The 7th IFAC Symposium on Intelligent Autonomous Vehicles 2010

Full Program »

Experimental Assessment of Different Feature Tracking Strategies for an IPT-based Navigation Task

Francisco Bonin-Font
University of the Balearic Islands

Alberto Ortiz
University of the Balearic Islands

Gabriel Oliver
University of the Balearic Islands

Visual navigation strategies based on feature tracking can be
inappropriate for real-time robot applications if the approach does
not combine celerity with a high degree of accuracy. In this paper,
different techniques for image feature extraction and matching
(SIFT, SURF, FAST and KLT methods) are evaluated and compared with
the purpose of finding the option that best combines performance
with maximum speed in a specific visual reactive navigation strategy
based on Inverse Perspective Transformation.


Powered by OpenConf®
Copyright ©2002-2010 Zakon Group LLC