A Human Motion Capture Sensor Using Pyroelectric Infrared Detectors

It has very good prospects on machine vision, physical fitness, human identification, intrusion detection, alarm and many other areas to get information on human movement patterns, especially use automatic test equipment to track, measure the body's natural movement and extract the related features, such as stepping, arm swinging, walking, running, and jumping, etc. This has important theoretical and commercial value.It is the mainstream research direction to use machine vision or video capture technology for the automatic detection of human movement patterns. To detect body movements more fast, low-cost, this paper introduced the methods using pyroelectric infrared sensor for human motion capture. On the basis of comprehensive analysis on pyroelectric infrared sensor about the human body detection, human tracking and human feature recognition and so on. The feasibility of detecting the human body movement patterns by pyroelectric infrared sensors was demonstrated by designing experiment and signal analysis.The pyroelectric infrared detection experiment system of human motion patterns was designed and constructed. Combined with hemispherical, semi-cylindrical Fresnel lens, human movement patterns were detected and the related signals were analyzed. The relationship between measuring height, distance and the detection results is analyzed and demonstrated. The movement of different human bodies was analyzed and it was found that the arm swinging may be the key factor resulting in the difference of signal from different people. The stepping-jumping experiments were carried out , and the action numbers were detected and measured by the peakfinder algorithm and K-order filter and the distinction between stepping and jumping movements was achieved; The body movement signals under the semi-cylindrical Fresnel lens was analyzed and the conclusion was found that the human face and other exposed parts of the skin detection might be the critical impact on the human body detecting result; Through the STFT signal analysis, signal high-frequency part of the spectral density was mined, and the distinction between stepping and arm swinging was realized.Based on the MSP430F149 MCU, PIR Human Motion Detection Analyzer was designed and produced, for achieving accurate real-time human motion detection; Based on the data acquisition card, PIR human-computer interaction systems was designed and produced, for achieving long-time stability in the human body movement signal acquisition and real-time analysis, and then combined with the virtual keyboard technology, a simple human-computer interaction function was achieved between computers running games and human movement.