Cloud video transcoding method and cloud server

Abstract

The invention provides a cloud video transcode method and a cloud server. The method comprises the following steps that: the cloud server decodes a video document to acquire a plurality of image frames of the video document; the cloud server respectively judges whether each of the plurality of image frames of the video document is a focus image frame or a background image frame; if image frames are focus image frames, then the cloud server acquires focus areas and background areas of the focus image frames and respectively performs lossless and lossy encoding to acquire encoded focus image frames; if image frames are background image frames, then performing lossy encoding on the background image frames to acquire encoded background image frames; and synthesizing the encoded focus image frames with the encoded background image frames into a transcode video document, and sending the transcode video document to a mobile terminal. According to the method of the embodiment, transcoding is high in efficiency, simultaneously image quality of transcode video documents is improved, video documents with different code rates can be provided according to different terminals, and user experience is improved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a cloud video transcoding method and a cloud server. Background technique [0002] Usually video files are only stored in one or several formats, bit rates, resolutions, and frame rates, which is difficult to meet the different needs of terminals with different screen sizes, different processor capabilities, and different network standards. Video files need to be transcoded. [0003] Current video transcoding methods mainly include: full-decoding and full-coding methods, partial full-decoding and full-coding methods, and DCT domain transcoding methods. The full solution and full encoding method needs to completely decode the video file to reconstruct the image, and then re-encode according to the terminal requirements, the calculation is complicated, and it cannot be transcoded in real time, and the image needs to be compressed twice during the transcoding, which reduce...

AI Technical Summary

Problems solved by technology

The full solution and full encoding method needs to completely decode the video file to reconstruct the image, and then re-encode according to the terminal requirements, the calculation is complicated, and it cannot be transcoded in real time, and the image needs to be compressed twice during the transcoding, which reduces the image quality; The encoding method decodes the video file and encodes it according to the macroblock information of the obtained macroblock. Although the transcoding efficiency has been improved, the image quality is reduced by compressing the image twice during the transcoding process; the DCT domain transcoding method After decoding the video file to the DCT domain, it is directly encoded according to the needs of the terminal, which is efficient, but the YUV data cannot be obtained, which will cause image drift. It can be seen that the existing video transcoding method is difficult to efficiently obtain high-quality transcoded video files, poor user experience

Images