area B, and a large area C (background). The large area A is divided into a medium area a1, a medium area a2, and a medium area a3 (background). ), The large area B is divided into the middle area b1, the middle area b2, and the middle area b3 (background), and the middle area a2 is further divided into the small area a21, the small area a22, and the small area a23 ( The object is divided into background objects. The small area may be further divided into objects. For example, it is possible to change the setting of how far to divide, and to divide up to a set level.

  FIG. 4 shows the object structure of the original image P shown in FIG. Each object has a hierarchical structure. Actually, there may be a lower object that is not completely included in the upper object, and in such a case, a relationship in which a related line is directly drawn by jumping from the upper hierarchy to the intermediate hierarchy can be defined. The total sum of stress may be obtained for each related line (line connecting objects) and added, and the number of related lines is not particularly limited.

  “Division of object” in step S102 of FIG. 2 is performed when image information is not structured, such as when a hard copy original is read using the scanning function of the scanner unit 11c or a separate multifunction machine. If the original image that is the target of color conversion has already been constructed with an object structure (for example, various patterns and fonts placed in the presentation creation software), it is not necessary to define the object structure immediately. You can get started. Even if the area has already been structured, if a part of the area is unstructured information, a process for dividing the area into objects is required.

  When the division and structuring of the object (step S102) is completed, the “stress characteristic” for the distortion between the related objects is subsequently defined and determined (step S103). Distortion is defined as an index indicating a change in color difference between two target objects. That is, the difference between the color difference between two objects in the first color space and the color difference after color conversion into the second color space between the two objects is defined as distortion. The force acting in the direction of eliminating the distortion is used as stress, and the stress characteristic with respect to the distortion generated between other objects is determined for each of the objects constituting the original image.

  FIG. 5 shows an example of distortion in the CIELAB color system. When the color difference between the object a1 and the object a2 in the first color space before the color conversion is ΔE, and the color difference between the object a1 and the object a2 after the color conversion into the second color space is ΔE ′, the distortion is abs (ΔE −ΔE ′). Note that abs (numerical value) represents the absolute value of the numerical value. In the example of FIG. 5, the color of the object a1 does not change before and after the color conversion, and only the color of the object a2 changes. Therefore, the distance moved by the object a2 by the color conversion is the distortion Q.

  ΔE is calculated based on the coordinate values of the CIELAB color space. However, the method of calculating the color difference is not limited to this, and various equivalent color spaces such as ΔE00 and other proposed calculation methods, CIEluv, and the like. And may be calculated in a color space to which a model relating to “color appearance” such as CIECAM02 is applied. Furthermore, since the human visual characteristics have a high spatial resolution capability with respect to changes in brightness, it is considered that maintaining the brightness relationship is more important. Assuming a virtual color space in which the change of the coordinate value in the brightness direction is set larger than the definition of the original uniform color space, ΔE is calculated in the virtual space, and the change in the brightness direction is regarded as a larger distortion. The brightness relationship may be maintained more strongly.

  FIG. 6 illustrates various stress characteristics. The horizontal axis is graphed as strain, and the vertical axis is plotted as stress. FIG. 6A illustrates a linear stress characteristic with respect to positive and negative strains. That is, a simple elasticity such as a spring is defined, and the greater the slope of the straight line, the stronger the stress (the color difference is easily maintained).

  The stress characteristics may be non-linear. For example, FIG. 6B shows an example of a stress characteristic in which a greater stress is applied in the contraction direction (negative distortion, the direction in which the color difference is reduced) than in the extension direction (positive distortion, the direction in which the color difference is increased). Is shown. That is, the color difference expansion direction is an acceptable stress characteristic. Further, it may be defined only by the puncture force. In practice, the stress characteristic can be configured by a look-up table in which a strain value is input and a stress value corresponding to the strain value is output.

  Furthermore, the stress is divided into the stress relating to the luminance or brightness component and the stress relating to the other component among the components constituting the color of the object (including the luminance component), and the stress relating to the distortion of the brightness or brightness component is the other stress. The characteristics of each stress may be individually set so as to be larger than the stress related to the component. In FIG. 6C, the stress characteristic 21 with respect to the distortion in the luminance direction and the stress characteristic 22 with respect to other components (components such as hues not including the luminance component) are individually set, and the stress with respect to the distortion in the luminance direction is the luminance. An example of the stress characteristics 21 and 22 when set to be larger than the stress with respect to a component such as a hue not including the component is shown.

  Since human visual characteristics have a high spatial resolution capability against changes in brightness, it is considered that maintaining the brightness relationship has a more important meaning. Therefore, as shown in FIG. 6C, different stress characteristics can be given to the brightness direction and the color difference (in this figure, the change in the “hue” and “saturation” directions without brightness). As a result, it is possible to provide a characteristic that maintains the brightness relationship between the objects.

  By setting the stress characteristics for each component, nonlinear stress characteristics are set as the entire color difference. When a plurality of strains are defined for each component, the stress is calculated from each strain and added. Note that, when iterative verification is performed as in the optimization method described later (FIGS. 18 and 19), there is no practical problem with addition of both vectors and scalars.

