远程视频实时人脸识别系统设计

Design of remote video real-time face recognition system

由于目前的嵌入式处理速度与PC相比还有一段差距,对于复杂的图像处理略显乏力,所以统筹两者优点,设计实现了一种高效的远程视频实时人脸识别系统。设计采用嵌入式ARM作为开发平台,并接入摄像头和无线网卡来实现前端部分视频的采集和传输。PC作为接收端,并配置开源的视觉处理工具OpenCV实现人脸识别和达到视频监控的目的。通过两者的结合既利用嵌入式优势,也获得了PC的处理速度,而且由于网络化,所以对监控的环境,距离等可以随意的调整,对工业要求或特定场合有一定的借鉴作用。

Because the speed of current embedded processing is slower than PC,it is not good at complex image processing,so co-ordinate the two advantages,and design an efficient real-time face recognition system for remote video.This design uses the embedded ARM as the development platform,and access camera and wireless network card to capture and transmit video.As the receiving end,PC configures the open-source visual processing tool OpenCV to achieve the purpose of face recognition and video surveillance.Through the combination the design gets advantages of embedded and processing speed of PC.What's more,it can adjust the environment and distance due to the network,and also can provide some reference for industrial requirements and specific occasions.