摘要
针对违章停车人工检测方式存在的准确率低、成本高且难以实时判断等缺点,文章提出一种监控视频中的移动目标侦测算法,用来检测和识别违章停车。该算法采用混合高斯模型进行背景建模,用以检测监控场景中的运动目标,并通过支持向量机(support vector machine,SVM)分类器判断运动目标是否为监控车辆,如果是监控车辆,则根据车辆停留时间将车辆分类,一旦发现违停车辆,系统会发出报警。实验结果表明,该算法准确率高、实时性好。
Aiming at the shortcomings of manual detection methods for illegal parking, such as low accuracy rate, high costs and difficulties in achieving real-time judgment, an illegal parking detection and classification algorithm based on video surveillance is proposed. The moving objects in video are detected by using mixture Gauss/an model, and classified by the support vector machine(SVM) classifier to determine whether they are vehicles. Then the vehicles are classified according to the retention time, and the system will send out alarm if any illegal parking is founded. The experimental results show that the algorithm has high accuracy and achieves real-time performance.
出处
《合肥工业大学学报(自然科学版)》
CAS
CSCD
北大核心
2015年第12期1639-1642,共4页
Journal of Hefei University of Technology:Natural Science
基金
安徽省科技攻关计划资助项目(1301b042023)
关键词
监控视频
违章停车检测
混合高斯模型
支持向量机分类器
surveillance video
illegal parking detection
mixture Gaussian model
support vector machine(SVM) classifier