In this paper, an algorithm for multi-spectrum image fusion using weighted bi-orthogonal self-adaptive wavelet transform is put forward, which can make up for the defects that there are faintness of image details in multi-spectrum image fusion of lower contrast image. The self-adaptive method of wavelet coefficients local model maximum which are weighted is used to fuse the high frequency components and the syncretism adaptive method is also chosen to fuse low frequency coefficients. The capability of multi-spectrum image fusion is evaluated by calculating mean grads of image. The experimental results show that the fusion rule of our proposed method is more effective.