论文部分内容阅读
车辆检测是智能交通系统(ITS)中的一个关键技术,也是视觉监控车辆跟踪的重要步骤。由于交通场景的复杂性,采集到的视频序列中存在很多面积较大的各种类型噪声。本文通过分析人眼对车辆识别的几个特点,定义了非目标形态和面积抑制,提出了一种新的噪声去除算法。为了提高计算效率,对序贯算法进行了改进,简化了整个算法的计算。实验结果表明,该算法具有较好的噪声抑制效果和较强的边界提取能力。
Vehicle detection is a key technology in Intelligent Transportation Systems (ITS) and an important step in the visual monitoring of vehicle tracking. Due to the complexity of traffic scenes, there are many types of noise with large area in the collected video sequence. In this paper, by analyzing several characteristics of human recognition of vehicles, a non-target shape and area suppression are defined, and a new noise removal algorithm is proposed. In order to improve the computational efficiency, the sequential algorithm is improved, which simplifies the calculation of the whole algorithm. Experimental results show that this algorithm has good noise suppression and strong boundary extraction ability.