【摘要】 在实验室视频监控中,应用图像识别技术改变人工不间断监视、视频回放等传统监控方法,提高实验室监控的智能性具有重要现实意义。本方案中利用图像形状特征分析技术,设计了一种适应固定摄像头监控画面的基于图像指纹匹配的简化快速图像识别算法,提出了基于图像指纹匹配算法和OpenCV的实验室监控方案,实现了实验室异常检测、实验室人脸识别、对外开放时间统计、监控图像变化率等功能,并自动生成图形报表。实验结果表明该方法可运用于实验过程监控、实验室状态监控、实验室利用、防火防盗等,并且在提取关键区域后,抗干扰能力加强,突出了重点区域,图像识别速度和精度都有大幅度的提高。更多还原

【Abstract】 In the laboratory video monitoring,the traditional monitoring methods are artificial uninterrupted monitoring, video playback etc,Tt has practical significance to use image recognition technique which can make the laboratory monitoring more intelligent.In this project,we designed a simple and fast image recognition arithmetic which is based on image fingerprint matching and adapt to the fixed camera monitoring picture,and put forward a new solution about laboratory monitoring based on image fingerprint matching algorithm and OpenCV.It implemented the laboratory anomaly detection,laboratory face recognition,opening time statistics,monitoring image change rate,etc.,and automatically generate graphical reports.The experimental results show that the method can be applied to the experimental process monitoring,laboratory status monitoring,laboratory use,fire alarm and so on,and after the Extraction of key areas,it strengthen the anti-interference ability and Image recognition speed and accuracy are improved greatly.更多还原