  In addition, the stress characteristic can be changed to various shapes, and the stress may suddenly increase with a certain threshold as a boundary, and this shape can be used to define various barriers. That is, when color conversion is performed, a change in color tone for a predetermined object can be limited within a predetermined color region. For example, “color name” is defined as a color attribute, a stress barrier is provided at the boundary of the color area recognized as the color name, and an object having a color inside the barrier is not color-converted to an area outside the barrier. You may restrict to.

  FIG. 7A shows an example of a color name barrier. The color area 25 defines an orange range, and the color area 26 defines a red range. For example, in the case of a red color, even if a change in color tone in a certain range can be recognized, there is a range of “red” as the color. In some cases, it is better to avoid exceeding this range, for example, to change to a color tone of “orange”. In such a case, a barrier is defined.

  An example of stress characteristics for obtaining such a barrier is shown in FIG. By causing the stress to increase rapidly or become infinite where the strain exceeds a certain value, a barrier is formed. By defining such a stress characteristic for the object, the barrier can function only for the necessary object. By providing a color name barrier in this way, it is possible to prevent changes that change the color name of an object, such as “red” to “orange”, while maintaining the degree of freedom of white adjustment. Balance materials can be created.

  For example, in the case of a red color, even if a change in color tone in a certain range can be recognized, there is a range of “red” as the color. In some cases, it is better to avoid exceeding this range, for example, to change the color tone to “orange”. In such a case, a barrier is defined. Actually, the stress suddenly increases at the point where the color conversion result exceeds the barrier. By defining the stress characteristic as shown in FIG. 7B for the object, the barrier is applied only to the necessary object. Can function.

There are the following criteria for setting the stress characteristics.
(1) Stress is high when connected, (2) Stress is high when there is a difference in luminance (3) Stress outside the contour is weak in the contoured region

  FIG. 8 schematically shows a state in which stress characteristics are set between the objects of the image P having the object structure shown in FIG. The one-point difference line in the figure indicates the inclusion relationship, the solid line indicates the stress characteristic between the objects, and the thicker solid line indicates that the stress with respect to the strain is greater. In addition, description about the minute stress between objects is abbreviate | omitted.

  Details of the setting of the stress characteristics will be described later. For example, when one object is surrounded by a clear outline as in the large area C and large area B of the image P shown in FIG. Even if the relationship changes slightly, there is no loss in the discriminability as a document. Therefore, it is not necessary to set a large stress for strain between the large region C and the large region B, and the large region C and the large region B are connected with a weak stress as shown in FIG.

  Returning to FIG. 2, the description will be continued. Next, the image processing unit 16 calculates a color difference in the first color space between each object constituting the original image and another object (step S104). Note that the execution position of the process of step S104 may be moved between step S102 and step S103.

  Next, the color space information of the second image output device 17 to be used (information indicating the range of the color area that can be expressed in the second color space) is acquired (step S105).

  Next, each object of the original image is arranged in the second color space so that the total sum of stresses when the objects of the original image are color-converted into the second color space is minimized (step S106). The image information of the original image based on the arrangement result is color-converted (step S107), and the process is ended (END).

  The above-described color conversion method of the present embodiment will be described in more detail by taking as an example the case where the original image P2 of FIG. 9 including characters, photographs, graphs, and the like is a processing target.

FIG. 10 shows the object structure of the original image P2 shown in FIG. This is an example in which characters, graphs (patterns), and photographs of continuous tone objects are included as large components. The color tone of the object in the upper layer is handled as follows.
(1) If the adjacent object is a single object, this color tone is used.
(2) In the case of a background and various characters, the stress is considered for each element. → In the calculation, the background and the character are considered as objects in the same region.
(3) In the case of continuous gradation such as a photograph, an average value of pixels in the vicinity of the adjacent region is used.

  As shown in FIG. 9, there is no clear outline between the background 31 of the large area A and the photograph 32. From this point, the relationship between the objects of the background 31 and the photograph 32 is “stress strong”. It should be. However, if a photograph has gradation and the gradation near the photographic area is distributed over a wide range, it is considered that the photographic area itself has a clear outline even if it is fragmented. Therefore, in this embodiment, the stress characteristic 33 between the background 31 and the photograph 32 is set low.

  Next, a method for dividing a continuous tone region such as the photograph 32 shown in FIG. Here, a division method is used in which several representative points are set from a continuous tone according to a predetermined rule, each representative point is treated as an object, and stress is set. This will be described in detail below.

  FIG. 11 shows the flow of object division processing by the first division method for the continuous tone area. Create color tone histogram of image (directly it is a histogram in 3D space, but for example, principal component analysis of image information may be used to create principal component histogram using one or two axes with large contribution rate) (Step S201), a high density area (histogram peak) is detected from the area (Step S202), and if there are many, a representative point (effectively located in the outer shell) is extracted. (Step S203) The color tone of the object is set. The stress characteristics between the extracted objects are set uniformly (same stress characteristics between the objects).

  FIG. 12 shows each peak detected from the color tone histogram (two-dimensional) in the continuous tone area as a black circle on the color coordinate. In addition, among the detected peaks, those located in the outer shell are shown connected by broken lines. Thus, when there are many peaks, the peak located in the outer shell is set as the representative point (object constituting the continuous tone region) (step S204).

