Automatic generation method of map color matching scheme for different emotional states

Abstract

The invention provides an innovative map color matching scheme automatic generation method oriented to different emotional states. According to the method, an objective function is constructed according to a color emotion quantization formula, a swarm intelligence optimization algorithm, color similarity, color distinction degree, kernel density and the like are adopted as constraint conditions, constraint search optimization is carried out on the objective function, and therefore a map color design scheme fitting specific emotions is obtained. And finally, fusing the emotional color scheme with the original map content to generate a map product with stronger emotional expressive force and richer infectivity. According to the method, the consideration of emotion is added into the design of map colors, so that the aesthetic concept of map making is expanded, the emotional relation between a user and geographic data is deepened, and the map is promoted to convey more vivid and perceptual information.

Description

Technical Field

[0001] The invention relates to the fields of color science and computer graphics, and in particular to a method for rapidly and randomly generating map color schemes with different emotions by using a computer. Background Art

[0002] Map color is the basic visual variable of the map. Matching or distinguishing colors is very important for obtaining map information. An excellent map color scheme can promote users' understanding of spatial information and reduce users' cognitive load. At the same time, with the popularization of smart devices, users have higher and higher requirements for personalized and emotional experience of device interfaces. Color has an important impact on people's emotions and feelings. Different colors will bring different emotional experiences to people.

[0003] Existing map color optimization technology still has limitations. Most of the coloring methods in the existing technology only consider visual perception factors, and have not yet considered the emotional characteristics contained in color. If we can further explore the relationship between color and emotion on the basis of existing methods, it will not only make the color matching of the map more coordinated and moderate, but also create a specific emotional atmosphere through color, arouse people's resonance, and enhance the expressiveness and appeal of administrative division maps.

[0004] Therefore, how to automatically generate colors that are close to the user's current emotional state has become a problem that needs to be solved urgently. Summary of the invention

[0005] In order to solve the problem that the existing map coloring technology does not consider the emotional characteristics of color, the present invention provides a method for automatically generating map color schemes for different emotional states. The method generates corresponding map colors according to the different emotional states of the user, aiming to improve the effect of the map in conveying emotions.

[0006] The present invention discloses a method for automatically generating map color schemes for different emotional states, comprising the following steps:

[0007] S1 sets color value constraints, randomly generates color groups within the color value constraints, and constructs an objective function according to a color emotion quantification formula, wherein the randomly generated color groups are within a CIELAB color space;

[0008] S2 obtains the user's emotional state by predefining or using facial expression recognition technology, sets search constraints for the objective function according to the emotional state, and uses the firefly algorithm to search for the optimal solution of the objective function that meets the search constraints in the randomly generated color group; the emotional state includes a positive emotional state and a negative emotional state;

[0009] S3 uses the optimal solution of the objective function obtained by the search as the actual color group, transforms the CIELAB color space into the HSB color space, and uses the transformed actual color group as the map color scheme to color the unprocessed original map to obtain the color emotion map under the current emotional state.

[0010] Furthermore, the objective function is to randomly generate the average value of the color emotion value of each color in the color group, and the color emotion value of a certain color is expressed as:

[0011]

[0012] In the formula, AP represents the emotional value of the color; L represents the brightness; a represents the red-green color; and b represents the yellow-blue color.

[0013] Furthermore, the color value constraint conditions include color similarity constraint, color distinction constraint and kernel density constraint.

[0014] Furthermore, the color similarity constraint is set as follows:

[0015] H min <ΔH(c1, c2) <H max

[0016] S min <ΔS(c1, c2) max

[0017] B min <ΔB(c1, c2) max

[0018] In the formula, ΔH represents the difference in hue angle between two colors, ΔS represents the difference in saturation between two colors, ΔB represents the difference in brightness between two colors, c1 and c2 are any two colors in the map color scheme, and H min , S min and B min are the minimum values ​​of hue, saturation and brightness of the global variables, H max , S max and B max They are the maximum values ​​of hue, saturation and brightness that control the global variables respectively.

