计算机应用 ›› 2012, Vol. 32 ›› Issue (06): 1581-1584.DOI: 10.3724/SP.J.1087.2012.01581

• 图形图像技术 • 上一篇    下一篇

复杂环境下的运动车辆检测

傅沈文   

  1. 广东农工商职业技术学院 机电系, 广州 510507
  • 收稿日期:2011-11-14 修回日期:2012-01-16 发布日期:2012-06-04 出版日期:2012-06-01
  • 通讯作者: 傅沈文
  • 作者简介:傅沈文(1967-),女,江西高安人,副教授,主要研究方向:信号处理、电气控制。

Moving vehicle detection in complex environments

FU Shen-wen   

  1. Department of Mechanics and Electronics, Guangdong AIB Polytechnic College, Guangzhou Guangdong 510507, China
  • Received:2011-11-14 Revised:2012-01-16 Online:2012-06-04 Published:2012-06-01
  • Contact: FU Shen-wen

摘要: 针对目前采用的车辆检测方法的优缺点,提出了一种新的车辆区域检测方法,能够消除阴影干扰。该算法首先运用选择性背景更新法进行背景相减,获取感兴趣区域,然后提出基于图的区域分割算法,对感兴趣区域进行再分割。该方法充分考虑了视频图像全局和局部的空间信息,根据分割区域的大小自动自适应地调节对图像局部细节的忽略程度,从而获取局部区域像素信息较为一致的分割块。最后基于分割过程中所具有的马尔科夫属性,运用条件随机域的方法建立分割后验概率分布,求取最大后验概率确定标号,并对具有相同标号的相邻分割进行合并。

关键词: 车辆检测, 阴影消除, 图区域分割, 马尔科夫属性, 条件随机域

Abstract: A new vehicle detection method is proposed, according to the advantages and disadvantages of the currently used vehicle detection methods. The proposed method can eliminate the effects of shadow interference. Firstly, obtain the interested area by background subtraction using selective background updates method. Then, do re-segmentation for the interested area via the proposed regional segmentation algorithm based on graph. Global and local space information of the video image is under fully consideration, so the algorithm can adaptively and automatically adjust the ignored degree for image local details according to the size of the segmentation area, and get segmentations whose local area pixel information is more uniform. Finally, based on the Markov property during video image segmentation, construct the segmentation posterior probability distribution by conditional random field method. By calculating the maximum posterior probability, determine the label, and merge the adjacent segmentations having same label.

Key words: vehicle detection, shadow elimination, graph regional segmentation, Markov property, conditional random field