基于卷积神经网络的视频监控识别方法
Video Surveillance Recognition Method Based on Convolutional Neural Network
姜全越1
作者信息
- 1. 北京国电电力新能源技术有限公司,北京 100101
- 折叠
摘要
文章基于视频监控系统的总体结构,引入L2正则化,提出基于卷积神经网络(Convolutional Neural Network,CNN)的视频监控识别方法,以解决传统系统在目标识别和异常检测方面存在的问题.另外,采用UCF-Crime数据集对所提方法进行测试,结果表明该方法在不同监控场景下能够成功检测异常区域,具有较好的实际应用潜力.
Abstract
Based on the overall structure of video surveillance system,this paper proposes a video surveillance recognition method based on convolutional neural network by introducing L2 regularization,to solve the problem of traditional systems in object recognition and anomaly detection.In addition,the proposed method is tested with UCF-Crime data set,the results show that the proposed method can successfully detect abnormal areas in different monitoring scenarios,and has good practical application potential.
关键词
卷积神经网络/视频监控/正则化/目标检测Key words
convolutional neural network/video surveillance/regularization/target detection引用本文复制引用
出版年
2024
NETL
NSTL
万方数据