Curvelet, as a new multiscale analysis algorithm, is more appropriate for the analysis of the image edges such as curve and line characteristics than wavelet, and it has better approximation precision and sparsity description. When the curvelet transform is introduced to image fusion, the characteristics of original images are taken better and more information for fusion is obtained. The proposal of the second generation curvelet theory makes it under-stood and implemented more easily. Then the second generation curvelet transform based image fusion method is proposed. Firstly, the source images are decomposed using curvelet transform, then the curvelet coefficients are fused with the fusion regular in the corresponding scales, and finally the fused coefficients are reconstructed to obtain fusion results. Multi-focus images are taken as experimental data, mean square error, difference coefficient, correlation coefficient are used to evaluate the results, and comparison with results based on wavelet transform is also carried out . The results show that when decomposition level is 2, this method gets fusion result of similar quality with wavelet, but for other decomposition levels this method gets much better fusion results than wavelet.