Nowadays, the issue of video surveillance has been considered very important and is of interest to many researchers. One of the applications of video surveillance is to care for the elderly and analyze their behaviors. This paper suggests an applied method for categorizing and recognizing human behaviors using hidden Markov model. For this purpose, behavioral characteristics were firstly extracted and feature vector was created. Then these features were categorized by the hidden Markov model. Ultimately, abnormal behaviors were identified and detected by the system. The results obtained by applying the proposed method represent an accuracy rate about 94% and in comparison with other similar methods, the proposed method has a higher degree of efficiency to determine the proper behavior of individuals.