基于非下采样Contourlet变换的多模态医学图像融合

A enw method of multimodality medical image fusion based on the nonsubsampled Contourlet transform

基于非下采样Contourlet变换,对配准后的原始图像进行非下采样Contourlet分解后,采用区域能量比加权法进行低频分量的融合,用平均梯度和局部区域能量相结合的方法来制定高频分量。实验结果证明,本方法可以更多地继承原始图像特征,细节信息更饱满。

Based on non-subsampled contourlet transform,the original registered image is dealt with non-subsampled Contourlet decomposition and then with low-frequency component fusion by means of regional energy ratio weighting method.The high-frequency coefficients of the fused image are determined by combining average gradient with local energy.Experimental results show that the method may ensure the fused image to inherit more original detailed features.