The patent badge is an abbreviated version of the USPTO patent document. The patent badge does contain a link to the full patent document.
The patent badge is an abbreviated version of the USPTO patent document. The patent badge covers the following: Patent number, Date patent was issued, Date patent was filed, Title of the patent, Applicant, Inventor, Assignee, Attorney firm, Primary examiner, Assistant examiner, CPCs, and Abstract. The patent badge does contain a link to the full patent document (in Adobe Acrobat format, aka pdf). To download or print any patent click here.
Patent No.:
Date of Patent:
Sep. 23, 2003
Filed:
Aug. 21, 2000
Kevin Manbeck, Cranston, RI (US);
Chengda Yang, Auburndale, MA (US);
Donald Geman, Amherst, MA (US);
Stuart Geman, Providence, RI (US);
MTI Film LLC, Providence, RI (US);
Abstract
A method for determining labels for video fields by identifying the state of the field is disclosed. Some examples of a video field's state include the origin of the field as film or video, its relative location with respect to edit points, and in the case of film-originating material, the location within the standard sequential pattern which results from converting film to video. To determine the label of a video field, the conditional probability distribution for a particular sequence of states given the entire video sequence is calculated. This may be optimized by using dynamic programing to maximize the conditional probability function and thus the labels. To determine the conditional probability, first the joint probability distribution is determined for the observed video fields and the states. This joint probability is calculated by creating a data model and a structure model for the video sequence. From the joint probability distribution an equation representing the conditional probability of having a particular sequence of states given a particular video field is determined. By maximizing this distribution over all states, which may be performed through dynamic programming the most likely state is determined which is consistent with the given video sequence data. Once the states are determined, labels corresponding to the states may be inserted within the video sequence.