红外和可见光的匹配跟踪在军事、遥感等领域有着广泛的应用。针对灰度和图像特征存在比较大差异的红外和可见光图像，本文采用了最大互信息算法，结合形态学梯度和小波分解。互信息算法优点在于不需要对多模图像灰度间的关系做任何假设，不足之处在于它对图像空间信息的忽略而且计算时间较长。本文互信息结合多结构元的形态学梯度检测的图像边缘，可以使得图像匹配精度提高，还能改善局部极值的问题，再利用小波分解对图像进行压缩降低分辨率，可以减少互信息计算量。最后的实验数据表明在配准过程中互信息的计算速度得到了优化，匹配精度得到了提高，实现快速和精确匹配。- Infrared and visible light matching tracking has wide application in military, remote sensing and other fields. There is a relatively large difference for grayscale image features infrared and visible light images, this article uses the maximum mutual information algorithm, combined with morphological gradient and wavelet decomposition. Mutual information algorithm the advantage not need to make any assumptions about the relationship between the multi-mode image gray inadequacies that its image spatial information is ignored and the computation time is longer. This paper mutual information combined with the multi-structural element morphological gradient image edge detection, can make the image matching accuracy is improved, but also to improve the problem of local minima, and then take advantage of the wavelet decomposition to reduce the resolution of the image is compressed, can reduce the amount of mutual information calculated . Finally, the experimental data show that the mutual