In order to improve the capability of low-light-level imaging system for image processing of moving object, a self-adaptive image enhancement algorithm was proposed. This method is based on the combination of multi-frame image processing and target speed estimation, which can decrease the image flicker noise effectively, improve the image contrast, and is applicable for moving target image processing. Test was made to the method on a platform with a Field Programming Gate Array(FPGA) as the core device. The result shows that: With the help of this processing method, the Signal to Noise Ratio (SNR) of images improves a lot from 2. 1 dB to 11. 5 dB when target is still; and the SNR will not decrease as dramatically as general methods do when the target is moving