Agent Vi's Core Technology

 

Agent Vi’s core technology is based on its patented software architecture “Image Processing over IP Networks” (IPoIP™). The IPoIP™ architecture was designed to allow for an optimal deployment of Video Analytics, especially in large scale installations, while maintaining an open architecture approach which allows the end-user to mix and match various other components into the surveillance system.

Agent Vi’s IPoIP™ architecture distributes the Video Analytics task between an edge device (IP camera or encoder) and a server. A “slim” software component (the “Agent”) is embedded within the edge device and performs a preliminary image processing task of “feature extraction”. The Agent’s task requires only limited processing resources and therefore may be applied to most IP edge devices, without burdening the device’s processor. The image features extracted by the Agent are sent to a server for further analysis, where the actual analysis and detection is performed. The Video Analytics function performed at the server only requires processing of the features sent by the Agent, and not the entire image. Accordingly, a single server can handle a large number of video channels simultaneously (due to the analysis of feature data only from each video channel).

With Agent Vi’s unique architecture, a typical quad core server is able to simultaneously perform the full range of Video Analytics functionalities on up to 200 cameras, without compromising on Video Analytics performance.