基于Contourlet变换的多传感器图像融合

Multi-sensor Image Fusion Based on Contourlet Transform

提出了一种新的基于Contourlet变换的融合多传感全色和多光谱影像的算法;由于Contourlet变换具有良好的多方向性和多尺度,所以比其他方法更适应于进行多传感图像融合,对于Contourlet变换后得到的低频和高频分量系数,采用平均法选择低频区域系数,选择区域"能量"较大的高频系数作为融合影像的高频系数;实验结果表明,基于所提出的Contourlet变换的融合结果,优于其他常用的融合方法。

This paper proposes a novel method of fusing panchromatic and multispeetral multl-sensor images using eontourlet transform. contourlet transform can offer a much richer set of directions and shapes, and thus it is more effective than other in fusing multisenser images. The images are firstly decomposed by contourlet transform, use average method to choose the low frequency eontourlet coefficients, use energy method choose high frequency contourlet coefficients. The experimental results show the fused image produced by the eontourlet transform is of better quality than that obtained through the other transform.