基于人体头肩部与步态检测的闸机通行逻辑

The Gate Transit Logic based on Detecting Human Head and Shoulder

城市轨道交通门式闸机通行逻辑算法的应用领域主要是自动售检票系统。由于闸机直接面对乘客,它的安全性、可靠性、准确性、稳定性等性能直接决定了地铁自动售检票系统的命运。针对现有的闸机通行逻辑遇到的问题,提出了一种基于人体头肩部与步态检测的闸机通行逻辑算法,通过步态检测传感器与人体头肩部检测传感器共同作用,很好地解决了尾随现象,以及人与物的所属关系及其区分。

The algorithm of gate recognition in urban rail transit is mainly applied in automatic fare collection （AFC） system. The gate provides services for passengers directly, so its security, reliability, accuracy and stability are very important for the service quality of AFC. The author analyzes the logic problems in existing gate recognition, points out that these problems make the gate working ineffectively and unfriendly in specific and abnor- mal situations. These problems also initiate disorder in ticketing accounts and the loss of tickets. The author proposes a new method integrated with a logic algorithm supported by human gait detection sensors and head-shoulder detection sensors. This method can correctly judge type, location, speed and direction of the moving object in the passing channel. The principle of this method is to combine the information collected from sensors in gate and cameras over gate, makes different information to supplement and confirm mutually. In this way, the method provides logic of gait recognition without misjudgment and missing judgment, also it provides a brand-new idea for the existing logic of gate recognition and obtains good results in practice.