基于视频数据挖掘的城市轨道交通车站行人交通行为特征提取系统研究

Study on Characteristics Extraction System of Pedestrian Traffic behavior for Subway Stations Based on Video Data Mining

城市轨道交通车站行人交通行为特征数据对于站内行人安全管理,以及疏导设施安全设计具有重要参考价值。针对城市轨道交通车站人流密集条件下行人交通行为数据采集速度慢、结果精度差的问题,设计一套基于视频数据挖掘的城市轨道交通车站行人交通行为特征提取系统。运用Hadoop架构,通过对采集到的连续相邻俯拍视频进行存储及拼接、提取二维平面中行人交通行为基础数据、判别行人交通行为类别,实现行人各交通行为特征参数的提取并运用分布式存储技术对其存储,最后通过实验数据验证系统的有效性。

Characteristic data of pedestrian traffic behavior in subway stations has important reference value for the passenger safety management and the safety design of facilities. In view of the low acquisition speed and poor precision of pedestrian traffic behavior data in subway stations under high passenger flow, a characteristics extraction system of pedestrian traffic behavior for subway stations was designedon the basis of video data mining. The Hadoop architecture was used to analyze and store the data, and the continuous adjacent videos were spliced and stored. Pedestrian trajectories and basic pedestrian motion parameters in a two-dimensional plane were extracted, and pedestrian traffic behavior types were discriminated. Then the extraction of characteristic parameters for pedestrian traffic behavior was realized and the distributed storage technology was used to store them, and finally the system effectiveness was verified by experimental data.