A nonparametric learning approach to vision based mobile robot localization

A nonparametric learning algorithm is used to build a robust mapping between an image obtained from a mobile robot s on board camera and the robot s cur rent position The mapping uses unprocessed pixel values by pixel image Because the learning algorithm is nonparametric it uses the learn ing data obtained from these raw pixel values to au tomatically choose a structure for the mapping with out human intervention or any a priori assumptions about what type of image features should be used The learning data consisting of a series example image in puts and corresponding position values is collected in a calibration phase where the robot randomly tra verses its intended workspace This process of build ing visual localization maps for mobile robots is com pletely general and can be applied to any implemen tation which uses on board cameras We demonstrate the feasibility of this approach on a mobile platform performing in a robotics laboratory workspace This workspace is visually cluttered with humans and other objects continually moving within the robot s environ ment The mapping learned in this environment is ro bust to these dynamic visual features and consistently reports timely localization information at greater than Hz to within acceptable limits