A method for quantitative evaluation of architectural color combinations based on image classification and recognition

Abstract

The present invention provides a method for quantitatively evaluating architectural color combinations based on image classification and recognition, comprising the following steps: step (1). performing comprehensive imaging of a building or building complex to be analyzed; step (2). adjusting the acquired image; step (3). performing object recognition of the building or building complex components; step (4). performing color recognition; step (5). constructing a color space distribution image of the building or building complex components; step (6). performing color harmony analysis on the image of the spatial distribution of the building object colors; step (7). performing harmony evaluation on other newly added building colors; and step (8). performing sensory evaluation analysis on different building object color distribution combination color systems. The present invention conducts a comprehensive study on progressive coupling, qualitative and quantitative, dynamic and static association, and sensory engineering for traditional colors, providing practical and feasible guiding opinions for the inheritance and sustainable development of Huizhou colors.

Description

Technical Field

[0001] The present invention relates to the technical field of color analysis of buildings and building complexes, and in particular to a method for quantitatively evaluating building color combinations based on image classification and recognition. Background Art

[0002] Sensibility is the ability to feel, or rather, to acquire feelings. Judgments about perception are often based on time and space and involve value judgments. The concept of Kansei Engineering was proposed by scholars at Hiroshima University in Japan. It is defined as a comprehensive interdisciplinary discipline that uses modern technology to measure, quantify, and evaluate user perceptions. In recent years, the rapid development of computer technology, big data technology, and artificial intelligence has provided excellent hardware conditions and technical support for the effective study of spatial color in Huizhou's regional architectural complexes.

[0003] Color practices in rural areas are relatively limited. Rural color planning started relatively late, and compared to the complexity of urban color planning, rural areas primarily approach color from perspectives such as landscape perception, rural settlement protection, and the shaping of rural landscape characteristics. They focus on the harmony of rural architecture and natural colors, the preservation of cultural traditions, and community development, often conducted within the framework of tourism planning. Related practices primarily rely on a combined top-down and bottom-up approach, using color planning management clauses to limit and guide both the inheritance and development of color. On the one hand, color restrictions primarily rely on mandatory constraints and protection of color landscapes, prohibiting the disorderly use of colors; on the other hand, color guidance primarily encourages and innovates color landscapes, allowing for the orderly renewal of colors. Overall, color practices in regional architectural complex spaces primarily rely on color restrictions, supplemented by color guidance. The combination of these two approaches maintains the order and vitality of color. Summary of the Invention

[0004] The technical problem solved by the present invention is to provide a method for quantitatively evaluating architectural color combinations based on image classification and recognition, so as to solve the problems raised in the above background technology.

[0005] The technical problem solved by the present invention is achieved by the following technical solution: a method for quantitatively evaluating architectural color combinations based on image classification and recognition, comprising the following steps:

[0006] Step (1). Perform multi-angle comprehensive imaging of the building or building complex to be analyzed;

[0007] Step (2). Adjust the clarity and color of the acquired image;

[0008] Step (3) performing object recognition of the building or building complex components and structural recognition of the building or building complex on the processed image;

[0009] Step (4). Identify the color intensity, saturation, and hue of the building or building complex components in the processed image;

[0010] Step (5). Constructing an image of the spatial distribution of the colors of the components of a building or a building complex within the building or building complex through a three-dimensional model;

[0011] Step (6). Perform color harmony analysis on the image of the spatial distribution of the building object's color;

[0012] Step (7) Processing the spatial distribution data of the building object color after the harmony analysis to form a building object color distribution combination color system; Based on the generated building object color distribution combination color system, the harmony of other newly added building colors is evaluated;

[0013] Step (8). Conduct sensory evaluation analysis on the color distribution and combination of different building objects.

[0014] The method for adjusting the clarity and color of the acquired image is to acquire N different first images of the same scene, fuse the N different first images to obtain N corresponding second images, subtract the corresponding second images from the first images pixel by pixel to obtain a corresponding residual noise image, then fuse the residual noise image with the corresponding input image to obtain a first denoised image, and perform corresponding format conversion on the first denoised image to obtain a final denoised image.

