System Design On Human Body Behavior Based On Machine Vision And MMS Alarm

Among the problems of the traditional monitoring alarm system are various kinds of under-reporting due to human weaknesses, false alerts due to overly sensitive sensors, difficulties of collecting and analyzing data, enormous workload of computer and slow response times of several processing parts after an alert is set up.In this paper, aiming at passive traditional monitoring alarm system, we designed a digital, intelligent and human-oriented video image recognition monitoring and MMS warning system from the perspective of machine vision and wireless communication. In this system, PC detects moving objects in video sequences using inter-frame difference method, analyzes and judges h intrusion and climbing behavior of human motion to the doors and windows, and then sends the results of the analysis to the lower computer, which triggers camera modules with VC0706 chips built-in to capture live images and sends the images as MMS to user’s mobile phones via external GPRS modules. The key to the whole system is the image processing procedures on the PC and MMS alarm system on the lower computer and the hardware design.The results show that the body posture behavior detection procedure is able to analyze and judge intrusion and climbing behavior of human motion and trigger the MMS alarm. The whole system is stable in running and expected effect is attained and is worth further development.