A video transmission method, device, equipment and medium

Abstract

The present application discloses a video transmission method, device, device, and medium. The method includes: acquiring an image sequence, and performing real-time encoding on the image sequence based on a preset encoding bit rate to obtain corresponding key frames or encoded frames. The target frame group is obtained; the target frame group includes a key frame and several coded frames after the key frame; based on the data packet segmentation length, the obtained key frame or coded frame is segmented in real time to obtain multiple data packets; Filling the data packet header information into the data packet according to the set data packet format to obtain the encapsulated data packet; sending the encapsulated data packet to the receiving end based on the first preset time interval, so that the receiving end based on the data of the encapsulated data packet The header information reassembles the key frame and the coded frame to obtain a video picture. It can improve H264 and H265 video stream transmission congestion and frame loss in poor network environments such as strong signal interference and weak signals, or network environments that limit throughput.

Description

technical field [0001] The present invention relates to the field of video transmission, in particular to a video transmission method, device, equipment and medium. Background technique [0002] During the transmission of the H264 or H265 stream, there is roughly one key frame and several coded frames every 2 to 4 seconds, and the size of the key frame is more than tens of KB, or even more than a hundred KB, and the data volume of the coded frame is small, usually It is only a few KB, so when transmitting key frames, there will be a burst of large data transmission, and the Wifi network is insufficient for the burst of large data transmission, not only easy to reach the transmission bottleneck, but also easy to lose packets, resulting in instantaneous congestion. At present, on the basis of TCP (i.e. TransmissionControl Protocol, Transmission Control Protocol) or UDP (i.e. User Datagram Protocol, User Datagram Protocol) for code stream transmission, after sending a data pack...

AI Technical Summary

Problems solved by technology

[0002] During the transmission of the H264 or H265 stream, there is roughly one key frame and several coded frames every 2 to 4 seconds, and the size of the key frame is more than tens of KB, or even more than a hundred KB, and the data volume of the coded frame is small, usually Only a few KB, so when the key frame is transmitted, there will be a burst of large data transmission, and the Wifi network is insufficient for the burst of large data transmission, not only easy to reach the transmission bottleneck, but also easy to lose packets, resulting in instantaneous congestion

At present, on the basis of TCP (that is, Transmission Control Protocol, Transmission Control Protocol) or UDP (that is, User Datagram Protocol, User Datagram Protocol), the code stream transmission waits for the receiver to feedback the receiving status after sending a data packet. If it fails, it will continue to resend the packet, that is, it will perform real-time detection on each data packet on the Wifi network to determine whether it is successfully sent. If no feedback is received after timeout, the data packet will be resent. It is easy to cause the sent data packets to be continuously resent, resulting in the failure of subsequent data packets to be sent, causing video stream delays, congestion, frame loss, etc.

Images