Translator Disclaimer
6 May 2009 Optical signatures for autonomous mobility
Author Affiliations +
Active sensing has long been used for autonomous navigation in ground vehicles and involves the classification of objects based primarily on 3D spatial information. Differentiating critical targets such as humans from other objects and terrain has proven difficult due to the diversity of human spatial configurations that could be present in a given scene. Our approach to this problem and classification in general is to augment existing sensor technology with spatially independent information. In this paper we report on new signatures-based enhancements to the current technology to enhance classification with additional aspects of discrimination. The experimental investigation involved the measurement and evaluation of the polarization characteristics of objects of importance to short range autonomous ground mobility. A monostatic polarimeter was constructed to operate at 1550 nm and analyze the backscatter from materials over a wide range of incident angles. Several materials were analyzed with special focus on discerning humans from foliage clutter. The results are discussed here in the context of current UGV perception technology and applications. We have further interest in extending this concept to include active spectropolarimetric sensing and present results from numerical calculations.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Michael A. Powers and Christopher C. Davis "Optical signatures for autonomous mobility", Proc. SPIE 7324, Atmospheric Propagation VI, 73240U (6 May 2009);


Back to Top