[0019] Furthermore, the color distinction constraint is expressed as:

[0020] ΔE(c1,c2)>θ

[0021] Where ΔE represents the perceived difference between two colors, c1 and c2 are any two colors in the map color scheme, and θ is the color difference threshold.

[0022] ​​Furthermore, the Gaussian kernel function P(x) is used to represent the kernel density constraint. P(x) is the matching probability of color x in the CIELAB color space. The larger P(x) is, the higher the probability that color x matches the subject information. The formula is as follows:

[0023]

[0024] where x is a color in the CIELAB color space, x i is the i-th color in the color sample, xx i is the distance between color x and the i-th color sample, n is the number of color samples, h is the bandwidth, and K is the kernel function.

[0025] Furthermore, the search constraints in S2 are set according to the user's emotional state, including hue, saturation and brightness constraints;

[0026] If the acquired emotional state is negative, the global minimum hue angle constraint of the color is set to 30-35, and the maximum hue angle constraint is set to 90-100; the global minimum saturation constraint is set to 30-50, and the maximum saturation is set to 70-90; the global minimum brightness constraint is set to 30-50, and the maximum brightness is set to 70-90;

[0027] If the acquired emotional state is positive emotion, the global minimum hue angle constraint of the color is set to 30-35, and the maximum hue angle constraint is set to 90-100; the global minimum saturation is set to 50-70, and the maximum saturation is set to 90-100; the minimum brightness is set to 50-70, and the maximum brightness is set to 90-100.

[0028] Furthermore, the process of searching for the optimal color group using the firefly algorithm is as follows:

[0029] The color emotion calculation formula is set as the objective function, the brightness value of the firefly is set as the optimization objective function value, and the position of a firefly corresponds to a randomly generated color group;

[0030] Fireflies keep moving, updating their positions, and eliminating those that do not meet the search constraints. The distance a firefly moves depends on the attraction of brighter fireflies, and each time a firefly moves to a new position, the brightness of the firefly will be recalculated;

[0031] After many iterations, the fireflies finally gathered in a stable position, indicating that the objective function has found the optimal solution.

[0032] Furthermore, the firefly brightness is set as the optimization objective function value, specifically:

[0033] When the emotional state is positive, the firefly brightness represents the maximum value of the optimization objective function. When the firefly brightness converges to a stable maximum value, it means that the average color emotion value of the randomly generated color group that meets the search constraints is the largest at this time, and the optimal solution of the objective function reflects the color combination of positive emotions.

[0034] When the emotional state is negative, the firefly brightness represents the minimum value of the optimization objective function. When the firefly brightness converges to a stable minimum value, it means that the average color emotion value of the randomly generated color group that meets the search constraints is the smallest at this time, and the optimal solution of the objective function reflects the color combination of negative emotions.

[0035] Furthermore, the actual color group is converted into the HSB color space in step S3, and the steps are as follows:

[0036] First, convert the color in CIELAB space to XYZ color space. The formula is:

[0037]

[0038] Then convert the color in the XYZ space to the RGB color space, the formula is expressed as:

[0039] R=3.2406X-1.5372Y-0.4986Z

[0040] G=-0.9689X+1.8758Y+0.0415Z

[0041] B=0.0557X-0.2040Y+1.0570Z

[0042] Finally, the color in the RGB space is converted to the HSB color space. The formula is expressed as:

[0043] B=max(R,G,B)

[0044]

[0045] Compared with the prior art, the significant advantage of the present invention is that the map color scheme is emotionally enhanced without changing the shape of map symbols or adding additional annotations. The present invention combines the knowledge of color psychology and utilizes a swarm intelligence optimization algorithm to automatically generate a color scheme that matches the preset emotion according to the set constraints and objective functions, and combines it with the map to create a personalized visual effect that matches the emotional experience for users. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flowchart of a method for automatically generating emotional color schemes for maps with different emotional states;

[0047] Figure 2 The original map of the Singapore subway line selected in Example 1;

[0048] Figure 3 is a map colored with positive emotional colors automatically generated by the method of the present invention in Example 1;

