基于灰色关联分析和IHS变换的图像融合算法

New and Better Image Fusion Method Based on Grey Relational Analysis and IHS Transform

针对IHS融合算法的光谱畸变问题,提出一种基于灰色关联分析和IHS变换的图像融合算法。该算法利用灰色关联分析法检测出SAR图像像素中的边缘点和非边缘点;通过IHS变换,在线性加权融合时赋予边缘点较大的权值。实验表明,该算法在保持光谱信息方面具有显著的优越性,并且能够有效提高融合图像的空间分辨能力,具有简洁性和抗噪性,明显优于传统的IHS融合算法和改进的基于直方图匹配的IHS融合算法。

A better image fusion method was proposed based on grey relational analysis and IHS transform to fuse the TM multipectral image and high-resolution SAR image, Firstly, grey relational analysis was used to classify pixels in the SAR image into non-edge pixels and edge pixels, Secondly, by using IHS transform, component and SAR image were fused with different weighting strategies, whose weight for edge pixels was higher than that for non-edge pixels, The experimental result indicates that the new method is compact and able to resist noises better, and it enhances spatial detail information of the fusion image efficiently and preserves spectral information of the original TM image much better.