Quality landmark selection from image streams based on distinctiveness and repeatability

Abstract

Mobile robots moving in broad workspace should have abilities to select usefulness landmarks for stable self-localization. In this paper, we proposed a method of landmark selection from image streams captured by a camera mounted on a mobile robot. To select stable visual landmarks for mobile robots, distinctiveness and repeatability are considered. Several neighboring feature points form a visual landmark, and its distinctiveness is evaluated in each image. After that, under the assumption that robots are able to seek such feasible landmarks actively, repeatability of the landmark is evaluated. Weighting techniques using feature position relationships is proposed, and landmark selection criteria using variation coefficient is defined. They result to select high visibility landmarks. Proof experiments using a real mobile robot shows the effectiveness of the proposed method.