运用时空域信息优化的虚拟检测线组式视频车辆检测算法

The Space-time Domain Information Optimized Vision Vehicle Detection Algorithm Based on Virtual Detection Line Group

本文提出了一种运用时空域信息优化的虚拟检测线组式视频车辆检测算法。针对车窗反光、环境瞬变等因素对车辆检测精度的影响,着重分析了对算法的优化方法。经过时空域信息的优化,这种算法的鲁棒性较运用单条虚拟检测线的算法有很大提高。算法的实验结果论文中也已给出,表明该算法有良好的前景。

This paper presents a vision vehicle detection algorithm based on virtual detection line group, which is optimized with space-time domain information. The main focuses are on the optimizing methods of the algorithm aimed at the reflex by vehicle windows and the instant change of environmental factors which affect the vehicle detection accuracy. The robustness of the optimized algorithm is greatly improved than that of algorithm using single virtual detection line. The experimental result of the algorithm indicates a promising future.