PT Unknown
AU Angel Sappa
   Fadi Dornaika
   David Geronimo
   Antonio Lopez
TI Registration-based Moving Object Detection from a Moving Camera
BT IROS2008 2nd Workshop on Perception, Planning and Navigation for Intelligent Vehicles
PY 2008
BP 65–69
AB This paper presents a robust approach for detecting moving objects from on-board stereo vision systems. It relies on a feature point quaternion-based registration, which avoids common problems that appear when computationally expensive iterative-based algorithms are used on dynamic environments. The proposed approach consists of three stages. Initially, feature points are extracted and tracked through consecutive frames. Then, a RANSAC based approach is used for registeringtwo 3D point sets with known correspondences by means of the quaternion method. Finally, the computed 3D rigid displacement is used to map two consecutive frames into the same coordinate system. Moving objects correspond to those areas with large registration errors. Experimental results, in different scenarios, show the viability of the proposed approach.
ER