  FIG. 13 shows the flow of object division processing by the second division method for the continuous tone area. In the second division method, the convex hull surrounding the “hem” of the tone histogram is obtained, the vertex of the convex hull is obtained, and the vertex is set as the representative point of the object. Specifically, a tone histogram of an image is created (step S221), a region having a predetermined density or more is extracted from the region (step S222), and a convex hull of each extracted region is defined (step S223). The vertexes of the hull are extracted as objects (step S224), and the stress characteristics between the extracted objects are set to be uniform (same stress characteristics between the objects).

  In either division method, the actual image information is continuous tone, so there is no clear boundary between objects in the image, and the application of color conversion to an actual image can be performed with a clear boundary. Absent. Therefore, as shown in FIG. 14, a weighting coefficient corresponding to the distance between the processing target image information and each object is added to the color change of a plurality of objects in the vicinity of the real image information, and the weighting coefficient and the conversion of the object are converted. Color conversion of the processing target image information may be performed based on the value.

  Next, additional stress setting will be described.

  When stress characteristics are defined only between objects, the point at which the total sum of stresses is lowest may not be uniquely determined. When the color gamut (second color space) of the output device is very large with respect to the color distribution of the object (when the total sum of stresses is 0 no matter where it is placed), or as a simple polygon In some models, there may be some arrangement patterns that reduce stress.

  In such a case, as shown in FIG. 15, between the primary color points 45a to 45f of the editing environment (first color space 41) and the chromaticity point (initial color) 44 of the target object itself in the editing environment. If a stress characteristic (indicated by a bidirectional arrow in the figure) is defined, a point 43 at which the total sum of stresses is minimum even under the above conditions is stably (uniquely) in the second color space 42. ) You will be able to ask. In addition, the stress generated with respect to the distortion generated between the primary color points 45a to 45f and the chromaticity point 44 of the target object itself may be small.

  Next, processing for dividing an image into objects will be described with reference to FIG. First, an image to be divided (image data) is acquired (step S301), and whether or not the image data has already been constructed with an object structure (for example, various patterns and fonts arranged by presentation creation software). It discriminate | determines (step S302). If it is structured (step S302; Yes), each structured area is set as an object, and each area is labeled (step S303). It is determined whether there is an area that needs to be further structured in addition to the structure already formed. If there is no need for further structure (step S304; No), the process proceeds to step S307.

  When the image is not constructed with the object structure (step S302; No) or when it is necessary to further structure (step S304; Yes), the contour is extracted based on the edge information for the processing target image (step S304). In step S305, a closed region is detected and labeled, and an excessively segmented region is arranged as necessary to define an object. (Step S306). Then, if the connection and inclusion relation of each object, and the image to be processed are already constructed in the object structure, the tree diagram (for example, FIG. 4) is obtained from the grouping definition, drawing pasting, layer structure information, etc. Hierarchies and connection relationships (object structures) are defined (step S307).

  The method of dividing into objects is not limited to the above method, and a general arbitrary region extraction process may be used.

  Next, a specific method for the process of determining the stress characteristic will be described. FIG. 17 shows an outline of the process for determining the stress characteristics. Prior to the processing, object division and tree diagram creation (object structure specification) are performed by the method described above. In the process of determining the stress characteristics, points are calculated for multiple connected objects (between objects connected by related lines) using multiple evaluation parameters such as the presence or absence of the boundary of the boundary and the contact state of the objects, and the stress is based on the total number of points. Define properties (stress intensity against strain).

  Specifically, any two connected objects in the tree diagram of the object structure are selected as evaluation targets, and the relation strength between the evaluation target objects is evaluated (step S401). Subsequently, a basic point corresponding to the attribute of each object is assigned (step S402). For example, 3 points are given as basic points for photographs, 5 points for graphs, and 7 points for titles. The value of the basic point given to the attribute of each object is set in advance and stored in a non-illustrated non-volatile memory or the like of the image processing unit 16.

  Next, it is checked whether or not there is a contour (contour line) at the boundary between objects to be evaluated. If there is a contour (step S403; Yes), points are assigned according to the strength of the contour (step). S404). The strength of the contour is evaluated by, for example, the thickness of the contour line, color, line type, and the like.

  Next, when there is contact between the objects to be evaluated (step S405; Yes), a score corresponding to the length of the contact distance is given (step S406). The longer the contact distance, the higher the score because it is necessary to ensure the discriminating power between the evaluation target objects.

  If the object to be evaluated is in the relationship between the character and the background (step S407; Yes), a score is assigned according to the character attributes (existence of decoration, font type, line thickness, etc.) (step S408). The lower the discriminability between the background and the characters, the higher the score. Here, in the case of decorative characters, in view of the fact that the character color of the color close to the background is often used because of the premise of ensuring distinctiveness due to texture etc., a high score is given for decorative characters. Yes.

  Next, if there is a color relationship between the objects to be evaluated (step S409; Yes), a score corresponding to the type of the color relationship is given (step S410). Here, 1 point is assigned for black and white, 3 points for palette color, and 5 points for full color. In the case of a full color, there is a possibility that a color approximate to the palette color may be used. Therefore, the score is increased so that a change in color tone due to color conversion hardly occurs.

  Thereafter, the points given by each evaluation parameter are summed (step S411), and a stress characteristic corresponding to the summed point is set between the objects to be evaluated (step S412).

  It is checked whether or not the above evaluation has been completed for all objects having a connection relationship. If the evaluation has not been completed (step S413; No), the evaluation target is switched to the next object combination (step S414) and step S401. Perform the following processing. When the evaluation is completed for all the objects having the connection relationship (step S413; Yes), the processing is ended (END).

