门禁系统中的异常行为检测

发布时间：2018-05-24 06:28

本文选题：门禁系统 + 目标检测　；参考：《计算机应用研究》2017年03期

【摘要】：针对门禁系统中尾随、蛮力开门两种异常通过的行为,结合门禁系统的身份验证功能,采用图像处理技术对违规行为进行检测,以进一步提升门禁系统的安全性能。设计并使用了一种基于轨迹分析的人数统计方法,通过目标检测与跟踪算法得到目标在监控场景中的运动轨迹,根据轨迹判定目标进出门禁的情况;然后根据监控场景的特点,设计了一种基于计数线的目标计数规则,结合目标运动轨迹实现人数统计;最后结合门禁系统的身份验证功能,设定异常通过行为的检测规则,检测定义的异常行为。实验结果表明,该方法能够有效地检测通过门禁系统时发生的异常行为,尤其对于行人比较稀疏和没有明显遮挡的情况,具有很好的检测效果,该方法平均检测准确率能够达到90%。
[Abstract]:In order to improve the security performance of the entrance control system, the image processing technology is used to detect the illegal behavior by combining the identity verification function of the entrance guard system with the trailing and brute force two kinds of abnormal passing behavior in the entrance control system, so as to further improve the security performance of the entrance control system. This paper designs and uses a statistic method based on trajectory analysis to get the moving track of the target in the monitored scene through the target detection and tracking algorithm, and then determines the entry and exit of the target according to the trajectory, and then according to the characteristics of the monitoring scene, A target counting rule based on counting line is designed to realize the population statistics combined with the moving track of the target. Finally, combined with the authentication function of the access control system, the detection rules of the abnormal passing behavior are set up to detect the defined abnormal behavior. The experimental results show that this method can detect the abnormal behavior when passing through the access control system effectively, especially for the cases where pedestrians are sparse and have no obvious occlusion. The average detection accuracy of this method can reach 90%.
【作者单位】：西南科技大学计算机科学与技术学院;
【基金】：西南科技大学研究生创新基金资助项目(15ycx055)
【分类号】：TU855;TP391.41

【相似文献】