Research on Optimization Recognition Method of Digital Image Target Point Based on Machine Vision

Abstract

In order to enhance the auto-focus detection ability of digital images exposed by a single strong light source, an optimized recognition method of digital image target points based on machine vision tracking learning is proposed. Establishing a single strong light source exposure digital image feature point enhancement detection model, carrying out feature matching of the single strong light source exposure digital image under information enhancement technology, establishing a single strong light source exposure digital image three-dimensional reconstruction model, constructing a fuzzy feature detection algorithm of the single strong light source exposure digital image, and carrying out RGB decomposition of the single strong light source exposure digital image through fast low illumination image feature point identification feature matching, The spatial matching function of digital image under fast low illumination image feature point recognition is obtained. Under the machine vision tracking recognition model, fast low illumination image feature point recognition information fusion is carried out. And combined with spatial visual information enhancement method, the matching filter detection of digital image exposed by single strong light source is carried out. Through wavelet feature decomposition and enhanced information method, the target point optimization recognition of digital image exposed by single strong light source is carried out, and the signal-to-noise ratio of digital image exposed by single strong light source is improved. The results show that this method can be used to identify the target points of digital images exposed by high single strong light source.