An image fusion algorithm based on local energy processing of wavelet coefficient is analyzed. To overcome its disadvantage of insufficient consideration for local pixel independent information,an improved algorithm is proposed. The improved algorithm combines the local energy selection and weighted average fusion rule by using the phase information of window center pixel. When the high frequency wavelet coefficients of the two source images are the same in phase,maximum rule of local energy is used to select the high frequency wavelet coefficient of the fusion images. When the two coefficients are reversal in phase,weighted average rule of local energy is used to obtain the coefficient. Experiment results show the effectiveness of the improved algorithm.