Note that the evaluation parameter is not limited to the above, and there may be any number, and a score to be added may be determined as appropriate according to the characteristics of the parameter. For example, the following items may be evaluated.
(1) Internal contrast of the object; in this case, the higher the contrast, the smaller the stress, that is, the lower the contrast, the higher the number of points to be given.
(2) Specificity of the object; for example, the stress of an object with high visual discrimination, such as a face pattern, may be small (reducing or reducing the number of points given in the case of a face pattern).
(3) Concatenated relationship: When an object is in contact with a large number of objects, it is considered that the distinguishability of the object surrounded by various objects is enhanced, so that the stress may be small. That is, when there are few objects to contact, the number of points to be given is increased, or the points are reduced as the number of objects in contact increases.
(4) Object form; for example, the stress of a shape with a vertex or a shape with high independence may be small.

  It may be configured to be able to select which of a plurality of prepared evaluation parameters is used. For example, multiple types of evaluation parameter groups including multiple evaluation parameters are set, and any evaluation parameter group is used according to the document pattern (business report / IR for investor / document type such as conference presentation) and operator preference. It may be configured to be selectable. Further, it is preferable that the user can set an arbitrary evaluation parameter or evaluation parameter group.

  Further, a pattern having a predetermined characteristic such as a face pattern is automatically extracted, and this is set as one object, and a constraint condition is automatically set for the object so that a change in color tone of the object is limited. May be. For example, a color area for a “human face” is set in advance, and a barrier stress characteristic that does not deviate from the color area for a “human face” is automatically set for an object extracted as a face pattern. May be configured.

  Next, how to obtain device color space information will be described. The color space of devices such as the first image output device 14 and the second image output device 17 can be acquired using a general-purpose ICC profile creation tool and a colorimeter.

Further, for example, it can be simply obtained and used by the following method.
(A) Based on the characteristics of primary colors (for example, Y (yellow), M (magenta), C (cyan), K (black)), the density addition law and the Neugebauer equation (see JP 2003-276340 A) And approximate the outer gamut of the color gamut.
(B) In a screen projection type projector, it is obtained from a model in which white flare (indoor scattered light reflected from a white screen) is added to the current RGB work.

  Next, a method for minimizing the total sum of stress will be described. There are a plurality of objects in one image, and various stress characteristics are defined between them. The condition for minimizing the total sum of stress may be obtained by a known mathematical, physical, or analytical method.

  FIG. 18 shows a method for obtaining an optimal solution that minimizes the sum of stresses by trial and error from the initial conditions. The outline of the method is as follows. Color conversion is performed according to the conventional example, and the total stress is obtained in that case. Next, the converted value of the object is moved in the direction of the stress vector, and the total sum of stress is obtained again. This is repeated, and it is assumed that a practical optimum solution is obtained when the sum of stresses becomes sufficiently small or when the amount of decrease in the sum per cycle becomes small.

  More specifically, color conversion is performed by any method conventionally used (step S501), and the amount of distortion between objects caused by the color conversion is calculated (step S502). Then, based on the stress characteristic defined between the objects, the stress corresponding to each strain amount is calculated (step S503). A total sum of these stresses is obtained, and a decrease amount from the total sum of stresses obtained last time is calculated (step S504). Note that the calculation of the reduction amount passes the first time.

  When the amount of decrease is not less than the predetermined value or for the first time (step S505; No), the color tone is slightly changed in the direction in which the stress is applied to each object (step S506), and the process returns to step S502. The direction of movement is calculated assuming that the barrier or color gamut boundary slides with a friction coefficient of zero.

  If the amount of decrease is equal to or less than the predetermined value (step S505; Yes), the process is terminated assuming that an optimal solution has been found.

  However, in the above processing, when there are many local minimum values in the sum of stresses, there are cases where the minimum value (optimum value) cannot be found by being trapped by these. Therefore, it is also effective to escape from the minimal trap by techniques such as “preparing some initial values” and “giving random changes in color change with a certain probability”.

  FIG. 19 shows the flow of processing when the above improvements are made. One condition is selected from a plurality of initial conditions (step S521). Next, normal color conversion (which may be any known method) is performed using the initial conditions (step S522). The amount of distortion between the objects generated by the color conversion is calculated (step S523), and the stress corresponding to the amount of distortion is calculated based on the stress characteristic defined between the objects (step S524). The total sum of these stresses is obtained, and the amount of decrease from the total sum of stresses obtained last time is calculated (step S525). The first time the amount of decrease is passed.

  When the amount of decrease is not less than the predetermined value or for the first time (step S526; No), the color tone is slightly changed in the direction in which the stress is applied to each object (step S527). Note that the direction of movement is calculated assuming that the barrier or color gamut boundary slides with a friction coefficient of zero. In addition, the color tone is randomly changed at a rate of once per predetermined number of times (step S528), and the process returns to step S523.

  When the decrease amount is equal to or smaller than the predetermined value (step S526; Yes), the initial condition is changed (step S530) except when the conversion result becomes appropriate or sufficiently repeated (step S529;; No). ), The process returns to step S521, and the process is continued. If the conversion result is appropriate or has been sufficiently repeated (step S529; Yes), this processing ends (end).

