Research on the application of multi-feature fusion in ceramic art design

In this paper, median filtering and linear stretching are used to preprocess the classified ceramic images, respectively, and the LoG operator is utilized to detect the edges of ceramic photos. Combining the Hough transform and TMOAA algorithm to measure the edge length, thickness, and diagonal dimension of ceramics to obtain the basic elements of ceramic art design. The FAST algorithm is used to make up for the shortcomings of the SIFT algorithm in real-time, to extract the features of the overall image of ceramics and the vessel image, and then combined with the thinking framework of ceramic art design to generate the best conceptual model of ceramic art design. Extract and detect the ceramic vessel image and carry out similarity rate calculation, quantify the enamel color characteristics to generate a color design scheme, calculate the color difference value, and evaluate the aesthetic value of ceramic art design works from the perspective of ceramic art design aesthetics. Porcelain enamel color difference ΔE = 2.0369, ΔE ≤ 5.00~6.00, in line with the color quality standards, indicating that the color design scheme generated based on the color semantic elements has a certain applicability. Evaluating the best conceptual model of ceramic art design by aesthetic standards, the works generated based on the seven ceramic patterns have different aesthetic values and different forms, which meet the various demands for ceramic artworks nowadays.