Art Intelligence System (AIS) Based on Edge Information Visualization and Image Reconstruction
Jun Hui Wu · 2025
With the continuous improvement of the application of computer image processing technology in complex scenarios, how to perform intelligent analysis of artistic images has become the mainstream of research. Once again, this study proposes an art intelligent system based on edge information visualization and image reconstruction. The proposed AIS framework consists of three core modules: (1) image texture feature detection module. This module adopts local three-value mode (LTP) and pixel density analysis to accurately describe image textures to accurately analyze features; (2) Image edge detection module. This module combines anisotropic diffusion filtering and improved Canny operator to improve edge detection accuracy while suppressing noise; (3) Art image reconstruction and enhancement module. This module realizes semantic reconstruction and image quality optimization through semi-autoregressive language decoder. The experiment used a self-built dataset containing 10,000 high-resolution art images and visualized and quantitatively analyzed it. The quantitative analysis part evaluates the algorithm performance through indicators such as the number of edge pixels, edge contrast measurement and peak signal-to-noise ratio. Experimental displays that the improved Canny operator is better than the traditional Sobel operator in terms of edge pixel count (6602 vs. 2503) and EBCM value (109.8928 vs. 100.3524), proving the effectiveness of the proposed algorithm.