Detection of fume hood operator based on image processing technology

摘要

It is a key parameter whether there is an operator or not in front of the fume hood when controlling it. The detection method is required to meet the needs of real-time, reliability, low-cost and low complexity. In this paper, a new algorithm based on frame difference and histogram similarity comparison is presented to determine whether there exists an operator or not accurately under the situation such as unclear image, personnel accounts for most of the image and personnel stillness in short time. The algorithm is simple and reliable, and could easily used in embedded system. The experiment results indicate that the algorithm has achieved a good accuracy and adaptability in detecting the fume hood operator under the conditions of operator free activities, short time stillness, and so on.