A novel method of histogram analysis for background extraction of video images was proposed,which was derived from the pixel-based histogram analysis.The method make full use of the statistical properties of pixels between temporal frames and the correlation of local pixels in a single frame.When carrying out histogram analysis for background extraction,the proposed method is based on a 2×2 block of pixels,instead of a single pixel,so that the number of calculations are dramatically reduced.and can extract a sound background image from video sequence,simultaneously.Experiment was presented for comparison between the proposed method and the conventional pixel-based histogram analysis,and the results show that the proposed mehtod is faster and delivers better extracted backgrounds.