Rope Deployment Method for Ropeway-Type Vermin Detection Systems
Rope Deployment Method for Ropeway-Type Vermin Detection Systems
复制标题
DOI：
10.1109/aina.2018.00060
复制
发表时间：
2018-05
期刊：
影响因子：
--
通讯作者：
Kodai Ogura;Kei Nihonyanagi;R. Katsuma
中科院分区：
文献类型：
--
作者：
Kodai Ogura;Kei Nihonyanagi;R. Katsuma
近年来，鹿、野猪或猴子等害虫对农村地区的破坏在频率和严重程度上都有所增加。这一问题有望通过由多个具有无线通信设备的传感器节点组成的无线传感器网络来解决。这些系统通过探测害虫并通过声音和光线等信号将它们赶走，从而减少了损害。然而，由于其固定的监控摄像机，一般的监控系统并不总是能够应对植物生长和其他减少监控面积的遮挡。本文提出了一种将监控摄像机移动到绳索上的索道式害虫检测系统，并提出了一种在预期监控场景下，将所需绳索数量最小化的方法。为了用尽可能少的摄像机进行有效的监测，该方法通过聚类过程将多个目标区域分组为一个目标区域。然后可以从一个位置监视分组区域。随后，我们的算法找到最有效的绳索部署，完全监控分组区域。在模拟中，该方法比一般聚类方法（k-means聚类）少26%的监控摄像机监控所有目标区域。
In recent years, damage to rural areas by vermin such as deer, wild boars or monkeys has increased in both frequency and severity. This problem is expected to be counteracted by wireless sensor networks constructed from multiple sensor nodes with wireless communication devices. These systems reduce the damage by detecting vermin and repelling them by signals such as sounds and light. However, owing to their fixed monitoring cameras, general monitoring systems cannot always cope with plant growth and other obscurations that decrease the monitored area. This paper proposes a ropeway-type vermin detection system that moves the monitoring cameras on ropes, and a method that minimizes the number of required ropes in the expected monitoring scenario. For efficient monitoring with as few cameras as possible, the method groups several target areas into one by a clustering procedure. The grouped area can then be monitored from a single position. Subsequently, our algorithm finds the most efficient rope deployment that completely monitors the grouped areas. In simulations, the proposed method monitored all target areas with 26% fewer monitoring cameras than a general clustering method (k-means clustering).