Image fusion effect evaluation based on fuzzy Choquet integration

Aiming at the problems of evaluating fusion effect of the fused images, an effect evaluation method based on fuzzy Choquet integration is proposed. Entropy of the fused images, mean of cross entropy, root-mean-square of cross entropy, mean gradient of the fused images are selected to form a set of single-factor evaluation measure. The membership functions and the fuzzy measure of single-factor evaluation measure are defined based on human knowledge. A comprehensive evaluation measure is obtained through integrating various single-factor measures with Choquet integration. The method integrates the information of multiple single-factor measures and introduces human knowledge into the evaluation for image fusion. In this way the evaluation is more comprehensive and objective. When the method is applied to evaluating the fusion effect of CCD and SAR images, its evaluation results are consistent with the theoretical analysis result and visual effect, which shows that the method is effective.