[0015] The method for identifying objects of components of buildings or building complexes is to perform structural modeling on objects in an input image based on a random field structure model to obtain a structured expression of the object; based on the structured expression of the object, a gradient back propagation algorithm is used to learn structural parameters, solve the gradient, and a stochastic gradient descent algorithm is used for learning and training to obtain an object recognition model.

[0016] The structural recognition method of a building or a building complex is to realize the spatial layout of the component objects based on the recognition results of the component objects of the building or the building complex, and to form the structure of the building or the building complex by combining the recognition results with its physical spatial structure.

[0017] The wavelet model is used for processing in step (3), which includes the following steps:

[0018] 1). Constructing wavelet model:

[0019]

[0020] where scale c = 2 j , choose the smoothing function with variance σ 2Two-dimensional Gaussian function of ;

[0021] At this time, let k = -2*4 j *σ 2

[0022] The simplified wavelet model is:

[0023]

[0024] 2). Calculate the amplitude of the wavelet transform along the horizontal and vertical directions:

[0025]

[0026] Where m and n represent the size of the selected transformation window;

[0027]

[0028] 3). Determine the input and output membership functions;

[0029] 4). Establish identification rules;

[0030] 5). Refine the structural image;

[0031] 6). Identify and annotate the structural image.

[0032] The input-output membership function is:

[0033]

[0034] The range of the input variable is [0, 1].

[0035] The identification rules are:

[0036] If B1, B2, B3, and B4 are all in the middle, then pixel A5 is an edge point;

[0037] If B2, B3, and B4 are medium and B1 is low, then pixel A5 is an edge point;

[0038] If B1, B3, and B4 are medium and B2 is low, then pixel A5 is an edge point;

[0039] If B1, B2, and B4 are medium and B3 is low, then pixel A5 is an edge point;

[0040] If B1, B2, and B3 are medium and B4 is low, then pixel A5 is an edge point;

[0041] If B1, B2, B3, and B4 are all high, then pixel A5 is an edge point;

[0042] Other cases are non-edge points;

[0043] Among them, A1-A9 represent the grayscale values ​​of the image in the window, B1-B4 represent the grayscale values ​​obtained after the 3x3 window omnidirectional wavelet transform. 0 ,45 0 ,90 0 ,135 0 The amplitudes after wavelet transformation in 4 directions.

[0044] The refinement processing method is as follows: for all edge points A0, if the following conditions are met, it can be determined that it can be deleted: the number of edge points E(A0) in its 8-connected domain is 2≤E(A0)≤6; in the 8-connected domain of point A0, it only contains one 4-connected domain edge point.

[0045] For all deletable points, if one of the following conditions is met, the point is retained; otherwise, the point is deleted: A2 and A6 are edge points, while A4 is deletable; A4 and A8 are edge points, while A6 is deletable; A4, A5, and A6 are all deletable.

[0046] The color harmony analysis method in step (6) is to divide the color space distribution image into blocks, count the number of adjacent same-color blocks to calculate the distribution area of ​​the same-color blocks, calculate the proportion and distribution position of the same-color blocks in the overall image, and simultaneously calculate the color distribution relationship of the color intensity, saturation, and hue of different color blocks in the image; and perform quantitative evaluation based on the harmony evaluation model.

[0047] The sensory evaluation analysis in step (8) includes using an eye tracker, an SD questionnaire, and an electroencephalogram (EBW) test device to conduct collaborative experiments, collect experimental data on the visual, psychological, and emotional effects of color on subjects in architecture-related fields, interpret the visual, psychological, and emotional effects of color experience, and then conduct a correlation analysis on the visual, psychological, and emotional effects.

[0048] Compared with the existing technology, the beneficial effects of the present invention are: the present invention conducts a comprehensive study on progressive coupling, qualitative and quantitative, dynamic and static correlation and sensory engineering for traditional colors, and provides practical and feasible guiding opinions for the sustainable goals of the inheritance and development of Huizhou colors. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the original embodiment of the invention.

[0050] Figure 2 This is an example diagram after processing in step (2) of the invention.

[0051] Figure 3 This is an example diagram for processing step (3) of the invention.