  In the process of FIG. 19, the final solution may not match the best solution (the best solution is obtained in the middle of trial and error). For this reason, although the description is omitted in FIG. 19, the optimum solution during trial and error is always retained, and after repeated sufficiently, the retained optimum solutions are compared to obtain a better optimum solution. It has become.

  As described above, as a result of performing color conversion so that the total sum of stresses is minimized, the stress may be zero in some cases, but the stress may remain. Of these remaining stresses (which are referred to as residual stresses), a large one causes the color tone relationship between objects on which the residual stress acts to deviate, and this causes a decrease in discrimination. Therefore, when residual stress equal to or greater than a predetermined threshold remains between the objects, the object is decorated by an amount corresponding to the residual stress to ensure the necessary discrimination.

  There are two types of decorations: those that give outlines to objects and those that add visible textures to object areas. Which one to use is preferably selected with or without the outline of the original object region.

  It should be noted that since the area identification by color difference and the area identification by texture and contour are visually different, they cannot be identified, but the observation conditions (approximate observation distance, environment and image brightness) can be specified. Under the conditions, a texture having discriminability equivalent to that of patterns (objects) having a color difference ΔE can be statistically determined by sensory evaluation.

  FIG. 20 is an example of a graph obtained by sensory evaluation of the relationship between the color difference ΔE and the strength of the contour and texture structure. The strength of the texture structure is, for example, the magnitude of the difference in shading of the texture pattern. Note that the effects of the texture and the contour are greatly affected by the visual contrast of the structure in the observation state, and the effect is greatly impaired when the structure is excessively miniaturized (the spatial frequency is high). For this reason, it is preferable to estimate the viewing conditions of the image and use a texture or contour having a pattern of effective spatial frequency characteristics.

  FIG. 21 shows a flow of processing for converting the residual stress into decoration. As advance preparation, the relationship between the color difference and the decoration pattern as shown in FIG. 20 is acquired by sensory evaluation (step S600). Subsequent processing is performed on an object having a residual stress of a predetermined threshold or more. First, calculate the color difference (distortion) that should be eliminated from the required amount of residual stress and the stress characteristics defined between the objects for a set (two) of objects that are the targets for elimination of residual stress. (Step S601). Further, the structure of the object (the relationship between the contour and the strength of the internal texture) is evaluated (step S602). If the object is evaluated as “contour is strong” (step S603; Yes), the texture having the strength corresponding to the color difference to be eliminated is obtained from the graph of FIG. The texture is added to one of the objects to be eliminated (step S605), and the process is terminated. When it is not evaluated that “the contour is strong” (step S603; No), the contour of strength corresponding to the color difference to be resolved is obtained from the graph of FIG. 20 prepared in advance, and the obtained contour is determined for the object to be eliminated. Assigned to one (step S604), the process is terminated.

  Since decoration affects all objects related to the object to be decorated, in principle, the one with the smaller number of stress characteristics defined for the object (the one with fewer related objects), or It is reasonable to apply to the spatially contained one.

  Next, processing for receiving an additional instruction (conversion option setting) from the operator regarding color conversion will be described.

  In the color conversion according to the present invention, when the change in the color tone of the object becomes large, a message is displayed to the operator to request an additional instruction. FIG. 22 shows a flow of processing for receiving an additional instruction regarding color conversion from the operator. After determining the color tone for color conversion so as to minimize the total sum of stresses using the method of the present invention described above (step S621), the maximum value among the color change amounts generated by the color conversion is a predetermined allowable value. It is determined whether or not (step S622). That is, it is determined whether or not the amount of change in the tone of an arbitrary object due to color conversion or the distortion generated between any two objects exceeds a predetermined threshold (allowable value).

If it is equal to or smaller than the allowable value (step S622; No), the process ends (end). When the allowable value is exceeded (step S622; Yes), the monitor device 14 receives a message 51, a simulation image 52 obtained as a result of color conversion, and an additional instruction regarding color conversion from the operator as illustrated in FIG. An instruction input screen 50 including an operation button 53 for displaying is displayed (step S623). In the case of the instruction input screen 50 shown in FIG. 23, as additional instruction options (conversion options),
(1) Restrict to the same color name range,
(2) Emphasize the original color tone,
(3) Maintain the maximum color tone,
These three options are displayed.

  If the operator wants to keep the color tone of the editing environment, for example, if there is an object whose color tone change is to be suppressed, the operator clicks on the area of the simulation image 52 to select it, selects a conversion option with a radio button, and presses the re-conversion button 53a To do.

  When the above operation is performed (step S624; Yes), a constraint condition corresponding to the selected conversion option is set for the selected object (step S625), and the process proceeds to step S621. The color conversion is performed again so that the sum of stresses is minimized. When the continue button 53b is operated (step S624; No), the process of receiving an additional instruction from the operator ends (end). In this case, the output by the second image output device 17 is executed in the current color conversion state (the state of the simulation image 52).

  For example, when the conversion option “restrict to the same color name range” is selected on the instruction input screen 50 in FIG. 23, as shown in FIG. 7, barriers corresponding to the color regions 25 and 26 for each color name (see FIG. 7b) is set as a constraint condition. 23, when the conversion option “focus on the original color tone” is selected on the instruction input screen 50 in FIG. 23, the chromaticity point (initial color) in the first color space of the object is set for each object. A stress characteristic is set between them, and a color difference from the initial color is suppressed. When the “maximum color tone maintenance” conversion option is selected, the stress characteristic is set so that a greater stress acts on the initial color.

