Design And Implementation Of Color Matching Website Based On GAN And Color Emotion

With the renovation of science and technology and the continuous improvement of life quality,people have higher requirements on the design of works.Obviously,the color scheme is the foundation of the design of the work,so its importance needs no further explanation.The beauty and depth of the design work interface depends on how to choose the color scheme by ourselves.To some extent,the existing online color matching websites provide users with some color matching schemes.These color matching websites,however,also have some defects that include the professionalism of tool color matching websites and the difficulty for users without professional color matching knowledge to get started.Specific color matching websites often adopt color matching uploaded and designed by professional designers,and users browse through scores and other methods for selective display.Accordingly,the process is timeconsuming and the number of color schemes is limited.The proposal of GAN satisfies the applications and requirements in many fields,and it introduces new impetus to these fields.Inspired by a series of successful applications,this paper proposes an automatic color matching algorithm based on deep convolution generating adversarial networks.This algorithm invariably generates so many excellent color matching schemes to solve the problem of limited number of color matching schemes in traditional color matching websites.However,there may be one or more colors in a set of color schemes that are not suitable for the needs of the design of the product interface in the process of choosing a color scheme.At this time,I hope that other diversified colors can not only be referenced and matched with the existing colors,but are complied with people's aesthetic standard.Then the idea of generating an adversarial network based on deep convolutional conditions to achieve a selected color scheme was emerged.In this algorithm,the condition variable c is defined as the number of selected colors.In addition,this paper systematically studies the relationship between color and emotional psychology,and quantitatively classifies color schemes based on the basic color system component range of the HSV model.Finally,the relationship between color emotions and color schemes is established,and the function of generating color schemes by using color and emotion keyword retrieval is realized.Additionally,this paper has conducted a special study on this test to verify whether the color scheme generated by emotional keyword retrieval can arouse the corresponding emotional resonance of users.According to the outcomes of the testers,statistics and analysis were performed.85.5%of people argued that the color matching was reasonable and comfortable,and 77.3%of the people's emotional feedback met the defined color scheme and corresponding emotional mapping.Therefore,the overall scheme has practical feasibility.Ultimately,the automatic color matching algorithm model and the locking and generating color matching algorithm model were trained on the nearly 20,000 color scheme data sets obtained through the crawler technology.And each function module of the color matching website has been specifically developed and implemented.