Shoe Sole Contour Extraction Based on Two-dimensional Maximum Entropy Threshold Segmentation

Aiming at the problem of low efficiency and accuracy of the glue application process in the shoemaking process, propose a method of shoe sole contour extraction based on improved maximum entropy threshold segmentation. First, transform the original image into a grayscale image, and use the fusion median filtering and wavelet transform denoising algorithm to remove the influence of noise in the image. Secondly, use the fast two-dimensional maximum entropy threshold segmentation algorithm to threshold the denoised image. Finally, uses the Sobel operator to extract the final edge, and the experiment verifies the effectiveness of the method proposed in this paper.