论文部分内容阅读
介绍了一种城市交通流量视频检测系统及其车辆识别新算法。通过检测道路上的运动目标并利用目标的运动信息和外形特征进行车辆识别。考虑到目标的阴影会影响目标的形状特征 ,导致目标识别错误 ,提出了一种新的阴影检测算法———利用目标的灰度特征对运动目标进行阴影检测和分离。实验结果表明该系统在模拟城市混合交通环境下 ,能够克服目标阴影的影响 ,准确检测和识别车辆 ,同时能够满足实时性的要求
A city traffic flow video detection system and its new vehicle recognition algorithm are introduced. Vehicle identification is performed by detecting the movement target on the road and using the movement information and the shape characteristics of the target. Considering that the shadow of the target will affect the shape characteristics of the target and lead to the target recognition error, a new shadow detection algorithm is proposed, which uses the gray feature of the target to detect and separate the moving target. The experimental results show that the system can overcome the influence of the target shadow, simulate the detection and recognition of vehicles correctly and meet the requirements of real-time while simulating the urban mixed traffic environment