  Conversion options (conditions relating to color conversion) received from the operator are not limited to those exemplified, and other types of conditions may be used.

  As described above, when the amount of change in the color tone of an arbitrary object resulting from the color conversion or the distortion between any two objects exceeds an allowable value, whether or not the color conversion is possible or a condition change related to the color conversion is changed to the operator. Since the inquiry is made, the color conversion unintended by the operator is prevented. In addition, since conversion options are listed on a screen such as the instruction input screen 50 and a selection instruction is received, it is possible to easily perform color conversion reflecting the operator's instruction, and a device with a very large characteristic difference. Even in the color conversion between them, it is possible to easily create an effective document.

  Next, a case where a multi-page image is processed will be described. When an image file is composed of images of a plurality of pages, it is desirable to perform color conversion in consideration of the relationship between pages. For example, when there are a plurality of presentation materials, a pattern having the same meaning (for example, a company logo mark) may appear repeatedly. In such a case, it is also preferable to add a constraint condition so that repeatedly appearing patterns (similar objects) are maintained in the same color tone.

  FIG. 24 shows the flow of processing when color-converting an image of a plurality of pages. With respect to the object for one page, the operation of defining the object structure and the stress characteristic (step S701) by the above-described method of the present invention is repeatedly performed for all the pages until completion (step S702; No). When all pages have been completed (step S702; Yes), next, a set of similar objects (objects having a relation strength greater than or equal to a predetermined level) is extracted from the images of the plurality of pages to be processed (step S703). A stress characteristic is set between similar objects to which it belongs (step S704). Since there is a possibility that a plurality of similar objects exist in a series of image groups, the processes of steps S703 and S704 are performed on all sets of similar objects. That is, if steps S703 and S704 have not been completed for all sets of similar objects (step S705; No), the process returns to step S703.

  When the setting of stress characteristics for all sets of similar objects is completed (step S705; Yes), the color of the object is adjusted so that the total sum of stresses against distortion is minimized for all pages of the object. (Step S706), and the process ends (END).

  As a result, the stress works between similar objects belonging to the same group, so that the color tones of similar objects approximate. Note that it is desirable to set stress characteristics between similar objects so that a larger stress acts between similar objects than between other objects.

  The color conversion method of the extracted similar object is not limited to the above method, and an appropriate method may be selected and used. For example, there are the following other modes. Note that the process shown in FIG. 24 corresponds to another mode by replacing the process of step S703 with the process (J) shown below and replacing the process of step S706 with the process (K) shown below. It becomes.

<Another aspect 1>
(J) A stress characteristic is imparted between the extracted object and the initial color.
(K) For each page, the color tone is adjusted so that the total sum of stresses is minimized.
Note that it is desirable to set the stress characteristic between the initial colors so that a larger stress is applied than the stress characteristic between other objects.

<Another aspect 2>
(J) Define a barrier near the extracted object.
(K) For each page, the color tone is adjusted so that the total sum of stresses is minimized.

<Another aspect 3>
(J) A process of minimizing the sum of stress is performed for each page, and a barrier is provided in the vicinity of the color tone after conversion of the object of the page having the largest sum.
(K) For each page, the color tone is adjusted so that the total sum of stresses is minimized.

There are the following methods for extracting similar objects.
(1) The same element of a design in which the same pattern is repeated is a similar object. For example, it is possible to determine that the edit history information is acquired and the same element or the copied and pasted element of the same template that has not been additionally adjusted has higher similarity.
(2) The operator may be able to instruct the same element that should be particularly noted.
(3) A color object that repeatedly appears at a similar timing in a design in which the same pattern is repeated can be judged highly similar even if its shape is different.
(4) Extract unique patterns such as faces, pencils, and pointing marks, and estimate that patterns appearing contiguously are similar. For example, even when the same pointing mark 61 as shown in FIG. 25 repeatedly appears and the shapes of the “balloons” 62 and 63 attached to the respective pointing marks 61 are different, the color tone of the “balloon” is unified. In this case, these “balloons” 62 and 63 are set as a set of similar objects.

  FIG. 26 shows an example of the flow of similar object extraction processing. When receiving designation from the operator (step S721; Yes), the designated plurality of objects are set as similar objects (step S722). Note that the similarity (relevance strength) of the object designated by the operator is temporarily evaluated, and a warning message may be displayed when the combination is impossible.

  If there is history information on the process of creating the image or document data to be processed (step S723; Yes), the object generated by the combination of copy and paste is searched, and the one having the same gradation value is extracted from among them. Then, a set of similar objects is set (step S724).

  When the template is used (step S725; Yes), the template having the same component requirement and the same gradation is extracted and set to a set of similar objects (step S726).

  Further, the remaining objects are grouped based on the pattern characteristics (step S727). Then, one group is extracted, and when the similarity (relevance strength) is high in both form and color tone, it is set as a set of similar objects (step S728). When the evaluation in step S728 has not been completed for all groups (step S729; No), the evaluation target is switched to the next group (group) (step S730), and the process proceeds to step S728. When the evaluation in step S728 is finished for all groups (step S729; Yes), the process is finished (end).