[0052] Figure 4 This is an example diagram for processing step (4) of the invention.

[0053] Figure 5 Schematic diagram of the process of the invention.

[0054] Figure 6 This is a flow chart of the method of step (2) of the invention.

[0055] Figure 7 Schematic diagram of the invented wavelet model processing flow.

[0056] Figure 8 This is a flow chart of step (8) of the invention. DETAILED DESCRIPTION

[0057] In order to make the technical means for realizing the present invention, the creative features, the purpose and the effect easily understood, the present invention is further described below with reference to specific diagrams. In the description of the present invention, it should be noted that, unless otherwise clearly stipulated and limited, the terms "installation", "connection" and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection. It can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be internal communication between two components.

[0058] like Figure 1-4 As shown, a method for quantitatively evaluating architectural color combinations based on image classification and recognition includes the following steps:

[0059] Step (1). Perform multi-angle comprehensive imaging on the building or building complex to be analyzed; a single photo is easily affected by the objective environment and subjective experience, while a large number of photos and videos can objectively reflect the physical color properties of the material, reduce the objective and subjective influences, and further improve the accuracy of identification and evaluation through multi-angle image analysis. At the same time, the color of the same object can be compared to reduce color difference. At this time, the nature and color accuracy of the specific object can be improved by imaging objects such as doors, columns, plaques, couplets, trees, walls, stone carvings, bluestone slabs, and ancient roads separately; perform multi-angle comprehensive imaging on the building or building complex to be analyzed, and choose cloudy weather with good visibility. The color collection time should be selected between three hours after sunrise and five hours before sunset, that is, between 10:00 and 15:00. It is suitable for materials in various complex conditions, such as mottled walls, water flow, light and shadow, etc., but attention should be paid to the existence of image noise and corresponding noise reduction measures in the later analysis.

[0060] Step (2). Adjust the clarity and chromaticity of the acquired image; and perform manual color contrast adjustment by using a color matching instrument to obtain color and a color card to obtain color. By acquiring N different first images of the same scene, the N different first images are fused to obtain N corresponding second images. The corresponding second image is subtracted pixel by pixel from the first image to obtain a corresponding residual noise image. The residual noise image is then fused with the corresponding input image to obtain a first denoised image. The first denoised image is converted to a corresponding format to obtain a final denoised image. Figure 2 As shown, it can be further improved Figure 1 Excellent color accuracy can greatly reduce subsequent processing errors and improve analysis reliability;

[0061] Step (3). The processed image is subjected to object recognition of the building or building complex components and structural recognition of the building or building complex; the objects in the input image are structurally modeled based on a random field structure model to obtain a structural expression of the object; based on the structural expression of the object, the structural parameters are learned using a gradient back propagation algorithm, the gradient is solved, and the stochastic gradient descent algorithm is used for learning and training to obtain an object recognition model; the structural recognition method of the building or building complex is to realize the spatial layout of the component objects according to the result of the object recognition of the building or building complex components, and to combine the recognition result with its physical spatial structure to form the structure of the building or building complex.

[0062] It can also be processed using a wavelet model, which includes the following steps:

[0063] 1). Constructing wavelet model:

[0064]

[0065] where scale c = 2 j , choose the smoothing function with variance σ 2 Two-dimensional Gaussian function of ;

[0066] At this time, let k = -2*4 j *σ 2

[0067] The simplified wavelet model is:

[0068]

[0069] 2). Calculate the amplitude of the wavelet transform along the horizontal and vertical directions:

[0070]

[0071] Where m and n represent the size of the selected transformation window;

[0072]

[0073] 3). Determine the input and output membership functions;

[0074] The range of the input variable is [0, 1].

