Systems and methods for scene categorization

Number: 10701365
Published: 2020-06-30
Filed: 2017-12-11
Assignee: Disney Enterprises, Inc.
Inventors: Arana; Mark, McPhillen; Jared
CPC: H04N19/139; H04N19/177; G11B27/34; H04N19/179; G11B27/19; G11B27/28; H04N21/23418
Verdict: Set aside scene categorization, content metadata/business
Source: Google Patents · FreePatentsOnline

Abstract

Systems and methods are provided for manually or automatically identifying delineating scenes in media content. Scene categorization parameters setting forth thematic and/or cinematic criteria are used to analyze frames or groups of pictures (GOPs) making up the media content. Those frames or GOPs having one or more characteristics that meet the thematic and/or cinematic criteria may be deemed to belong or be a part of a particular scene in the media content. Downstream applications or processes can be optimized based on the scene segmentation.

Background

TECHNICAL FIELD(1) The present disclosure relates generally to digital media encoding, and more particularly, some embodiments relate to systems and methods for identifying scenes for downstream applications, such as transcoding.DESCRIPTION OF THE RELATED ART(2) Due to the growing number of digital media distribution platforms that are available, a single piece of media content, such as a movie, is often digitized or transcoded into different formats and/or bit-rates. Many content distributors create multiple versions of a single video, song, or other digital media program for multiple avenues of distribution. For example, for a single piece of content, a video distributor might maintain a plurality of different files having the same resolution but different bitrates. These groups of files may be used in web distribution using adaptive variable bitrate streaming techniques. Moreover, different portions of media content may be utilized to generate previews, highlight clips, and the like for distribution.BRIEF SUMMARY OF THE DISCLOSURE(3) In accordance with one embodiment, a computer-implemented method comprises receiving a media content comprising a plurality of media blocks, and receiving scene delineation parameters setting forth at least one of thematic and cinematic criteria. The method further comprises identifying sets of media blocks meeting at least one of the thematic and cinematic criteria as defining scenes in the media content. Further still, the method comprises f

Claims

1. A non-transitory computer-readable medium having computer executable program code embodied thereon, the computer executable program code configured to cause a computer system to: recognize existence of one or more characteristics present in two or more groups of pictures (GOPs) of media content, wherein each of the two or more GOPs starts with a respective I-frame followed by one or more respective inter-frames; determine whether the one or more characteristics present in the two or more GOPs meet thematic criteria indicative of a scene of the media content; in response to determining that the one or more characteristics present in the two or more GOPs meet the thematic criteria, determine a start of the respective I-frame of the first one of the two or more GOPs and an end of the last one of the two or more GOPs; and transmit the two or more GOPs and the determined start and end of the two or more GOPs to a transcoder.