公开(公告)号：DE10035043A1

公开(公告)日：2001-03-01

申请号：DE10035043

申请日：2000-07-19

Applicant: IBM

Inventor： BERGMAN LAWRENCE DAVID ,  CASTELLI VITTORIO ,  CHANG YUAN-CHI ,  LI CHUNG-SHEN ,  SMITH JOHN RICHARD

IPC: G06F17/30

Abstract: The method involves forming a first convex envelope of N input data sets, whereby M data sets represent apex points of the first convex envelope and are associated with a first plane of the indexing structure and forming a second convex envelope of the rest of the N -M data sets, whereby P data sets represent apex points of the second envelope and are associated with a second plane and at least part of one plane is located geometrically in another plane. An Independent claim is also included for an arrangement for building a multi-dimensional indexing structure for use with linear optimising requests and for a product for processing a request.

公开(公告)号：GB2464123A

公开(公告)日：2010-04-07

申请号：GB0818176

申请日：2008-10-04

Applicant: IBM

Inventor： CHANG JANE ,  NATSEV APOSTOL IVANOV ,  SMITH JOHN RICHARD

IPC: H04N5/91 ,  G06T1/00 ,  G11B20/00

Abstract: A computer implemented method, apparatus, and computer program product for temporal, event-based video fingerprinting. In one embodiment, events are detected (904) amongst a plurality of video frames, the events representing discrete points of interest in the video content, such as changes in scene, speaker, audio, frame intensity, colour, texture, shape, edges or motion. Consequently, a set of temporal, event-based segments are generated (906) using the events. Each temporal, event-based segment is a segment of the video content covering a set of events. A time series signal is derived (908) from each temporal, event-based segment using temporal tracking of content-based features of a set of frames associated with the each temporal, event-based segment. A temporal segment-based fingerprint is extracted (910) based on the time series signal for the each temporal, event-based segment to form a set of temporal segment based fingerprints associated with the video content. The fingerprints may be compared with unknown video content to determine whether such content is a near duplicate of original video content. This method has applications in video content filtering, searching and determination of video piracy.