Robust Feature Detection for a Mobile Robot using a Multi-View Single Camera

In this paper, we present a multi-view single camera and a novel model of scale and illumination invariant corner detection for a service robot in indoor environment. Vision-based simultaneous localization and mapping (VSLAM) has received much attention. It is used for VSLAM that are single cameras, multiple cameras in a stereo setup or omni-directional cameras. We propose a different approach which multiple mirrors are mounted on a vision system in a single-view-point (SVP) configuration. This vision system is easily to acquire no distortion image sequences without any preprocessing. Robust feature detection method is also described for VSLAM of a service robot in indoor environment. This method can detect scale and illumination invariant corner features in any environment.