Media data encoding apparatus and method

摘要

A method and apparatus for encoding media content. In the field of video compression, temporal compression generates I-frames and dependent P-frames and B-frames. The placement of I-frames is determined by the presence of scene switches in the video. In order to improve the ability of a user to seek to interesting parts of the video, a video encoder considers semantic data identifying narrative points of interest within the video. The encoding process generates encoded video data having I-frames present at scene transitions and also at narratively interesting parts of the video.

主权项

1. A method of encoding video media content into an encoded video sequence of independent frames and dependent frames, the method comprising:
analyzing the video media content to determine where scene changes occur within the media content; generating encoding parameters for the encoded video sequence, including placement of independent frames corresponding to the location of said scene changes; accessing narrative data stored in a narrative data file generated by an author of the video media content, said narrative data indicating narratively significant audio track sections of the media content; updating the encoding parameters so that further independent frames are generated at locations indicated by the narrative data; and generating the encoded video sequence in accordance with the updated encoding parameters so that the encoded video sequence contains independent frames at locations corresponding to identified scene changes and at locations indicated by the narrative data relating to narratively significant audio track sections.