Research on Image Fusion Algorithm for Video Optimization

Abstract

This paper aims to study video image fusion through two types of image fusion algorithms: pixel level and feature level, which can solve camera shake during video shooting, image blur caused by interference, target occlusion. Based on the above proposed methods, a video is optimized by wavelet transform fusion algorithm, Laplacian pyramid fusion algorithm and the scale-invariant feature transform (SIFT) fusion algorithm, respectively. Moreover, three indexes are used to evaluate the images obtained by using these two types of fusion algorithms. A human-computer interaction interface is constructed by utilizing MATLAB toolbox, and the simulation of video image fusion is completed in the interface, which proves that the fusion algorithm can achieve a good result than another two algorithms.