摘要： 针对遥感影像融合中出现的对比度差、边缘模糊等问题，利用支持向量机的统计学习优势，结合方向滤波器组的多方向特性，提出一种基于支持向量机的遥感影像融合框架。由于小波核相对其他核函数对复杂的信号具有更好的逼近能力，因此利用Morlet母小波构造核函数，研究基于小波支持向量机的遥感影像融合，提高遥感影像融合的精确度。实验结果表明，该方法的融合效果优于传统的图像融合方法。

关键词: 图像处理, 图像融合, 遥感影像, 小波支持向量机

Abstract: Aiming at some problems such as poor contrast and edge blur in remote sensing image fusion, by using the statistical learning advantages of Support Vector Machine(SVM) and combining with the multi- directional characteristics of filter bank, this paper proposes a framework for remote sensing image fusion based on SVM. Compared with other kernel functions, wavelet kernel function has better approximation ability on complex signals. So in this paper, through kernel function constructed by Morlet mother wavelet function, remote sensing images fusion based on wavelet SVM is researched. It can improve the accuracy of remote sensing image fusion. Experimental results show that the fusion effect of the method is better than traditional image fusion methods.

Key words: image processing, image fusion, remote sensing image, wavelet Support Vector Machine(SVM)

中图分类号: 

• TN911.73