Through the analysis of histogram distribution of the local spatio-te mporal features (gradient and optical flow) for different behavior videos,it is found that the statistics cha racteristics of gradient and optical flow for different behavior videos are obviously different respectively.In order to ensure the high descriptive of features to the behavior,a new method of human activity recognition is put forward by u sing the statistics characteristics of gradient and optical flow in this paper.Firstly,it is found that the histog ram distributions of gradient and optical flow for different behavior videos conform to the asymmetric generalized Gaussi an distribution (AGGD) through the mathematical statistic analysis.Secondly,the parameters of AGGD mo del are extracted respectively and fused to describe different behavior as the statistical features.Moreover,h uman behavior is recognized through calculating the Mahalanobis distance between the test video′s feature matrix and the train videos′.Finally,the performance is investigated in the KTH action dataset and Weizmann action datas et,and the average recognition rates are as high as 93.16% and 95.20% for the two action datasets,respectively. The results show that this method can generate a more comprehensive and effective representation for action videos .