Place recognition based on Gabor filters and SVMs for mobile robot self-localization

This paper proposes a novel appearance-based place recognition approach for mobile robot localization. The multi-channel Gabor filters is used to extract the global texture features of the scene images which are associated with the corresponding locations, and then these texture features are fed to support vector machine classifiers to determine the location of mobile robot. The algorithm has been tested on our designed and developed autonomous mobile robot CASIA-I. The experiment results indicate that the algorithm can reach up to 91.11% correct location rate, be robust to the various illumination and contrast, and satisfy the real-time self-localization demand of mobile robot.