论文部分内容阅读
作为智能监控系统的一部分,遗留物检测系统的主要作用是检测监控场景中有意或者无意遗留下来的行李、包裹、车辆等目标,并在检测出可疑遗留目标后发出警报,同时在视频中对遗留物进行标注。利用双背景建模方法检测遗留物掩膜,结合Camshift算法对疑似遗留物进行跟踪。在跟踪过程中,引入与搜索窗匹配的延迟因子方法避免遗留物的遮挡问题,成功实现并验证了该方法对于遗留物检测及报警的有效性。在此基础上,利用EmguCV开源计算机视觉库及Visual Studio2015和.NET框架,基于C#开发语言成功完成了视频中遗留物检测与报警系统的设计与实现。
As part of the intelligent monitoring system, the main function of the legacy detection system is to detect the target such as baggage, parcels, vehicles left behind intentionally or unintentionally in the surveillance scenario and to issue an alarm after the suspicious legacy target is detected and at the same time, Mark objects. The dual-background modeling method is used to detect the remnant mask, and the Camshift algorithm is used to track the suspected remnants. In the process of tracking, the delay factor matching with the search window was introduced to avoid the occlusion of the relics, and the effectiveness of the method in detecting and alarming the relics was successfully achieved and verified. On this basis, using EmguCV open source computer vision library and Visual Studio2015 and. NET framework, based on the C # development language successfully completed video in the detection and alarm system of legacy detection and design.