基于目标识别技术的智能报警系统

Intelligent Alarm System Based on Target Recognition Technology

由于长江流域有钉螺孳生的情况,为防止江滩上的行人受到血吸虫感染的威胁,本文设计了一种基于物联网的HOG函数提取、SVM模式目标识别以及肤色检测技术,以树莓派为控制核心的智能目标识别报警系统。当检测到人们有脱衣、游泳等危险行为倾向时,发出提醒和预警,对血吸虫病的防治工作具有重要意义。该系统实现了自动感应、人体识别、智能报警等功能。实验结果表明,所设计系统的实时性、准确性及稳定性均达到了设计要求,为智能目标识别报警系统的设计提供了一种可行方案,具有十分广阔的应用前景。

At present,target recognition in the field of Internet of Things is one of the hot research directions at home and abroad.Based on the oncomelania breeding in the Yangtze River Basin.The system is designed with a target recognition algorithm based on the Internet of Things,HOG function extraction and SVM pattern recognition technology,and skin color detection technology in order to prevent pedestrians on the river beach from being threatened by schistosomiasis infection.Raspberry Pi is the intelligent target recognition alarm system that controls the core.When it is detected that people have the tendency of undressing,swimming and other dangerous behaviors,it is of great significance to issue reminders and early warning for the prevention and treatment of schistosomiasis.The system realizes the functions of automatic induction,human body recognition,intelligent alarm and so on.The experimental results show that the real-time,accuracy and stability of the designed system meet the design requirements,which provides a feasible scheme for the design of intelligent target identification and alarm system,and has a very broad application prospect.