A Robust Descriptor for Tracking Vertical Lines in Omnidirectional Images and its Use in Robot Self-Calibration

This paper presents a robust descriptor for matching vertical lines among omnidirectional images and a method for automatically calibrating an omnidirectional camera with the robot reference system. The first part of this paper describes how to build the feature descriptor. We show that the descriptor is unique and distinctive for each feature and is invariant to rotation and slight changes of illumination. The robustness of the descriptor is validated through real experiments on a wheeled robot. The second part of the paper is devoted to the extrinsic self-calibration of the camera with the robot reference system. We show that by implementing an extended Kalman filter that fuses the information of the visual features with the odometry, it is possible to extrinsically and automatically calibrate the camera while the robot is moving.