When the traditional global auto-exposure algorithm is applied to high contrast special scenes such as dark background and bright target, there is often over-exposure of the target, which leads to the loss of the key information of the subject concerned. To solve this problem, an adaptive weight exposure method based on regional brightness detection is proposed. Firstly, the whole imaging field of view is divided into several regions, and then the position of the target in the field of view is identified by independent judgment. Finally, the automatic exposure control of gray statistics is carried out according to the preset region weight. The experimental results show that the proposed method can effectively improve the image quality of the bright target or dark target in the dark background, avoid the phenomenon of the target too bright or too dark, and better capture the texture features of the target. © 2021 IEEE.