Video Analytics and WebRTC

Abstract

Video Analytics which is also known as Video Content Analysis (VCA) is the capability of automatically analyzing Video and to detect events not based on a single image. It will analyze video streams run time to detect events related to particular type of condition. It will be helpful for applications such as Intrusion Detection, Crowd Management, and Baggage Detection.

WebRTC stands for Web Real-Time Communication. Aim of the Project is to Stream Real time streaming from Surveillance cameras to Browser which will sup-port camera list, multiple streaming and PTZ (pann/till/zoom). WebRTC contains main three components WebRTC Client, WebRTC gateway, Media Server. Client (Browser) supports multiple real-time streaming of surveillance cameras with support of PTZ. Media Server is a combination of RTSP Server and Streaming Server. To stream RTP packets from Media Server to Client it requires signaling mechanism to complete. WebRTC Client has its own specification such as it re- quires Web Socket, Ice-Stun, DTLS-SRTP Signaling, and can Play only Secured RTP packets and support VP8 codec only. Media Server does not support web-socket protocol, it supports rtsp signaling and supports H264 codec and stream RTP packets only. To request Streaming from Media Server and to support PLAY, PAUSE and STOP actions from client RTSP signaling mechanism is required to follow. WebRTC gateway is added as a mediator between WebRTC Client and Media Server. In WebRTC Gateway supports Web-Socket Signaling, RTSP Signaling, and Ice Mechanism for NAT Traversal. it supports DTLS to export keying material to encrypt RTP and RTCP packets. After following Signaling mechanism, Gateway will receive RTP and RTCP packets from Media Server, encrypt it as SRTP and SRTCP packets and route it to WebRTC Client.