Dynamic 3D special effect rendering synthesis method and system based on cloud native architecture

Abstract

The invention discloses a dynamic 3D special effect rendering synthesis method and system based on a cloud native architecture, and relates to the technical field of image processing, and the method comprises the steps: S1, carrying out the fragmentation of an original video, and dynamically distributing fragments according to the real-time load state of a cloud rendering node, the multi-cloud rendering node renders and splices the fragments in parallel through a dynamic three-dimensional special effect generation module to output a synthesized video file; s2, respectively decoding a video stream and an audio stream of the original video to obtain a video frame and an audio block, aligning the special effect frame with the video frame and the audio block through a bidirectional matching algorithm, and then synthesizing the special effect frame and the video frame through a video and special effect synthesis module to generate a final display frame, and the final display frame is updated through the Laplacian pyramid and the sub-pixel convolution, and the final display frame and the audio block are written into the synthesized video file according to the timestamp sequence, so that the problems of high energy consumption, unnatural special effect and video fusion boundary and insufficient audio and video synchronization precision in traditional special effect rendering are solved, and the quality of the generated video is remarkably improved.

Images