Video coding and decoding method and device

Abstract

The invention provides a video coding and decoding method and device. The method provided by the invention comprises the following steps: 1, receiving a feedback real time control protocol (RTCP) message packet; 2, acquiring a packet loss ratio; 3, computing reference frames: firstly, dividing P frames into P* frames and P' frames: (a) enabling each P' frame to select a P frame which is in front of the P' frame and in the same group of pictures (GOP) as a reference frame; and and (b) computing all P* frames and corresponding reference frames of the P* frames in one GOP according to the preset cross rules corresponding to different packet loss ratios, wherein the cross rules meet the following promise: the cross degree is increased with the increase of the packet loss ratio; 4, coding according to the acquired reference frames; and 5, outputting and displaying after decoding, and feeding back the RTCP message packet. The method and the device provided by the invention can dynamically select the reference frames of the P frames, have no need of always using the previous P frame as a reference frame, and can improve the fluency and the quality of videos as much as possible under the unstable network condition.

Description

technical field [0001] The present invention relates to video codec technology, in particular to a video codec method and device. Background technique [0002] At present, using RTP (Real-time Transport Protocol, Real-time Transport Protocol) for real-time video transmission is the most commonly used method in the industry, but since RTP is based on UDP (User Datagram Protocol, User Datagram Protocol), it provides a simple Unreliable data delivery service, so frequent network packet loss is unavoidable. Once packet loss occurs, there are currently two processing methods commonly used: continue to send data or stop sending data. If the first method is used, the subsequent video data will continue to be sent to the decoder for decoding and display. At this time, the image will move, but the mosaic phenomenon is likely to appear; in addition, if the second method is used, the subsequent affected data will be temporarily stopped. In this way, although there will be no mosaic p...

AI Technical Summary

Problems solved by technology

If real-time video transmission is required, only two frame types, IDR frame and P frame, need to be used in the I frame. Among them, the IDR frame uses intra-frame predictive coding, and it does not need to depend on any other frame. The H.264 standard will Regularly generate IDR frames, which can be used to synchronize images in the case of video packet loss; P frames use forward predictive coding, and the coding efficiency and compression ratio are higher than IDR frames, but P frames need to be used in the same GOP and in Its previous P frame is used as a reference for encoding. If the current P frame is lost, it is likely to cause decoding deviations of subsequent P frames, resulting in mosaic phenomena.

Images