[0075] 4) Establish recognition rules; if B1, B2, B3, and B4 are all middle, then pixel A5 is an edge point;

[0076] If B2, B3, and B4 are medium and B1 is low, then pixel A5 is an edge point;

[0077] If B1, B3, and B4 are medium and B2 is low, then pixel A5 is an edge point;

[0078] If B1, B2, and B4 are medium and B3 is low, then pixel A5 is an edge point;

[0079] If B1, B2, and B3 are medium and B4 is low, then pixel A5 is an edge point;

[0080] If B1, B2, B3, and B4 are all high, then pixel A5 is an edge point;

[0081] Other cases are non-edge points;

[0082] Among them, A1-A9 represent the grayscale values ​​of the image in the window, B1-B4 represent the grayscale values ​​obtained after the 3x3 window omnidirectional wavelet transform. 0 ,45 0 ,90 0 ,135 0 The amplitudes after wavelet transformation in 4 directions.

[0083] 5). Perform thinning on the structural image; for all edge points A0, if the following conditions are met, it can be determined that it can be deleted: the number of edge points E(A0) in its 8-connected domain is 2≤E(A0)≤6; in the 8-connected domain of point A0, it only contains one 4-connected domain edge point.

[0084] For all deletable points, if one of the following conditions is met, the point is retained; otherwise, the point is deleted: A2 and A6 are edge points, while A4 is deletable; A4 and A8 are edge points, while A6 is deletable; A4, A5, and A6 are all deletable.

[0085] 6). Identify and annotate the structural image.

[0086] Step (4). Identify the color intensity, saturation, and hue of the building or building complex components in the processed image; identify the color intensity, saturation, and hue of the doors, columns, plaques, couplets, trees, walls, stone carvings, bluestone slabs, ancient roads, etc. after identification; and mark them accordingly; at the same time, use a color matching instrument to pick colors and a color card to compare colors, and perform manual color contrast adjustment;

[0087] Step (5). Construct an image of the spatial distribution of the colors of the components of a building or building complex through a three-dimensional model; by strengthening the spatial expression of color and weakening the expression of the building structure, the role of color in the spatial layout can be further improved; the role of color in the building or building complex can be better analyzed to provide a reference for the subsequent color matching of the building or building complex;

[0088] Step (6). Perform color harmony analysis on the image of the spatial distribution of the color of the building object; divide the color space distribution image into blocks, and count the number of adjacent blocks of the same color to calculate the distribution area of ​​the blocks of the same color, calculate the proportion and distribution position of the blocks of the same color in the overall image, and calculate the color distribution relationship of the color intensity, saturation, and hue of different blocks in the image; and perform quantitative evaluation based on the harmony evaluation model. The beauty calculation can be used to preliminarily screen suitable color schemes and color matching methods, which can provide more and more accurate samples for the later perceptual evaluation, reduce the number of unnecessary samples in the future, and improve the generation efficiency of the color system;

[0089] Step (7). Process the spatial distribution data of the building object color after the harmony analysis to form a building object color distribution combination color system; the overall color matching of the building or building complex can be formed, and the building object color distribution combination color system can be different according to the sample form of different regions and eras; the cultural migration history of different regions and the color system change history of the same region can be explored based on the building object color distribution combination color system, so as to better analyze the local cultural changes; based on the generated building object color distribution combination color system, the harmony of other newly added building colors is evaluated; several basic color samples are established, using the RGB color system, so that the R, G, and B values ​​are all changed in 64-bit unit intervals, and the established several basic color samples are used as comparison templates for the scheme;

[0090] Calculate the correlation between the color scheme and the color sample. Each color scheme Xi (Ri, Gi, Bi) has a corresponding upper and lower adjacent color X in the set basic color sample. 01 (R 01 ,G 01 ,B 01 ) and X 02 (R 02 ,G 02 ,B 02); calculate X respectively i ,X 01 ,X 02 , the correlation between them, namely γ(X 01 ,Xi) and γ(X 02 ,Xi).

[0091] Xi,X 01 ,X 02 , the correlation calculation method between them is:

[0092]

[0093]

[0094] Where ζ is the discrimination coefficient, and the value of ζ ranges from [0,1].

[0095] The target color image evaluation value is solved, and the image vocabulary is obtained through the questionnaire summary, and the relative image evaluation value e of the sample and the image vocabulary is obtained. I :

[0096]

[0097] Where: a1, a 02 For Plan X 01 and X 02 Image evaluation weight coefficient; e 01 ,e 02 For Plan X 01 and X 02 Image evaluation value.

