An Adaptive Image Fusion Method Based on Local Statistical Feature of Wavelet Coefficients

According to the characteristics of the low frequency and the high frequency coefficients of wavelet decomposition, a new adaptive image fusion method based on local statistical feature of wavelet coefficients is presented in this paper. For the low frequency coefficients, taking the local energy of the image as a criterion, an adaptive fusion rule of combining weighted average with selection is used to obtain the approximate coefficients. For the high frequency coefficients, on the basis of local variance and covariance, an adaptive weighted average method is used to obtain the detail coefficients. The image entropy and the cross entropy can be computed to evaluate the performance of the proposed method and other wavelet-based image fusion methods. Experiments show that the proposed method can achieve better effect than other methods in the human visual perception and objective performance evaluation.