Video encoding method, decoding method, video transmission method, and electronic device
By differentially encoding key frames and non-key frames in the video and using visual representation information to generate a compressed video bitstream, the problem of video distortion caused by error accumulation is solved, achieving low bitrate, high-efficiency transmission and high-quality reconstruction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHEJIANG TMALL TECH CO LTD
- Filing Date
- 2026-03-18
- Publication Date
- 2026-07-17
AI Technical Summary
In existing video coding technologies, error accumulation leads to poor video quality after decoding, especially when the prediction residual quantization is large, resulting in serious distortion of the decoded video. In addition, the coding bitrate is high, which consumes a lot of bandwidth resources.
By identifying keyframes and non-keyframes in the video, visual representation information of non-keyframes is obtained, and keyframes and visual representation information are compressed and encoded separately to generate a compressed video stream containing keyframe bitstreams and non-keyframe visual representation bitstreams.
It significantly reduces the encoding bitrate, reduces the bandwidth consumption of video transmission, improves video transmission efficiency, and effectively alleviates the distortion problem caused by error accumulation, ensuring the quality of the decoded video.
Smart Images
Figure CN122420541A_ABST