Multi-graphics card data collaborative processing method, video conference system and cloud server

Abstract

The disclosure discloses a multi-graphics data collaborative processing method, a video conferencing system and a cloud server, and belongs to the technical field of video encoding and decoding. The method includes: allocating graphics cards for data processing to each conference terminal connected to the video conference, and completing the video data decoding of each conference terminal in parallel; the decoded original decoded data is respectively stored on the graphics card corresponding to each conference terminal; The nth graphics card assigned by the conference terminal obtains the original decoded data on the remaining graphics cards, and the nth graphics card scales each channel of the original decoded data according to the size specified in the private layout requested by the Xth conference terminal to obtain intermediate data; among them, Both X and n are non-zero natural numbers; the nth graphics card splices each channel of intermediate data after scaling, and then encodes to obtain a fused code stream, and feeds the fused code stream back to the X conference terminal. By implementing the technical solution of the present disclosure, one decoding channel can be shared by multiple encoding channels, and the number of encoding and decoding channels can be increased.

Description

technical field [0001] The present disclosure relates to video encoding and decoding technology, and in particular to a multi-graphics data cooperative processing method, a video conferencing system and a cloud server. Background technique [0002] At present, traditional CPU encoding for video conferencing not only has low encoding efficiency, but also has a limited number of encoding channels and consumes CPU resources. In order to increase the number of codecs and efficiency, graphics card codecs are introduced, and the number of codecs for a single graphics card is still relatively small. [0003] In a video conferencing scenario, it is generally necessary to integrate multiple conference sites at the same time, and finally encode a fusion screen. It may be necessary to configure a private layout for individual conference sites. In particular, each site needs to configure a private layout, and each site Layouts vary. Since the layout of each venue is different, codes c...

AI Technical Summary

Problems solved by technology

[0002] At present, traditional CPU encoding for video conferencing not only has low encoding efficiency, but also limits the number of encoding channels and consumes CPU resources.

In order to improve the number of codecs and efficiency, graphics card codecs are introduced, and because the number of codecs for a single graphics card is still small

Images