Research On Gait Recognition Based On Bézier Curve Model

Gates intelligent recognize and control system is an important component of the orbit traffic automatic fare collection system. It provides control signal and various alarm signals to the gates. Its kernel part impose computer and its assistant system to automatically recognize the crowd movement target and make intelligent classification, thereby to achieve the aim of automatic control of the gates. Compare to other biologic characters, the gait feature has the superiority of non-invasive,non-contact,difficulty of hide and disguise. Therefore, the gait recognition has a good prospect as a rising domain. At the same time, it is an indispensably part of the gates intelligent recognizes and controls system.This paper for the first time propose a new method to recognize the human gait ,the method is to use the Bézier curve to simulate the human movement track and then abstract features from the curve and carry out the recognition. We first take and preprocess the original data that collected by the camera, this points list is the input initialized of our system. Then we use these points to draw the Bézier curve. This curve is the human movement track. Then we abstract feature vector from the curve, the feature vector is submitted to the recognizer, which make matching use the new feature vector and old feature vectors which stored in the Database. The match result is given as a similarity; the similarity is as big as the feature vectors reflecting the same people and vice versa. Experiments have proved that this method goes well; it can obtain a high degree of recognition.Our work has proved the theory availability of the Bézier curve in the domain of human movement gait recognition through a great deal of experiments. We can apply the method in many complicated application by expanding the system we proposed. We can conclude that the Bézier curve has a wide prospect in the domain of gait tracing.