Planar flow-based planar region extraction for visual navigation using a single on-board camera

Abstract

Dominant plane detection is an essential task for autonomous navigation of mobile robots equipped with a vision system, as we assume that the mobile robot moves on the dominant plane. In this paper, the image point motion on a space plane is modeled, and the conception of planar flow which describes the image motion pattern of the points on a plane is introduced. Planar flow provides another geometric constraint on optical flow when there is planar region in image. Based on planar flow, the distance error function in image space between a point and the plane is defined, which can be used for planar point detection. To improve the robustness of planar region detection, some other prior knowledge, such as the pixel ratio or region size, can be combined with the basic process. Experiments are conducted to validate the proposed method. The experiments show that the proposed method can detect the planar region with low texture image.