Unified Real-Time Tracking and Recognition with Rotation-Invariant Fast Features

Author

Takacs, Gabriel ; Chandrasekhar, Vijay ; Tsai, Sam ; Chen, David ; Grzeszczuk, Radek ; Girod, Bernd

Author_Institution

Inf. Syst. Lab., Stanford Univ., Stanford, CA, USA

fYear

2010

fDate

13-18 June 2010

Firstpage

934

Lastpage

941

Abstract

We present a method that unifies tracking and video content recognition with applications to Mobile Augmented Reality (MAR). We introduce the Radial Gradient Transform (RGT) and an approximate RGT, yielding the Rotation-Invariant, Fast Feature (RIFF) descriptor. We demonstrate that RIFF is fast enough for real-time tracking, while robust enough for large scale retrieval tasks. At 26× the speed, our tracking-scheme obtains a more accurate global affine motion-model than the Kanade Lucas Tomasi (KLT) tracker. The same descriptors can achieve 94% retrieval accuracy from a database of 10⁴ images.

Keywords

augmented reality; feature extraction; gradient methods; image motion analysis; image recognition; mobile computing; optical tracking; transforms; video signal processing; KLT tracker; Kanade Lucas Tomasi tracker; RIFF descriptor; global affine motion-model; mobile augmented reality; radial gradient transform; real-time tracking; retrieval accuracy; rotation-invariant fast feature; video content recognition; Augmented reality; Filters; Handheld computers; Information systems; Laboratories; Large-scale systems; Mobile handsets; Real time systems; Robustness; Tracking;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on

Conference_Location

San Francisco, CA

ISSN

1063-6919

Print_ISBN

978-1-4244-6984-0

Type

conf

DOI

10.1109/CVPR.2010.5540116

Filename

5540116