CAJ | 학술논문

提出了一种基于物联网和视频分析技术的监狱智能监控系统。该系统主要由监控终端子系统以及犯人活动网络定位监测子系统组成，并利用视频分析技术对囚犯进行智能行为监控。实验结果表明：所提出的的透视变换及SVM的行为识别及目标分类方法可以有效监控罪犯行为，将监控人员从“紧盯屏幕”的低级监控模式中解放出来，能够很好地提高监控有效性，实现“人防、技防、物防”三位一体的最佳结合。
제출료일충기우물련망화시빈분석기술적감옥지능감공계통。해계통주요유감공종단자계통이급범인활동망락정위감측자계통조성，병이용시빈분석기술대수범진행지능행위감공。실험결과표명：소제출적적투시변환급SVM적행위식별급목표분류방법가이유효감공죄범행위，장감공인원종“긴정병막”적저급감공모식중해방출래，능구흔호지제고감공유효성，실현“인방、기방、물방”삼위일체적최가결합。
An intelligent monitoring system is presented based on the prison IOT and video analysis technology. The system mainly consists of the monitoring terminal subsystem based on embedded platform, positioning monitoring subsystem which is based on the activities of prisoners networking technology. Video analysis technology is used for intelligent behavior monitoring of prisoners. Experimental results show that,the proposed perspective transformation and SVM classification method can effectively monitor the behavior of the prisoners,which free the monitors from the low-level monitoring mode and can improve the efficiency of monitoring to achieve the best combination of"air de-fense,anti-technology,anti-matter".