基于差分均值背景提取和矩阵分区目标检测算法的研究  

Background Extraction Based on Differential Mean Method and Shadow Detection Using Matrix Subregion Partition

作　　者：韩超[1] 邓甲昊[1] 邹金慧[2] 韩敏[3] 

机构地区：[1]北京理工大学机电工程与控制国家级重点实验室,北京100081 [2]昆明理工大学信息工程与自动化学院,云南昆明650051 [3]北京理工大学信息与电子学院,北京100081

出　　处：《北京理工大学学报》2012年第12期1247-1251,1257,共6页Transactions of Beijing Institute of Technology

基　　金：机电工程与控制国家级重点实验室(北京理工大学)基金资助项目(60874100)

摘　　要：为提高对车辆图像的检测程度和实时性,针对智能交通系统,通过对实时路况的信息采集和视频图像的处理,提出了一种基于差分均值的背景提取计算方法和矩阵分区域的阴影检测方法,最终得到一个视频车辆的检测原型,从而实现对运动车辆的检测.实验结果表明,此种方法简单、计算量小、鲁棒性高,能快速地提取出背景图像,检测出比较完整的车辆阴影,可满足多运动目标的实时检测要求.For intelligent transportation monitoring systems, a background extraction method based on differential mean and shadow detection using matrix subregion partition are proposed to improve the quality and speed of the vehicle image detection. The real-time traffic information was collected and its video images were sophisticatedly processed. Finally, a video vehicle monitoring prototype was setup to realize the detection of moving vehicles. Experimental results show that this method has the advantages of simplicity, robustness and small amount of calculation. It could extract a background image quickly and detect a complete shadow of vehicles. This method could meet the requirements of real-time detection of multiple moving targets.

关 键 词：智能交通 背景提取 阴影检测 车辆目标检测算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]