Space encoding based human activity modeling and situation perception

Human activity refers to a collection of motions and actions from either an individual or a group. Human activity study aims to recognize subjects' behavioral patterns and understand their situations. Human activity analysis and situation perception can be used to enhance the performance of various applications ranging from healthcare to surveillance and energy efficient buildings, etc. This paper presents a space encoding based framework for human activity information acquisition/analysis and situation perception. The novelties of this paper include: 1) efficient activity information acquisition based on a space encoding method; 2) reliable human activity representation based on stable statistical distributions; 3) effective activity-to-situation correlation based on data learning methods. Both simulations and experiments are performed by using pressure sensors deployed on the floor. The preliminary results have demonstrated the advantages of the proposed method.