The performance of the image enhancement method in the spatial domain is limited by free parameters and global gray mapping functions. From the perspective of image fusion, an image enhancement algorithm based on subimage fusion is proposed to solve this problem. Firstly, the adaptive image enhancement is regarded as a process of image fusion and the enhanced images which are derived from enumeration are as the fusion images. Secondly, the weights of fusion images are estimated according to the information of details, saturation and luminance of fusion images. Eventually, the enhancement image is fused according to the estimated weights. Moreover, the proposed method is used to improve some conventional image enhancement methods in the spatial domain, e.g. piecewise linear transformation, Gamma correction, contraststretching transformation. The experimental results show that the proposed method is able to improve the visual effect of image enhancement in the spatial domain.