联合多判据的图像融合算法

Image Fusion Algorithm by Combining Multiple Criteria

常见的基于小波变换的图像融合方法只是从小波系数的某一特性的单一角度来考虑小波系数的选取,因而只适合图像中某一特性的融合.为寻找一种有效的小波系数融合方法,从数据融合的角度出发,提出了一种有效的融合算法.该算法采用投票表决法来联合小波系数选取中的多个判据,并综合考虑小波系数的选取.实验表明,与单一小波变换系数融合方法相比,本方法能有效地提高融合图像的质量.

As common image fusion schemes based on wavelet consider a certain aspects of the fusion task when fusing wavelet coefficients of the source images, it is most possible that these methods each feeds single property of the fused image. To counterbalance these methods, a new fusion scheme was proposed. This new method adopts the majority voting to synthesis plural decision criteria to select the wavelet coefficients for fusion. The experiments show this method is effective compared with the single fusion method in the quality of fusion result.