{"title":"使用YOLOv6卷积神经网络的可疑活动触发系统","authors":"S. Awang, Mohd Qhairel Rafiqi Rokei, J. Sulaiman","doi":"10.1109/ICAIIC57133.2023.10066970","DOIUrl":null,"url":null,"abstract":"Property theft is one of the crimes that increases in which leads to a major concern in Malaysia. Despite of having surveillance cameras (CCTV) everywhere, the crimes keep occur due to the lack of security system. The security system can be developed by utilizing the existence of CCTVs specifically home surveillance CCTV. Therefore, this paper introduces a security system known as Suspicious Activity Trigger System (SATS) that able to automatically trigger an alarm or an alert message whenever suspicious activity is detected from the CCTV video image. The activity will be detected in a video image using Deep Learning technique which is YOLOv6 Convolutional Neural Network (CNN) algorithm. The algorithm will detect an object which is a person in the video and classify it as a suspicious activity or not. If the activity is classified as the suspicious activity, the system will automatically display a trigger message to alert SATS user. The user can therefore take whatever appropriate measure to prevent being a victim. Experiments have been conducted using a dataset taken from Google Open Image. We also implemented the experiments on the self-obtained dataset. Based on the experiment, 92.53% for precision and 96.6% of the accuracy is obtained using this algorithm. Therefore, YOLOv6 can be implemented in the security system to prevent crimes in residency areas.","PeriodicalId":105769,"journal":{"name":"2023 International Conference on Artificial Intelligence in Information and Communication (ICAIIC)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2023-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Suspicious Activity Trigger System using YOLOv6 Convolutional Neural Network\",\"authors\":\"S. Awang, Mohd Qhairel Rafiqi Rokei, J. Sulaiman\",\"doi\":\"10.1109/ICAIIC57133.2023.10066970\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Property theft is one of the crimes that increases in which leads to a major concern in Malaysia. Despite of having surveillance cameras (CCTV) everywhere, the crimes keep occur due to the lack of security system. The security system can be developed by utilizing the existence of CCTVs specifically home surveillance CCTV. Therefore, this paper introduces a security system known as Suspicious Activity Trigger System (SATS) that able to automatically trigger an alarm or an alert message whenever suspicious activity is detected from the CCTV video image. The activity will be detected in a video image using Deep Learning technique which is YOLOv6 Convolutional Neural Network (CNN) algorithm. The algorithm will detect an object which is a person in the video and classify it as a suspicious activity or not. If the activity is classified as the suspicious activity, the system will automatically display a trigger message to alert SATS user. The user can therefore take whatever appropriate measure to prevent being a victim. Experiments have been conducted using a dataset taken from Google Open Image. We also implemented the experiments on the self-obtained dataset. Based on the experiment, 92.53% for precision and 96.6% of the accuracy is obtained using this algorithm. Therefore, YOLOv6 can be implemented in the security system to prevent crimes in residency areas.\",\"PeriodicalId\":105769,\"journal\":{\"name\":\"2023 International Conference on Artificial Intelligence in Information and Communication (ICAIIC)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-02-20\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2023 International Conference on Artificial Intelligence in Information and Communication (ICAIIC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICAIIC57133.2023.10066970\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 International Conference on Artificial Intelligence in Information and Communication (ICAIIC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICAIIC57133.2023.10066970","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
摘要
财产盗窃是日益增加的犯罪之一,在马来西亚引起了很大的关注。尽管到处都有监控摄像头(CCTV),但由于缺乏安全系统,犯罪事件不断发生。安全系统可以利用现有的闭路电视,特别是家庭监控闭路电视来开发。因此,本文介绍了一种称为可疑活动触发系统(SATS)的安全系统,该系统能够在从CCTV视频图像中检测到可疑活动时自动触发警报或警报消息。使用深度学习技术,即YOLOv6卷积神经网络(CNN)算法,将在视频图像中检测活动。该算法将检测视频中是否有人的物体,并将其归类为可疑活动。如果该活动被归类为可疑活动,系统将自动显示触发消息以提醒SATS用户。因此,用户可以采取任何适当的措施来防止成为受害者。实验使用了来自Google Open Image的数据集。我们还在自己获得的数据集上进行了实验。实验结果表明,该算法的准确率为96.6%,精密度为92.53%。因此,YOLOv6可以在安全系统中实施,以防止居民区的犯罪。
Suspicious Activity Trigger System using YOLOv6 Convolutional Neural Network
Property theft is one of the crimes that increases in which leads to a major concern in Malaysia. Despite of having surveillance cameras (CCTV) everywhere, the crimes keep occur due to the lack of security system. The security system can be developed by utilizing the existence of CCTVs specifically home surveillance CCTV. Therefore, this paper introduces a security system known as Suspicious Activity Trigger System (SATS) that able to automatically trigger an alarm or an alert message whenever suspicious activity is detected from the CCTV video image. The activity will be detected in a video image using Deep Learning technique which is YOLOv6 Convolutional Neural Network (CNN) algorithm. The algorithm will detect an object which is a person in the video and classify it as a suspicious activity or not. If the activity is classified as the suspicious activity, the system will automatically display a trigger message to alert SATS user. The user can therefore take whatever appropriate measure to prevent being a victim. Experiments have been conducted using a dataset taken from Google Open Image. We also implemented the experiments on the self-obtained dataset. Based on the experiment, 92.53% for precision and 96.6% of the accuracy is obtained using this algorithm. Therefore, YOLOv6 can be implemented in the security system to prevent crimes in residency areas.