[0049] Figure 4 is a map colored with negative emotional colors automatically generated by the method of the present invention in Example 1;

[0050] Figure 5 It is the original land use coverage map of the United States in 2019 selected in Example 2;

[0051] Figure 6 is a map colored with positive emotional colors automatically generated by the method of the present invention in Example 2;

[0052] Figure 7 This is a map colored with negative emotional colors that is automatically generated according to the method of the present invention in Example 2. DETAILED DESCRIPTION

[0053] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0054] like Figure 1 The figure shows a flow chart of a method for automatically generating map color schemes for different emotions provided by the present invention. The method will be described in detail in conjunction with a specific embodiment.

[0055] S1 sets color value constraints, randomly generates color groups within the color value constraints, and constructs an objective function according to a color emotion quantification formula, wherein the randomly generated color groups are within a CIELAB color space.

[0056] The color value constraints include color similarity constraints, color distinction constraints, and kernel density constraints. In this example, the color similarity constraint is implemented by specifying a specific hue angle, saturation value, and lightness value range to ensure that the generated colors meet the design requirements in these properties. The color distinction constraint is used to consider the difference or distance between colors to ensure that the generated colors are sufficiently different so that they can be clearly distinguished visually. The kernel density constraint is used to guide the distribution of data points in the color space to ensure that the generated colors are evenly distributed throughout the color space to avoid being too concentrated or dispersed. Together, these constraints ensure that the generated colors have good visual distinction and can meet design requirements.

[0057] Color similarity refers to the visual difference between two colors. According to the constraints of color similarity, ensure that the hue angle, saturation value, and lightness value of all colors are not too small or too large. The setting range is as follows:

[0058] H min <ΔH(c1, c2) <H max

[0059] S min <ΔS(c1, c2) max

[0060] B min <ΔB(c1, c2) max

[0061] In the formula, ΔH represents the difference in hue angle between two colors; ΔS represents the difference in saturation between two colors; ΔB represents the difference in brightness between two colors; c1 and c2 are any two colors in the map color scheme. min , S min and B min are the minimum values ​​of hue, saturation and brightness of the global variables. max , S max and B max They are the maximum values ​​of hue, saturation and brightness that control the global variables respectively.

[0062] The color discrimination constraint refers to the degree of difference between colors in a color group. The present invention uses the color difference index as a constraint for model optimization search, and measures the color discrimination by calculating the perceived difference of map colors. If a certain threshold is reached, the two colors are considered distinguishable. For the calculation of color difference, the CIEDE2000 color difference formula is used, as follows:

[0063]

[0064] ​​Where ΔL′ is the lightness difference between the sample and the standard; ΔC′ is the chromaticity difference between the sample and the standard; ΔH′ is the hue difference between the sample and the standard; k L , k C、 k H , is the weight factor, usually takes the value of 1; S L , S C , S H is a weighted function that depends on the average value of lightness and chromaticity; R T is a correction term for hue differences, depending on the chromaticity and hue of the sample and the standard.

[0065] The CIEDE2000 color difference formula includes compensation for neutral color, brightness, chroma and hue, and has high perceptual uniformity. Specifically, the color distinction constraint is expressed as:

[0066] ΔE(c1,c2)>θ

[0067] Where ΔE represents the perceived difference between two colors; c1 and c2 are any two colors in the map color scheme; and θ is the color difference threshold.

[0068] On the basis of the above, kernel density estimation is used to estimate the matching probability of map colors based on color samples, and the Gaussian kernel function P(x) is used to represent it. P(x) is the matching probability of color x in the CIELAB color space. The larger P(x) is, the higher the probability that color x matches the theme information. The function is as follows:

[0069]

[0070] where x is a color in the CIELAB color space, x i is the i-th color in the color sample, xx i is the distance between color x and the i-th color sample, n is the number of color samples, h is the bandwidth, and K is the kernel function.

