Research On How To Recognize Moving Persons' Behaviors Under Special Circumstance Based On Images

With the rapid development of urban railway traffic, Automatic Fare Collection (AFC) System has become more important. In urban railway traffic, gate is one of key apparatus and used to automatically examine tickets by means of its intelligent recognition system. The recognition system is the"brain"and core of gate in urban railway traffic. We are facing with the tremendous challenges in the aspects of how the passengers are recognized quickly and correctly using the recognition system of the gate, owing to complicated transits. This paper has made a further study on this issue, and the main content is as follows:To recognize moving objects'behaviors through the gate based on images is mainly used to recognize the number of person in a picture. The environment the system working in, makes the background of the picture relatively fixed, so it makes you can use a simple method to cut off the background of the picture, making it easy to recognize the number. Second, this program use a self-adapt threshold in the division of image, we do this to adapt the light change impacting the division process. Third, this program uses an edge detection method base on the assumption of the highest point to save process time effectively. At last, a new relating method between frames is used to raise accuracy in single frame recognition. The result is satisfied, it can recognize the number of persons through the gate, it can also recognize person from obvious object. The result indicates that it runs fast enough.