Self-Localization of a Mobile Robot Using Compressed Image Data of Average and Standard Deviation

Abstract

In this paper, an image-based self-localization method is proposed for a mobile robot. Images are compressed for each column, and the average and standard deviation of the pixels in each column are used. Environmental and observational data, which are the compressed image data at the registration and observational stages, are matched, and the position of the robot is obtained. The entire environment can be represented continuously with a small amount of data. A simple and robust matching method based on a voting process is introduced. The methods are evaluated through several experiments with omnidirectional images.