用于EVS的改进小波变换图像融合算法

摘要

为增强小波变换图像融合算法的实时性,提高视觉增强系统(EVS)可见光图像与红外图像实时融合的效率,提出了一种基于矩阵QR分解和小波变换的图像融合算法.该算法对原始图像的像素矩阵进行QR分解,再利用正交矩阵的性质,根据小波变换图像融合算法对QR分解得到的上三角矩阵进行分解融合,利用QR分解得到的正交矩阵逆变换得到融合图像.实验结果表明,该算法能获得较好的实时性,同时保证较好的融合效果.%To improve the real-time performances of wavelet transform image fusion algorithm and increase the real-time fusion efficiency of visible and infrared images in enhanced vision system (EVS), an algorithm based on matrix QR decomposition and wavelet transform image fusion algorithm is presented. Firstly, the images' pixel matrixes are decomposed with QR decomposition theory, and then according to the properties of orthogonal matrix, the upper-triangle matrixes obtained in QR decomposition are decomposed and fused with wavelet transform image fusion algorithm, finally fused image is obtained by inverse transform with orthogonal matrixes obtained in QR decomposition. Experiments show that the proposed algorithm has better real-time performances and maintains better fusion effect