摘要
为了加强公交安全防范,提出了一种基于视频图像的主要体现为过快移动速度的公交车人群异常情况检测方法。根据人群运动轨迹确立图像感兴趣区域;利用改进的Vi Be算法提取运动目标,缩小数据处理范围;用Shi-Tomasi角点检测算法提取特征点;最后利用带修正系数的金字塔Lucas-Kanada光流法提取运动目标速度信息,进行异常情况的检测。实验表明,与基本Vi Be算法相比,改进后的Vi Be算法对光照具有更好的鲁棒性,该人群异常检测方法正确率达86.4%以上。
In order to strengthen the bus safety precautions,this paper presents an image processing based detection algorithm to detect the anomalous crowd behavior which mainly refers to the rapid flow of the crowd in the bus.According to the trajectory of passengers,region of interest is determined.Moving targets are extracted and data processing range is reduced with an improved Vi Be algorithm.The Shi-Tomasi corner detection algorithm is used to extract keypoints.Through pyramid Lucas-Kanade optical flow with correction coefficients,speed information is collected to recognize anomalous behavior.Experimental results show that the improved Vi Be algorithm has more robustness to illumination than the Vi Be algorithm and the accuracy of the proposed algorithm is more than 86.4%.
出处
《南京理工大学学报》
EI
CAS
CSCD
北大核心
2017年第1期65-73,79,共10页
Journal of Nanjing University of Science and Technology
关键词
人群异常
感兴趣区域
运动目标检测
角点检测
光流法
anomalous crowd
region of interest
moving target detection
corner detection
optical flow