[0071] The present invention adopts a universal quantification model of color emotion proposed by Lichen-Ou (2018), which is recorded in Universal models of colour emotion and colour harmony (Color Research&Application.43.10.1002 / col.22243.), and uses the color emotion quantification formula in the text to calculate the color emotion value. The color emotion quantification formula is as follows:

[0072]

[0073] In the formula, AP represents the color emotion value, L represents the brightness, a represents the red-green degree of the color, and b represents the yellow-blue degree of the color. The objective function is the average value of the color emotion value of each color in the randomly generated color group.

[0074] S2 obtains the user's emotional state by predefining or using facial expression recognition technology, sets the search constraints of the objective function according to the emotional state, and uses the firefly algorithm to search for the optimal solution of the objective function that meets the search constraints in the randomly generated color group; the emotional state includes a positive emotional state and a negative emotional state.

[0075] Optionally, the facial expression recognition technology introduces expression recognition technology based on a convolutional neural network model to achieve real-time measurement of the user's emotional state.

[0076] In this embodiment, the emotional states including positive emotion and negative emotion are artificially predefined, and the higher the AP value, the more positive the color emotion; the lower the AP value, the more negative the color emotion value. The specific search constraints are set as follows:

[0077] If the acquired emotional state is negative, the global minimum hue angle constraint of the color is set to 30-35, and the maximum hue angle constraint is set to 90-100; the global minimum saturation constraint is set to 30-50, and the maximum saturation is set to 70-90; the global minimum brightness constraint is set to 30-50, and the maximum brightness is set to 70-90;

[0078] If the acquired emotional state is positive emotion, the global minimum hue angle constraint of the color is set to 30-35, and the maximum hue angle constraint is set to 90-100; the global minimum saturation is set to 50-70, and the maximum saturation is set to 90-100; the minimum brightness is set to 50-70, and the maximum brightness is set to 90-100.

[0079] According to the color value constraints set by S1, a color group is randomly generated, and the generated color group is substituted into the color emotion quantification formula. The average emotion value between several generated colors is calculated to obtain the overall trend of the group of colors in emotional communication. The color emotion value of each color in the color group is calculated to compare the advantages and disadvantages of the color groups. The higher the color emotion average value of the color group, the more effective the color group can express positive emotions; the lower the color emotion average value of the color group, the more effective the color group can express negative emotions.

[0080] In order to speed up the search, the firefly algorithm is used. The firefly algorithm is an efficient swarm intelligence algorithm inspired by the flashing behavior of fireflies. The following is a further description of how to use the firefly algorithm to generate a qualified color group in the color space.

[0081] In this embodiment, 50 fireflies are first set in the search space, and the colors of all fireflies are randomly initialized, as shown in the following formula.

[0082] H=H min +rand(0,1)(H max -H min )

[0083] S=S min +rand(0,1)(S max -S min )

[0084] B=B min +rand(0,1)(B max -B min )

[0085] Where rand(0,1) represents a random value in the range of (0,1). min , S min and B min They are the minimum values ​​of hue, saturation and brightness, all set to 0. max , S max and B max They are the maximum values ​​of hue, saturation and brightness respectively. Set the maximum value of hue to 360, and the maximum values ​​of saturation and brightness to 100. Each firefly represents a randomly generated color group.

[0086] Initially, all fireflies are randomly distributed in the search space. The color emotion calculation formula is defined as the objective function to calculate the brightness of a single firefly. Assuming that a firefly satisfies the relevant constraints, its brightness is assigned the value of the objective function. As the fireflies move, the color groups that are difficult to distinguish emotions are replaced. The distance moved depends on the attraction of brighter fireflies and the position of the fireflies is updated every time the fireflies move. When the fireflies move to a new position, their brightness will be recalculated. Eventually, after many iterations, the fireflies will gather at the brightest position.

[0087] When the emotional state is positive, the firefly brightness represents the maximum value of the optimization objective function. When the firefly brightness converges to a stable maximum value, it means that the average color emotion value of the randomly generated color group that meets the search constraints is the largest at this time, and the optimal solution of the objective function reflects the color combination of positive emotions.

