Real-time Vision Based Dynamic Sinkage Detection for Exploration Rovers

Al-Milli, S, Spiteri, C, Comin, F and Gao, Y (2013) Real-time Vision Based Dynamic Sinkage Detection for Exploration Rovers In: International Conference on Intelligent Robots and Systems, 2013-11-03 - 2013-11-08, Tokyo Big Sight, Japan.

Full text not available from this repository.

Abstract

Identification of the wheel sinkage of exploration rovers provides valuable insight into the characteristics of deformable soils and thus the ease of traversal is also identified. In this paper we propose a simple vision based approach that robustly detects and measures the sinkage of any shaped wheel in real-time and with little sensitivity to various operating conditions. The method is based on color-space segmentation to identify the wheel contour and consequently the depth of the sinkage. In addition, our approach also provides a dynamic sinkage analysis which potentially allows for the identification of non-geometric hazards. The robustness of the algorithm has been validated for poor lighting, blurring, and background noise. The experimental results presented are for a hybrid legged wheel from our in-house single-wheel test-bed.

Item Type:	Conference or Workshop Item (UNSPECIFIED)
Authors :	Name Email ORCID Al-Milli, S s.al-milli@surrey.ac.uk Spiteri, C Comin, F Gao, Y yang.gao@surrey.ac.uk
Date :	3 November 2013
Depositing User :	Symplectic Elements
Date Deposited :	17 May 2017 12:58
Last Modified :	17 May 2017 15:07
URI:	http://epubs.surrey.ac.uk/id/eprint/837441

Actions (login required)

View Item

Downloads