摘要
根据系列入室盗窃案件嫌疑人的活动特征,提出了一种利用证据理论检测嫌疑人员的异常轨迹方法。该方法在进行时空距离约束的基础上考虑了案发前后主体位置是否异常、案发期间主体位置是否异常、案发前后主体速度是否异常等3个因素,利用证据理论,综合这3个指标来判别轨迹的异常程度,进而检测轨迹主体的异常性。实验结果表明,在海量的时空数据中,此方法能有效排除大部分在案件时空距离约束内的异常“噪声”,缩小侦查范围,能够一定程度上解决有预谋入室盗窃案件嫌疑人的异常轨迹识别问题。
According to the activity characteristics of suspects in a series of burglary cases,a method for detecting abnormal trajectories of suspects using evidence theory was proposed.Based on the space-time distance constraints,this method considered three abnormalities,including subjects’position before and after the case,subjects’position during the case,and subject’s velocity before and after the case.The evidence theory was used to judge the degree of abnormality of trajectory through these three factors,and then the abnormality of the subject’s trajectory was detected.The results showed that in the massive space-time data,this method could effectively eliminate most of the abnormal“noise”within the space-time distance constraints of the cases,narrow the scope of the investigation,and help solve the abnormal trajectory identification of the suspects in premeditated burglary cases to a certain extent.
作者
张新宇
陈鹏
顾海硕
ZHANG Xinyu;CHEN Peng;GU Haishuo(School of Information Technology and Cyber Security,People's Public Security University of China,Beijing 100038,China)
出处
《中国人民公安大学学报（自然科学版）》
2022年第4期61-66,共6页
Journal of People’s Public Security University of China(Science and Technology)
基金
北京市自然科学基金项目(9192022)。
关键词
异常轨迹检测
D-S证据理论
时空距离约束
证据
有预谋入室盗窃
abnormal trajectory detection
D-S evidence theory
space-time distance constraint
evidence
premeditated burglary
