Dark areas often exist in gray image as light and other reasons

hence

in these areas the image detail can not be directly identified. So an improved image enhancement algorithm based on local statistics and partially overlapped sub-block ideas is proposed in this paper. The proposed algorithm can achieve real-time enhancement of low gray image under dark background. Compared with traditional histogram equalization method

the new algorithm not only reserves the visual characteristics of bright region but also brings relatively weak noise enhancement. Experimental results tested on TMS320DM642 real-time image processing system show that the proposed algorithm obtains excellent display performance

and the processing time for each frame is around 8 ms

which satisfies the real-time requirement completely.