  It should be noted that a set of similar objects may be detected by general pattern matching based on the form and the text, and a method for detecting similar objects may be arbitrarily determined.

  Further, the stress characteristics set for each set of similar objects may determine the magnitude of the stress according to the similarity (relevant strength). For example, when a perfect match such as a logo mark is required, the stress characteristic is set so that an extremely large stress acts. Compared to this, the stress is set to be smaller than that of the logo mark for a similar object that allows a certain degree of difference such as a balloon. The relationship between these objects and the stress to be set may be prepared as a preset table for a plurality of possible cases, for example.

  FIG. 27 schematically shows the result of color conversion by the method of the present invention. FIG. 27A shows a state in which the image P shown in FIG. 3 is divided into objects, the stress acting between the objects is represented by a solid line between the objects, and this is arranged in the first color space M1 of the editing apparatus. (Corresponding to FIG. 28 (b)). A small region and weak stress are not shown. FIG. 27B shows a second image output device 17 having the image P represented as shown in FIG. 27A in the first color space M1 by the color conversion method according to the present embodiment. 2 shows a state in which color conversion is performed into the second color space M2.

  In the color conversion method according to the present embodiment, since color conversion is performed by defining a stress characteristic with respect to the relationship between objects, the color tone relationship between objects given strong stress characteristics, that is, a thick solid line in the figure The color conversion is performed so as to maintain the color tone relationship between the objects connected in step (b), and the color conversion is performed as shown in FIG. In addition, as in the relationship between the large region B and the other large regions A and C, a large distortion occurs between objects having weak stress characteristics, but as described above, the relationship has a small influence on the explanation effect of the material ( In this way, it is possible to selectively arrange distortions between objects that do not easily affect discrimination even if there is a color difference.

  Compared to the color conversion shown in FIGS. 28C and 28D, the change in color tone between related objects (the middle areas in the large area A and the middle areas in the large area B) is suppressed to a small level. An image that effectively uses the second color space M2 (high saturation) is obtained.

  As described above, in the color conversion method according to the present embodiment, based on the structure of a plurality of objects arranged in the color conversion target image, the color difference between objects having a significant color relationship is preferentially maintained. Therefore, even if the image editing environment (first color space) and the display environment (second color space) are different, effective image printing is not required without requiring special image readjustment. And projection, and presentations can be made with materials that have the intended effect.

  The effect of the present invention occurs when the number of objects is three or more. In addition, since the effect of the present invention is more apparent when many relationships between objects can be defined, efficient image processing is realized when the number of objects is three or more. In addition to the number of objects, the condition of “perform the color conversion of the present invention when the relationship between defined objects is greater than or equal to a predetermined value (for example, 3)” or “between objects with defined relationships” The condition of “perform color conversion of the present invention when the type of stress characteristic is 2 or more” may be used, or a condition in which these are combined (“and” or “or”) may be used.

  Although the embodiments of the present invention have been described with reference to the drawings, the specific configuration is not limited to these embodiments, and the present invention can be changed or added without departing from the scope of the present invention. include.

  For example, the color conversion is performed so that the total sum of stresses is minimized, but if the total sum of stresses is made smaller than that obtained by color conversion by the conventional method without optimization until the total amount is minimized, It is possible to obtain a certain degree of effect, that is, an effect of suppressing a change in color tone that affects the distinguishability (or maintaining a meaningful color tone relationship).

  In the embodiment, the stress characteristics are determined based on the object structure. However, the stress characteristics are determined based on other factors, for example, the designation of stress by the user, the designation of the color relationship to be maintained, or the like. May be.

  In other words, the stress characteristic corresponds to setting the degree of association between objects. In other words, the stronger the stress is, the more the color relationship is maintained, so that the color conversion by the color conversion becomes smaller between the objects having the higher degree of association. Note that the degree of association does not have to be defined as a stress characteristic. The degree of association is set to several ranks, and color conversion is performed so that the amount of change in color due to color conversion decreases between objects for which the degree of association of the higher rank is set. That's fine. For example, an allowable amount of color change is determined for each rank, and the color arrangement of an object having a color change exceeding the allowable amount as a result of color conversion by a conventional method is moved so that the color change is reduced. It is also possible to repeat the method and adjust so that all the objects are within the allowable amount of the corresponding rank.

