复杂环境下视频目标检测及其在交通系统中的应用

设计并实现了一种自动检测和记录车辆违法行为的视频系统.使用局域阈值分割算法,有效地提取复杂光影下的道路标线,利用图像抖动估计和补偿技术克服了摄像机抖动所造成的影响,提出对局部图像的边缘进行统计判别,以消除车辆阴影的干扰.实践表明,此系统能较好地工作在多种天气条件下及较复杂的环境下.

Video Object Detection in Complex Environments and its Application to Traffic System

A sort of video system, which is used for automatically detecting and recording lawbreaking deeds of vehicles, is designed and implemented. The local threshold segmentation algorithm is used for extracting the lines on the roads efficiently in the complex illumination. The interference of camera dithering is overcome by image dithering estimation and compensation techniques. A new method of counting edges in local areas is proposed to remove the influence of shadow of vehicles. Practices show that the system is able to work in various weather conditions and complex environments.