突出边缘特征的图像融合方法研究

Research on Enhancing Edge Feature Image Fusion Method

介绍了像素级图像融合及特征级图像融合方法,分析了两种融合层次的优点。在此基础上,提出了两种突出边缘特征的图像融合方法,两种方案均对人们感兴趣的边缘特征进行增强,并在融合过程中考虑了边缘提取时的假边缘和双边缘情况。实验结果显示,边缘增强方案1提高了边缘定位的准确性,边缘增强方案2有效解决了方案1中融合影像缺少空间细节信息的问题。两种方案有效地增强了融合后图像的边缘特征,有利于图像的后续判读及图像识别。

This paper introduced pixel-level image fusion method and feature-level image fusion method, and analysed the advantages of two level image fusion methods. On the basis of discussion, two enhancing edge feature image fusion methods were given, which enhanced the interested edge feature and avoid the occurrence of the false edges and double edges. The experiment showed that method improved the edge veracity and solved the bottleneck problem of the first method. Both methods were helpful for image recognition.