图像融合技术以其综合多传感器信息的优越性日益受到诸多领域的重视。为了使其应用在医学、遥感、计算机视觉、气象预报、军事目标识别等方面更迅速、深入地开展，有效、实用的融合算法是至关重要的。本文在小波变换金字塔结构的基础上，提出了一种多尺度多算子融合方法，对热红外图像和可见光图像的融合进行了研究。结果融合效果很好，目标和背景区别显著，而且边缘不突兀。由于这种方法对小波分解的层数要求不高，因此计算量不大，便于并行处理及硬件实时化实现，具有广阔的应用前景。 Image fusion technology is increasingly valued by many fields due to its superiority of synthesizing multi-sensor information. In order to make its application more rapid, intensive and effective, practical fusion algorithm in the fields of medicine, remote sensing, computer vision, weather forecast and military target recognition, it is very important. Based on the wavelet transform pyramid structure, a multi-scale and multi-operator fusion method is proposed in this paper. The fusion of thermal infrared images and visible images is studied. The results of the fusion effect is good, the difference between the target and the background significant, and the edge is unobtrusive. Because of this method, the wavelet decomposition layer is not demanding, so the calculation is not large, easy to parallel processing and hardware real-time implementation, has broad application prospects.