[0098] Step (8). Conduct a perceptual evaluation analysis on the color distribution and combination of different architectural objects; the perceptual evaluation analysis includes the use of eye trackers, SD questionnaires, and brain wave testing equipment to conduct collaborative experiments, collect experimental data on the visual, psychological, and emotional effects of color systems on subjects in architecture-related fields, interpret the visual, psychological, and emotional effects of color experience, and then conduct a correlation analysis on the visual, psychological, and emotional effects. The subjects freely browse dynamic and static scene information, and coordinate with the relevant audio-visual environment and temperature and humidity environment. At the same time, the eye tracker is used to accurately track and record the eye movement data of different subjects browsing the virtual scene, and then analyze the visual focus position and its change trajectory, etc., to explore the relationship between visual perception and color information. At the same time as the eye movement test, the fast Fourier transform method is used to convert the subject's brain wave data, and the analysis software is used to extract the data reflecting the person's comfort. Using multivariate analysis methods such as dispersion analysis and multiple regression analysis, we analyzed the differences in brain waves during the dynamic and static changes in the Huizhou architectural complex space, thereby understanding the impact of various parameters of the complex environment on the subjects' physiological level. In conjunction with the SD method survey and analysis after the experiment, we conducted psychological evaluations on the subjects and compared them with the results of eye movement and brain wave analysis to explore the relationship between psychological and physiological quantities in the evaluation of complex color environments, thereby verifying the physiological evaluation results of color value experience.

[0099] Perceptual evaluation analysis establishes a perceptual evaluation between subjective understanding and objective reality to analyze the value experience of Huizhou colors. Therefore, only research based on perceptual engineering can effectively grasp the emotion of color and achieve the analysis of value experience. Perceptual engineering is a young but highly mature discipline. If we want to capture specific, individual experiences as universal scientific experience, we need to capture perceptuality in the most intuitive form of time and space, transforming empirical data into the subject of scientific theoretical investigation. Similarly, the perceptual experience that color brings to people is concentrated in visual, psychological, and physiological perception.

[0100] Due to the strict requirements of EEG experiments and eye movement experiments on the experimental environment, we should first strictly control the light environment, audio-visual environment, and physical environment (temperature and humidity) of the collaborative experiment based on the existing experimental environment. In particular, the light environment should meet the brightness of the eye tracking of the eye tracker, and should also meet the impact of changes in the light environment of the EEG experiment on EEG analysis. Combined with the SD psychological experiment, we will finally study the impact of changes in the various characteristic indicators of Huizhou colors on human vision, psychology, and emotion at the physiological level.

[0101] Based on the aforementioned color cognitive results from both static and dynamic perspectives, we used eye trackers, EEG testing equipment, and SD questionnaires to collect and organize the visual attention, psychological tendencies, and EEG waves of subjects in architecture-related fields in response to different color effects. During the experiment, the choice of static and dynamic images should be carefully considered, ensuring that they reflect the characteristics of spatial elements and avoid artifacts such as the photographic angle that could interfere with the experimental results and affect the relationship between color and element information. In particular, when using EEG testing equipment and eye trackers in collaborative experiments, it is important to ensure that the lighting, physical, and audiovisual environments meet the requirements of both EEG and eye tracker testing. When measuring the emotional effects of participants in different complex environments, it is important to increase the diversity of these complex environments to more clearly reflect the differences in brainwaves.

[0102] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for quantitatively evaluating architectural color combinations based on image classification and recognition, characterized by: The following steps are involved: Step (1): Perform multi-angle comprehensive imaging of the building or building complex to be analyzed; Step (2). Adjust the clarity and color of the acquired image; Step (3). Perform object recognition of the building or building complex components and structural recognition of the building or building complex on the processed image; Step (4). Identify the color intensity, saturation, and hue of the building or building complex components in the processed image; Step (5). Constructing an image of the spatial distribution of the colors of the components of the building or building complex within the building or building complex through the three-dimensional model; Step (6): Performing color harmony analysis on the image of the spatial distribution of the building object's colors; Step (7). Process the spatial distribution data of the building object colors after the harmony analysis to form a building object color distribution combination color system; based on the generated building object color distribution combination color system, evaluate the harmony of other newly added building colors; Step (8). Conduct sensory evaluation analysis on the color distribution and combination of different building objects.

2. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 1, characterized in that: The method for adjusting the clarity and color of the acquired image is to acquire N different first images of the same scene, fuse the N different first images to obtain N corresponding second images, subtract the corresponding second images from the first images pixel by pixel to obtain a corresponding residual noise image, then fuse the residual noise image with the corresponding input image to obtain a first denoised image, and perform corresponding format conversion on the first denoised image to obtain a final denoised image.

3. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 1, characterized in that: The method for identifying objects as components of buildings or building complexes is to perform structural modeling on objects in an input image based on a random field structure model to obtain a structured expression of the objects; Based on the structured expression of the object, the gradient back propagation algorithm is used to learn the structural parameters, solve the gradient, and use the stochastic gradient descent algorithm for learning and training to obtain the object recognition model.

4. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 3, characterized in that: The structural recognition method of a building or a building complex is to realize the spatial layout of the component objects based on the recognition results of the component objects of the building or the building complex, and to form the structure of the building or the building complex by combining the recognition results with its physical spatial structure.

5. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 1, characterized in that: The wavelet model is used for processing in step (3), which includes the following steps: 1). Construct wavelet model: where scale c=2 j , choose the smoothing function with variance σ 2 Two-dimensional Gaussian function of ; At this time, let k=-2*4 j *σ 2 The simplified wavelet model is: ； 2). Calculate the amplitude of the wavelet transform along the horizontal and vertical directions: , Where m and n represent the size of the selected transformation window; ； 3). Determine the input and output membership functions; 4). Establish identification rules; Refine the structural image; 6). Identify and annotate the structural image.

6. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 5, characterized in that: The input-output membership function is: ; The range of the input variable is [0, 1].

7. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 5, characterized in that: The identification rules are: If B1, B2, B3, and B4 are all in the middle, then pixel A5 is an edge point; If B2, B3, and B4 are medium and B1 is low, then pixel A5 is an edge point; If B1, B3, and B4 are medium and B2 is low, then pixel A5 is an edge point; If B1, B2, and B4 are medium and B3 is low, then pixel A5 is an edge point; If B1, B2, and B3 are medium and B4 is low, then pixel A5 is an edge point; If B1, B2, B3, and B4 are all high, then pixel A5 is an edge point; Other cases are non-edge points; Among them, A1-A9 represent the grayscale values ​​of the image in the window, B1-B4 represent the grayscale values ​​obtained after the 3x3 window omnidirectional wavelet transform. 0 ,45 0 ,90 0 ,135 0 The amplitudes after wavelet transformation in 4 directions.

8. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 7, characterized in that: The refinement processing method is as follows: for all edge points A0, if the following conditions are met, it can be judged that it can be deleted: the number of edge points E(A0) in its 8-connected domain is 2≤E(A0)≤6; the 8-connected domain of point A0 contains only one 4-connected domain edge point; for all deletable points, if one of the following conditions is met, the point is retained; otherwise, the point is deleted: A2 and A6 are edge points, and A4 is deletable; A4 and A8 are edge points, and A6 is deletable; A4, A5, and A6 are all deletable.

9. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 1, characterized in that: The color harmony analysis method in step (6) is to divide the color space distribution image into blocks, count the number of adjacent same-color blocks to calculate the distribution area of ​​the same-color blocks, calculate the proportion and distribution position of the same-color blocks in the overall image, and simultaneously calculate the color distribution relationship of the color intensity, saturation, and hue of different color blocks in the image; and perform quantitative evaluation based on the harmony evaluation model.

10. The method for quantitatively evaluating architectural color combinations based on image classification and recognition according to claim 1, characterized in that: The sensory evaluation analysis in step (8) includes using an eye tracker, SD questionnaire, and brain wave testing equipment to conduct collaborative experiments, collect experimental data on the visual, psychological, and emotional effects of color on subjects in architecture-related fields, interpret the visual, psychological, and emotional effects of color experience, and then conduct correlation analysis on the visual, psychological, and emotional effects.

Images