[0088] When the emotional state is negative, the firefly brightness represents the minimum value of the optimization objective function. When the firefly brightness converges to a stable minimum value, it means that the average color emotion value of the randomly generated color group that meets the search constraints is the smallest at this time, and the optimal solution of the objective function reflects the color combination of negative emotions.

[0089] S3 uses the optimal solution of the objective function obtained by the search as the actual color group, transforms the CIELAB color space into the HSB color space, and uses the transformed actual color group as the map color scheme to color the unprocessed original map to obtain the color emotion map under the current emotional state. This process includes the following steps:

[0090] First, convert the color in the CIELAB space to the XYZ color space. The color space conversion formula is as follows:

[0091]

[0092] Then convert the color in the XYZ space to the RGB color space. The color space conversion formula is as follows:

[0093] R=3.2406X-1.5372Y-0.4986Z

[0094] G=-0.9689X+1.8758y+0.0415Z

[0095] B=0.0557X-0.2040Y+1.0570Z

[0096] Finally, the color in the RGB space is converted to the HSB color space. The color space conversion formula is as follows:

[0097] B=max(R,G,B)

[0098]

[0099] Using the CIELAB color model in the CIE color system as a transition, the brightness value, yellow-blue saturation, and red-green saturation under this color model are converted into hue, chroma, and lightness component values ​​under the HSB color model. These three components describe the color through the HSB color group from the perspective of visual perception, forming a color space with relatively unified values ​​and perceptions, which is more advantageous in the recognition of color emotions.

[0100] After converting the actual target color group of CIELAB color space into HSB color space, the target color group is used as the map color scheme to color the unprocessed original map to obtain the map that best matches the target emotion.

[0101] The example map of Example 1 of the present invention selects the Singapore subway map, which has a simple structure and is suitable for displaying complex traffic information; the example map of Example 2 of the present invention selects the US land use coverage map, which covers a variety of geographical information and is directly related to people's quality of life and environmental perception. The two maps have clear information transmission and rich emotional cognitive potential, respectively. Using these two maps as example demonstrations, the emotional difference between the maps before and after optimization can be more clearly observed.

[0102] In Example 1, Figure 2 The Singapore subway line map shown in FIG. 1 is used as the original map, and is processed by the method proposed in the present invention to obtain the following Figure 3 The map of positive emotions shown in Figure 4 In Example 2, the negative sentiment map is shown. Figure 5 The land use coverage map of the United States shown in FIG. 1 is used as the original map, and is processed by the method proposed in the present invention to obtain the following Figure 6 The map of positive emotions shown in Figure 7 Map of negative sentiment shown.

[0103] By optimizing the color schemes of these maps, it is clearly observed that maps designed with emotional colors perform better than the original maps in terms of visual effects and emotional resonance.

Claims

1. A method for automatically generating map color schemes for different emotional states, characterized in that: The steps include: S1 sets color value constraints, randomly generates color groups within the color value constraints, and constructs an objective function according to a color emotion quantification formula, wherein the randomly generated color groups are within a CIELAB color space; S2 obtains the user's emotional state by predefining or using facial expression recognition technology, sets search constraints for the objective function according to the emotional state, and uses the firefly algorithm to search for the optimal solution of the objective function that meets the search constraints in the randomly generated color group; the emotional state includes a positive emotional state and a negative emotional state; S3 uses the optimal solution of the objective function obtained by the search as the actual color group, transforms the CIELAB color space into the HSB color space, and uses the transformed actual color group as the map color scheme to color the unprocessed original map to obtain the color emotion map under the current emotional state.

2. The color scheme automatic generation method according to claim 1, characterized in that: The objective function is to generate the average value of the color emotion value of each color in the randomly generated color group. The color emotion value of a certain color is expressed as: In the formula, AP represents the color emotion value of the color; L represents the brightness; a represents the red-green color; and b represents the yellow-blue color.

3. The color scheme automatic generation method according to claim 1, characterized in that: The color value constraints include color similarity constraints, color distinction constraints and kernel density constraints.

