In the process of obtaining high dynamic range(HDR) image using the fusion of multiple shot images, the selection of exposure time in traditional method is blind, which makes the image information redundant and thus affects the fusion efficiency. In this paper, a method of multi-exposure control based on maximum local information entropy was proposed. The relationship between information entropy and exposure time of low dynamic scene was discussed. It was concluded that the image information entropy of a low dynamic range scene increased first and then decreased with the increase of exposure time. And information entropy achieved the maximum at a certain exposure time. For a high dynamic range scene, firstly, the range of exposure time was determined by using the approximate linear relationship between the gray level of the image and the exposure time. Secondly, the high dynamic range scene was divided into several low dynamic range(LDR) regions by using the histogram of the image. At last, the optimal exposure time of each region was searched. The method combined the local information entropy with the exposure time, which maked different exposure to different regions and avoided the shortcomings of the traditional exposure control effectively. Experimental results show that the image obtained with the proposed method has a good effect.