A novel multi-focus image fusion algorithm is proposed on the basis of stationary wavelet transform without reconstruction. The original images to be fused are firstly decomposed by stationary wavelet transform. The low-frequency coefficients and high-frequency coefficients can be obtained. The clarity of low-frequency coefficients is estimated by the energy of Laplacian and the decision map can be acquired by comparing the energy of Laplacian. For the high-frequency coefficients at different resolution and different direction, the absolute value of these coefficients are summed firstly, and then the local region energy is used as the feature to get the high-frequency decision map. The fused image can be obtained by combining the low-frequency decision map and high-frequency decision map. The simulated results demonstrate that the proposed algorithm can get clearer image when compared with the average method, the traditional wavelet-based algorithm and some modified algorithm. The objective evaluation indices value, such as Shannon, average gradient, space frequency, standard deviation and mutual information, are improved compared with the average method and the traditional wavelet-based algorithm. Especially, the mutual information value computed by our algorithm is 2.4 times compared with the algorithm of Ref [3]. Subjective observation and objective evaluation suggest that our algorithm is effective for multi-focus image fusion.