Abstract: Aiming at the flicker problem faced by vehicle streaming media rearview mirror under LED signal light environment, meanwhile, considering the impact of rapid local image motion and complex edge textures on image restoration, a real-time flicker mitigation system based on FPGA(Field Programmable Gate Array) is proposed in this paper. Firstly, the constraint conditions of the flicker pixels are established according to the features of flicker images such that the flicker region are efficiently detected through three consecutive frames. Then, a motion estimation algorithm based on neighborhood images groups is proposed, the image matching criterion is improved, and the neighborhood images vectors is clustered to enhance image inpainting effect in complex scene. Finally, a parallel pipeline processing architecture for video data streams is designed based on the ZYNQ(ZYNQ-7000 All Programmable SoC) platform to achieve system acceleration. The experimental results show that the proposed system can satisfy the real-time processing requirement of 800×480 resolution and 30 fps video, and has good mitigation effect on flickering images in complex scenes.