Visual element recognition based on profile coefficient and image processing technology

Abstract

A method based on the contour coefficient and image processing technology is proposed to better identify the visual elements. This article takes the contour of the image as the recognition feature, summarizes the methods of target contour feature extraction, contour shape representation, and similarity representation, and studies the processing methods of contour edge preserving and denoising, contour feature simplification and description methods, and contour matching methods. This problem can usually be solved by filling out a form. Generally, a simple iterative equation is given to express the direct relationship between the current table and the calculated table values. The dynamic programming algorithm of inner distance shape context, multi-scale convexity convexity, and triangle area representation finds the best sequence correspondence.