华南理工大学学报(自然科学版) ›› 2008, Vol. 36 ›› Issue (10): 43-46.

• 交通运输工程 • 上一篇    下一篇

基于梯度方向恒定性的运动车辆阴影检测

秦钟   

  1. 华南理工大学 土木与交通学院, 广东 广州 510640
  • 收稿日期:2007-10-26 修回日期:2008-01-15 出版日期:2008-10-25 发布日期:2008-10-25
  • 通信作者: 秦钟(1972-),男,博士,高级工程师,主要从事智能交通、图像处理研究. E-mail:qzscut@yahoo.com.cn
  • 作者简介:秦钟(1972-),男,博士,高级工程师,主要从事智能交通、图像处理研究.
  • 基金资助:

    国家自然科学基金资助项目(50578064)

Shadow Detection of Moving Vehicles Based on Texture Constancy in Gradient Direction

Qin Zhong   

  1. School of Civil Engineering and Transportation, South China University of Technology, Guangzhou 510640, Guangdong, China
  • Received:2007-10-26 Revised:2008-01-15 Online:2008-10-25 Published:2008-10-25
  • Contact: 秦钟(1972-),男,博士,高级工程师,主要从事智能交通、图像处理研究. E-mail:qzscut@yahoo.com.cn
  • About author:秦钟(1972-),男,博士,高级工程师,主要从事智能交通、图像处理研究.
  • Supported by:

    国家自然科学基金资助项目(50578064)

摘要: 交通参数的视频检测是智能交通系统的一个研究重点,其中运动车辆的分割是视频检测过程中的一个关键环节.目前,运动车辆阴影的检测与剔除是准确、有效地分割出运动车辆所面临的一个难题.文中发现并证明了梯度方向恒定性原理,在此基础上提出了一种基于梯度方向恒定性的阴影检测与剔除方法.该方法首先建立路面背景的梯度矢量图,根据与当前帧图像的梯度矢量图的比较结果,判断是路面背景还是运动车辆,然后对运动车辆区域进行形态滤波,弥补内部空洞和剔除杂点,进而准确分割出车辆.试验结果表明,该方法适应性强,车辆分割效果好.

关键词: 阴影检测, 车辆分割, 梯度方向, 交通视频

Abstract:

The video detection of tragic parameters is of vital importance to the intelligent transportation system, and the segmentation of moving vehicles constitutes one of the key steps in video detection. However, it is difficult to accurately and effectively detect and remove the shadow of moving vehicles in the segmentation process of moving vehicles. In this paper, the principle of texture constancy in the gradient direction is discovered and proved. Based on the principle, a method of detecting and removing the shadow of moving vehicles is proposed. In this method, a background gradient vectorgraph of highway surface is set up and compared with that of the present frame image to judge whether the graph represents the highway surface background or the moving vehicle. The moving vehicles are then accurately segmented after the morphological filtering for remedying inside cavities and rejecting isolated points. Test results indicate that the proposed method is applicable and effective..

Key words: shadow detection, vehicle segmentation, gradient direction, traffic video