Diversity,relevance and ambiguity of human behavior have caused great difficulties for behavior recognition.This paper proposes an improved FCM clustering algorithm for the problem that when clustering and modeling of behavioral characteristic descriptors in behavior recognition,the traditional FCM algorithm can not determine the number of clusters.The algorithm is based on the indexes of inter-cluster compactness and the separation of clusters to determine the initial cluster centers.Then model behavioral descriptors that have been clustered to reach the purpose of improving behavior recognition accuracy.The experimental results show that: the improved algorithm can classify and model behavioral descriptors better and improve the recognition accuracy.