计算机与数字工程2024，Vol.52Issue(4)：1169-1172,4.DOI:10.3969/j.issn.1672-9722.2024.04.037
基于固定背景的视频浓缩算法设计与实现
Design and Implementation of Video Concentration Algorithm Based on Fixed Background
摘要
Abstract
Surveillance video is to extract valuable content information from long-term surveillance videos through image pro-cessing and machine vision.In the field of video surveillance,only manual video screening will cause inefficiency and problems of under-reporting and false-reporting.The article uses the adjacent frame content detection method,the motion segment detection method to cooperate with the Harris feature point extraction algorithm and the SIFT feature point extraction algorithm to achieve im-age feature point registration.A dual-mode fusion background modeling and content detection method is proposed,and a 2D dis-tance-based motion content tracking method and GrabCut matting algorithm are used to design and implement the video concentra-tion algorithm.关键词
视频浓缩/时间空间/背景建模/运动检测/双模融合Key words
video enrichment/time and space/background modeling/motion detection/dual mode fusion分类
信息技术与安全科学引用本文复制引用
周冬梅,余洪嘉..基于固定背景的视频浓缩算法设计与实现[J].计算机与数字工程,2024,52(4):1169-1172,4.基金项目
四川省科技厅重点研发项目"视频监控中运动目标图像分割关键技术研究"(编号:18ZDYF3425)资助. （编号:18ZDYF3425）
