作者: Jose-Luis Blanco , Javier Gonzalez , Juan-Antonio Fernandez-Madrigal
DOI: 10.1007/978-3-540-72849-8_25
关键词: Artificial intelligence 、 Occupancy grid mapping 、 Map matching 、 Polar coordinate system 、 Robotics 、 Scale-invariant feature transform 、 Computer vision 、 Image registration 、 Computer science 、 Pattern recognition 、 Robustness (computer science) 、 Occupancy
摘要: In this paper we propose a new matching method for occupancy grid-maps under the perspective of image registration. Our approach is based on extracting feature descriptors by means polar coordinate transformation around highly distinctive points. The proposed presents modest computation complexity, although it can find matchings between features reliably and regardless their orientation. Experimental results show robustness estimates even dynamic environments. proposal has important applications into field mobile robotics.