基于非下采样剪切波变换的医学图像融合算法

为了解决单一模态医学图像的局限性,提出了一种基于非下采样剪切波变换(NSST)的多模态医学图像融合方法.该方法利用NSST将待融合的医学图像分解成低频系数和高频系数,并利用区域能量加权(WLE)的方法对分解后的低频系数进行融合,使用区域能量和平均梯度加权的方法对分解后医学图像的高低频系数进行融合,采用NSST逆变换重建融合后的图像.选择信息熵、平均梯度和空间频率3个参数作为融合图像的客观评价参数,结果表明,该方法取得的融合结果比离散小波、轮廓波和非下采样轮廓波变换等传统方法更好,计算效率更高.

Medical image fusion algorithm based on non-subsampled shearlet transform

In order to solve the limitation of single mode of medical images, a multi mode medicine image fusion method based on non subsampled shearlet transform （NSST） was proposed. In this method, the medical images to be fused were decomposed into low frequency and high frequency coefficients with the NSST. With the weighted local energy （WLE） method, the low frequency coefficients after decomposition were fused. With the local energy and average gradient weighted method, the high and low frequency coefficients of medical images after decomposition were fused. In addition, the fused images were reconstructed by the inverse NSST. Three parameters including the information entropy, average gradient and spatial frequency were selected as the objective evaluation parameters of fused images. The experimental results show that the fusion effect of the proposed method is better than that of such traditional methods as discrete wavelet transform, contourlet transform and non subsampled contourlet transform （NSCT）, and the proposed method has higher computational efficiency.