高感度映像とDeep Neural Networkを用いた夜間対応型知的防犯カメラ

カテゴリ: 論文誌(論文単位)

グループ名: 【D】産業応用部門

発行日: 2016/10/01

タイトル(英語): Study on Intelligent Security Camera Systems for the Automated Detection of Nighttime Snatching Incidents using Deep Neural Networks

著者名: 長山　格（琉球大学工学部情報工学科），宮原　彬（琉球大学大学院理工学研究科情報工学専攻），島袋　航一（琉球大学大学院理工学研究科情報工学専攻）

著者名(英語): Itaru Nagayama (Department of Information Engineering, University of the Ryukyus), Akira Miyahara (Graduate School of Engineering, University of the Ryukyus), Koichi Shimabukuro (Graduate School of Engineering, University of the Ryukyus)

キーワード: 夜間の路上，犯罪行為，Deep Neural Network，防犯カメラ，ひったくり　　nightly road，criminal incident，deep neural network，security camera，snatching

要約(英語): In this paper, we propose an intelligent security camera system for the automated detection of snatching incidents on streets during the night. Although over a half of all snatching incidents occur at night, this has not been considered in previous studies. Thus, an intelligent security camera system using a deep neural network and SAM (snatching action model) is presented in this paper, for the automated detection of snatching incidents at night. Certain characteristics of motion are determined from video streams, and using a deep neural network the system automatically classifies the situations in the video streams into criminal or non-criminal scenes. We consider many types of scenarios to perform experiments regarding snatching incidents on streets at night. The experimental results show that the system can effectively detect snatching incidents with an accuracy of 96.66%.

本誌: 電気学会論文誌D（産業応用部門誌） Vol.136 No.10 （2016） 特集：Okinawa型ロボット組み込み/サスティナブルシステム

本誌掲載ページ: 727-734 p

原稿種別: 論文／日本語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejias/136/10/136_727/_article/-char/ja/