基于小波系数邻域特征的图像融合

在利用小波变换进行图像融合的基础上,研究了融合因子的选取方法。由于小波变换在时域和频域中同时具有良好的局部特性,为了很好地利用小波变换的这种特征,提出了利用小波系数的邻域特征(邻域方差)来定义融合因子的思想。评价融合算法的性能应该从融合图像的信息增加量和融合图像的失真度两个方面来评价,融合图像的熵用来描述融合图像的信息含量,相对熵可以描述融合图像的失真度,评价结果显示出其方法的实用性。实验表明该算法用于医学图像的融合能得到很好的效果。

Image fusion based on neighborhood features of wavelet coefficients

Selection of fusion-factors is essential for an image fusion process, and it is therefore studied by using wavelet transformation to do image fusion. In order to make better use of the excellent spatial-frequency characteristics of wavelet transformation, it is suggested to use the neighboring region features of wavelet coefficients, such as variance, to define the fusion factors. The performance of fusion algorithm should be evaluated by both the increase in information and the distortion degree of fused images, the entropy of fused image is used to express the amount of information contained in the image, and the crossing-entropy of fused image is used to express the distortion degree of fused image. Experimental results indicate that the algorithm proposed is excellent, and can be used to obtain very good fused medical images.