Detect Human Body And Recognize Human Behavior Which Base On Polymerization Characteristics

In recent years, with the continuous development of smart devices, somatosensory equipment also entered a period of rapid development. In this type of product, the key technique is the behavior of the human body detection and human behavior identification. But there are still many unresolved issues, such as dealing with issues of dynamic sequence of images, the human body occlusion, etc.. For building fast and efficient human recognition and behavior detection system caused a lot of barriers to further the practical application of development of. In this study, the use of image pixels within the region related to the gradual and smooth transition between the conversion is continuous and smooth movement of inertia characteristics to improve human detection and recognition of human behavior and performance efficiency.This study presents the use of polymerization characteristics of human body to achieve detection and identification of human conception, design and implement a preliminary framework for the system. The polymerization characteristics mean the complex process of analysis, integration, calculating a series of characteristic dataset. Constructed interproximal specific algorithm which is used to detect and classify human body; build a cache parameters inversion algorithm to track the target; constructed inertia characteristics algorithm identifies behavioral; description and analysis of the three algorithms principles, ideas and implementation. A human detection and human behavior recognition system was tested which base on three algorithms.Experimental results show that the interproximal specific algorithms for the detection and classification of the human body have a good effect, with the original method can get more parameters, but also improve the efficiency and adaptability and sets up a good foundation for subsequent research. Experiments also show that cache parameters inversion algorithm and inertia characteristics algorithm has better adaptability in the human body tracking and human behavior recognition, reducing the cost of self-learning, while also improved in terms of efficiency, improved robustness and accuracy.