Robust Rule-Based Method for Human Activity Recognition

Human activity recognition is an active research field in computer vision and image processing. In this paper we propose a robust rule-based method for the task of recognition of human activities and scenarios in video image sequences. The methodology uses a context-free grammar based representation scheme to represent human actions. The proposed system consists of three major steps. Initially by using a single camera, in a variety of angles, the movement of the object is detected, and object silhouette is generated in each frame. Then, the proposed method for generating a silhouette is presented. The silhouette is used to determine human activities such as running and walking. In the last stage, a rule-based classifier is used to classify the action. The experimental results show that the system can recognize seven types of primitive actions with high accuracy.