基于模糊变换耦合最大熵的医学图像融合算法

针对融合医学图像在过渡区微小细节及边缘信息不够清晰,边缘容易失真等问题,提出了一种基于模糊变换耦合最大熵值的多模态医学图像融合算法.首先,将待融合图像划分为大小相同的非重叠图像块,引入模糊变换对得到的图像块进行处理得到模糊子块,并利用邻域最大熵融合规则对模糊子块进行融合,获取新的融合子块;然后,将新的融合子块进行逆模糊变换,利用选择最大值融合规则,将逆变换得到的子块进一步融合生成最终融合医学图像.实验结果表明:与当前医学图像融合算法对比,本文算法在主观评价与客观评价指标边缘强度、信息熵、互信息、峰值信噪比上具有更大的优势,其融合图像边缘更加清晰,细节丰富,克服了边缘模糊与伪轮廓,更能够有效完成医学图像融合,实现了多模态医学图像信息互补.

The Medical Image Fusion Algorithm Based on Fuzzy Transformation and Maximum Entropy

In order to solve the defects such as blurring small details and edge information of the fused image in the transition region as well as distorted edge, the medical image fusion algorithm based on fuzzy transformation and maximum entropy was proposed in this paper. Firstly, the preparing fused image was divided into non overlapping image blocks with the same size; secondly, the fuzzy subblocks was obtained by introducing the fuzzy transformation to process the image blocks, and the new fusion subblocks was got by using the maximum entropy fusion rule to fuse these fuzzy subblocks. Then the new fusion subblocks was transformed by inverse fuzzy, and final fusion medical image was got by selecting the maximum fusion rule to fuse the subblocks which were obtained by inverse fuzzy transformation. Experimental results show that this algorithm has excellent performance in subjective evaluation and objective evaluation as edge strength, information entropy, mutual information and peak signal to noise ratio, which the edge and details of the fused image was more clear to overcome the edge blur and the false contour comparison with common medical image fusion algorithms.