论文部分内容阅读
A framework is presented for robustly estimating the location of a mobile robot in urban areas based on images extracted from a monocular onboard camera, given a 2D map with building outlines with neither 3D geometric information nor appearance data. T