Video flaw traceability and fixed-point regeneration method and system

Abstract

The present application relates to the field of artificial intelligence and digital video processing technology, and discloses a video flaw tracing and fixed-point regeneration method and system, the method comprising: establishing a multi-level data blood relationship mapping table in video generation pipeline execution, associating video frames to video segments, split shots, assets and script location layers, and carrying generation parameter sets; receiving user specified operations on flaw areas and picture coordinates, querying the mapping table and combining computer vision models to reversely locate to source production links and generation parameters; presenting an editing interface to receive parameter modifications; calculating the affected range, only locally regenerating the affected segments, and comparing the modified parameters with the original parameters to generate a change increment, which is transmitted to the downstream; using a multi-mode stitching strategy to splice, performing timing smoothness and role consistency verification after splicing, and switching modes or issuing warnings if unqualified. The present application realizes second-level visual positioning of flaws, reduces computing power consumption, and ensures seamless connection of the repaired video.

Images