Patent · US Active

Background model for complex and dynamic scenes

US8358834B2 · kind B2 · utility

14Cited by

24References

22Claims

0Family size

Assignee

Behavioral Recognition Systems, Inc. · US

Inventors

Wesley Kenneth Cobb · Diemelstraße, US
Ming-Jung Seow · The Woodlands, US
Tao Yang · Zhejiang, CN

Key dates

Filing date	Aug 18, 2009
Grant date	Jan 22, 2013
Priority date	—
Expiry date	Mar 30, 2031

Classification

Technology area (CPC G)Physics
CPC primaryG06V20/44
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Techniques are disclosed for learning and modeling a background for a complex and/or dynamic scene over a period of observations without supervision. A background/foreground component of a computer vision engine may be configured to model a scene using an array of ART networks. The ART networks learn the regularity and periodicity of the scene by observing the scene over a period of time. Thus, the ART networks allow the computer vision engine to model complex and dynamic scene backgrounds in video.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.