PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Multiscale Keypoint Analysis based on Complex Wavelets
Pashmina Bendale, William Triggs and Nick Kingsbury
In: British Machine Vision Conference, 31 Aug - 3 Sep 2010, Aberystwyth, U.K..

Abstract

We describe a new multiscale keypoint detector and a set of local visual descriptors, both based on the efficient Dual-Tree Complex Wavelet Transform. The detector has properties and performance similar to multiscale Foerstner-Harris detectors. The descriptor provides efficient rotation-invariant matching. We evaluate the method, comparing it to a previous wavelet based approach and to several conventional detectors and descriptors on a new dataset designed for the automatic evaluation of 3D viewpoint invariance. The dataset contains over 4000 images of toy cars on a turntable under accurately calibrated conditions. Both it and the evaluation software are publicly available. Overall the method gives performance competitive with existing Harris-like detectors.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Conference or Workshop Item (Paper)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Machine Vision
ID Code:7166
Deposited By:William Triggs
Deposited On:07 March 2011