It is a block diagram which shows the structure of the image processing system which performs color conversion with the color conversion method which concerns on embodiment of this invention. It is a flowchart which shows the outline of the color conversion process which the image process part which concerns on embodiment of this invention performs. It is a front view which shows an example of an original image. It is explanatory drawing which shows the object structure of the original image shown in FIG. It is explanatory drawing which showed an example of the distortion produced by color conversion in the CIELAB color system. It is explanatory drawing which illustrated various stress characteristics. It is explanatory drawing which shows an example of the stress characteristic for making a color name barrier and a barrier. FIG. 5 is an explanatory diagram schematically illustrating a state in which stress characteristics are set between objects of the image P having the object structure illustrated in FIG. 4. It is a front view which shows the example of the original image containing a character, a photograph, a graph, etc. It is explanatory drawing which shows the object structure of the original image P2 shown in FIG. It is a flowchart which shows the object division process by the 1st division method regarding a continuous tone area | region. It is explanatory drawing which represented each peak detected from the histogram of the color tone of the continuous tone area | region on the color coordinate. It is a flowchart which shows the object division process by the 2nd division method regarding a continuous tone area | region. It is a flowchart which shows the real image processing method with respect to a continuous tone area | region. It is explanatory drawing which illustrated the state which defined the stress characteristic except between objects. It is a flowchart which shows the process which divides | segments an image into an object. It is a flowchart which shows the outline of the process which determines a stress characteristic. It is a flowchart which shows the process which calculates | requires the optimal solution from which the sum total of stress becomes the minimum. It is a flowchart which shows the process which calculates | requires the optimal solution from which the sum total of stress becomes the minimum. It is explanatory drawing which shows the relationship between color difference (DELTA) E and the strength of the structure of an outline and a texture. It is a flowchart which shows the process which converts residual capacity into decoration. It is a flowchart which shows the process which receives the additional instruction | indication regarding color conversion from an operator. It is a front view which shows an example of an instruction | indication input screen. It is a flowchart which shows the process in the case of carrying out color conversion of the image of several pages. It is explanatory drawing which shows an example (pointing mark from which the shape of a speech bubble differs) of a similar object. It is a flowchart which shows the extraction process of a similar object. It is explanatory drawing which shows the effect (The result which color-converted the image in 1st color space to 2nd color space by the color conversion method which concerns on this Embodiment). It is explanatory drawing which shows the example color-converted by the conventional method.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Image processing system 11 ... Image input part 11a ... Communication part 11b ... Imaging part 11c ... Scanner part 12 ... Image storage part 13 ... Editing process part 14 ... Monitor apparatus (1st image output device)
DESCRIPTION OF SYMBOLS 15 ... Instruction input part 16 ... Image processing part 17 ... 2nd image output device 17a ... Printer 17b ... Projection type display 21 ... Stress characteristic with respect to distortion of luminance direction 22 ... Stress characteristic with respect to other components 25 ... Color name "orange" 26 color area corresponding to the color name "red" 31 ... background of large area A 32 ... photograph 33 ... stress characteristic 41 ... first color space 42 ... second color space 43 ... sum of stress 44 ... The chromaticity point 45a to 45f of the target object itself ... Primary color point 50 ... Instruction input screen 51 ... Message 52 ... Simulation image 53 ... Operation button 53a ... Re-convert button 53b ... Continue button 61 ... Pointing mark 62, 63 ... balloon M1 ... first color space M2 ... second color space P, P2 ... original image

Claims (15)

In a color conversion method for color-converting an original image that exists in a first color space and includes a plurality of objects into an image in a second color space,
The difference between the color difference between two objects in the first color space and the color difference between the two objects after color conversion into the second color space is used as a distortion, and the force acts in a direction to eliminate the distortion. For each of the objects constituting the original image, the characteristic of the stress with respect to the distortion generated between the other objects is determined, and the color conversion is performed so that the total sum of the stresses becomes small. A color conversion method characterized by the above. The color conversion method according to claim 1, wherein the stress characteristic is determined based on an object structure of the original image. The color conversion method according to claim 1, wherein the color conversion is performed so that the sum is minimized. As a result of the color conversion, when the amount of change in the color tone of an arbitrary object or the distortion between two arbitrary objects exceeds a predetermined threshold, the user is inquired about whether or not the color conversion is possible or a condition change related to the color conversion. The color conversion method according to claim 1, wherein: 5. The color conversion method according to claim 1, wherein, when the color conversion is performed, a change in color tone for a predetermined object is limited within a predetermined color region. The color conversion method according to claim 1, wherein the stress characteristic is non-linear. The stress is divided into the stress related to the luminance or brightness component and the stress related to the other component, and the characteristics of each stress are set so that the stress related to the brightness or brightness component is larger than the stress related to the other component. The color conversion method according to claim 1, wherein: The color conversion method according to any one of claims 1 to 7, wherein a decoration corresponding to the residual stress is applied to an object whose residual stress after the color conversion is equal to or greater than a predetermined value. . The color conversion method according to any one of claims 1 to 8, wherein a setting of a restriction condition related to color conversion of a specific object is received. The color conversion method according to any one of claims 1 to 9, wherein an object belonging to the specific pattern is detected, and a predetermined constraint condition is set for color conversion of the detected object. When the color conversion is performed on the first image and the second image, the object constituting the first image has a relation strength higher than a predetermined value with any object constituting the second image. 11. The color conversion method according to claim 1, wherein when there is an object, a predetermined constraint condition is set for the color conversion of the object having a relation strength higher than a predetermined level. The constraint conditions are: (1) limiting the color change of the object within a predetermined color region; (2) the color of the object in the first color space and the color in the second color space after color conversion; The difference is set as a distortion, and the stress for the distortion is set. (3) The difference between the color determined based on the predetermined definition and the color in the second color space after the color conversion of the object is set as the distortion. The color conversion method according to claim 9, wherein the color conversion method is at least one of the following. In a color conversion method for color-converting an original image that exists in a first color space and includes a plurality of objects into an image in a second color space,
A color conversion method characterized in that a degree of association is set between each object constituting the original image, and the color conversion is performed such that a change in color difference between objects becomes smaller between objects having a higher degree of association. The color conversion method according to claim 13, wherein the relevance is determined based on an object structure of the original image. The color conversion method according to any one of claims 1 to 14, wherein the color conversion is executed only when the number of objects constituting the original image is three or more.