Analysis and design of pervasive computing intelligent surveillance system integrating optical sensors and behavioral analysis for industrial internet

Nowadays, the surveillance system performing as an indispensable part for modern industrial Internet has attracted much attention. With the continuous progress of microprocessor technology, sensor technology, particularly optical sensors, and network technology, it has entered the era of pervasive computing, which leads to the breakthrough of the surveillance system and the emergence of intelligent environment safety monitoring systems. However, the intelligence of surveillance systems is still a significant challenge. In this paper, we propose a surveillance system for detecting and recognizing people based on their behaviors from images or video sequences using computer vision techniques and optical sensor data, which is implemented by designing the system architecture through a hierarchical structure, applying key techniques such as background subtraction, modeling, and optical sensing analysis, and experimentally processing and testing a variety of behavioral samples. The alarm information generated by the proposed surveillance system aided by pervasive computing system can be delivered to remote terminals. Experimental results reveal that the environmental monitoring and alarm information can be processed in real time with devices, such as cell phones, telephones, and PCs. Therefore, the proposed approach can improve the performance of the surveillance systems.