4. The method for automatically generating a color scheme according to claim 3, characterized in that: The color similarity constraint is set as follows: H min <ΔH(c1,c2)<H max S min <ΔS(c1,c2)<S max N min <ΔB(c1,c2)<B max In the formula, ΔH represents the difference in hue angle between two colors, ΔS represents the difference in saturation between two colors, ΔB represents the difference in brightness between two colors, c1 and c2 are any two colors in the map color scheme, and H min , S min and B min are the minimum values ​​of hue, saturation and brightness of the global variables, H max , S max and B max They are the maximum values ​​of hue, saturation and brightness that control the global variables respectively.

5. The method for automatically generating a color scheme according to claim 4, characterized in that: The color discrimination constraint is expressed as: ΔE(c1,c2)>θ Where ΔE represents the perceived difference between two colors, c1 and c2 are any two colors in the map color scheme, and θ is the color difference threshold.

6. The method for automatically generating a color scheme according to claim 5, characterized in that: The Gaussian kernel function P(x) is used to represent the kernel density constraint. P(x) is the matching probability of color x in the CIELAB color space. The larger P(x) is, the higher the probability that color x matches the subject information, which is expressed as: where x is a color in the CIELAB color space, x i is the i-th color in the color sample, xx i is the distance between color x and the i-th color sample, n is the number of color samples, h is the bandwidth, and K is the kernel function.

7. The method for automatically generating a color scheme according to claim 1, characterized in that: S2: the search constraints are set according to the user's emotional state, including hue, saturation and brightness constraints; If the acquired emotional state is negative, the global minimum hue angle constraint of the color is set to 30-35, and the maximum hue angle constraint is set to 90-100; the global minimum saturation constraint is set to 30-50, and the maximum saturation is set to 70-90; the global minimum brightness constraint is set to 30-50, and the maximum brightness is set to 70-90; If the acquired emotional state is positive emotion, the global minimum hue angle constraint of the color is set to 30-35, and the maximum hue angle constraint is set to 90-100; the global minimum saturation is set to 50-70, and the maximum saturation is set to 90-100; the minimum brightness is set to 50-70, and the maximum brightness is set to 90-100.

8. The method for automatically generating a color scheme according to claim 7, characterized in that: The process of searching the optimal color group using the firefly algorithm is as follows: The color emotion calculation formula is set as the objective function, the brightness value of the firefly is set as the optimization objective function value, and the position of a firefly corresponds to a randomly generated color group; Fireflies keep moving, updating their positions, and eliminating those that do not meet the search constraints. The distance a firefly moves depends on the attraction of brighter fireflies, and each time a firefly moves to a new position, the brightness of the firefly will be recalculated; After many iterations, the fireflies finally gathered in a stable position, indicating that the objective function has found the optimal solution.

9. The method for automatically generating a color scheme according to claim 8, characterized in that: The firefly brightness is set as the optimization objective function value, specifically: When the emotional state is positive, the firefly brightness represents the maximum value of the optimization objective function. When the firefly brightness converges to a stable maximum value, it means that the average color emotion value of the randomly generated color group that meets the search constraints is the largest at this time, and the optimal solution of the objective function reflects the color combination of positive emotions. When the emotional state is negative, the firefly brightness represents the minimum value of the optimization objective function. When the firefly brightness converges to a stable minimum value, it means that the average color emotion value of the randomly generated color group that meets the search constraints is the smallest at this time, and the optimal solution of the objective function reflects the color combination of negative emotions.

10. The method for automatically generating a color scheme according to claim 9, characterized in that: S3 converts the actual color group of the CIELAB color space into the HSB color space, and the steps are as follows: First, convert the color in CIELAB space to XYZ color space. The formula is: Then convert the color in the XYZ space to the RGB color space, the formula is expressed as: R=3.2406X-1.5372Y-0.4986Z G=-0.9689X+1.8758Y+0.0415Z B=0.0557X-0.2040Y+1.0570Z Finally, the color in the RGB space is converted to the HSB color space. The formula is expressed as: B=max(R,G,B)

Images