Fig 1.
(a)The relative spatial relationship between the world coordinate system, the ground coordinate system, and the robot coordinate system. (b) The camera coordinate system and the image coordinate system.
Fig 2.
Processing flow chart of the proposed method.
Fig 3.
Extended image coordinate system and sub-regions.
Fig 4.
Subregions of camera coordinate system.
(a) xOz plane view. (b) yOz plane view.
Fig 5.
Relationship between plane primitive and the field of view.
Table 1.
Intersection status check table.
Fig 6.
The plane constraint for the mobile robot.
Fig 7.
The plane used for positioning is unidirectional.
Fig 8.
Prototype of the mobile robot system.
(a)Overall view of the mobile robot prototype. (b)The four Mecanum wheels of the mobile robot. (c)The structure of the mobile robot system.
Fig 9.
The layout of the experiment workspace and the point cloud example for localization.
(a)Experimental workspace and the identified numbers of the planes and their vertices. (b)Overall view of the workspace. (c)Example of observed planes in view. (d)Example of the filtered ROIs when identifying the planes.
Table 2.
The structured parameters of a-priori planes.
Fig 10.
(a)The change of the plane region to the view field region (VFR). (b)Size of point cloud inside filter ROI when the supposed initial x-coordinate of the mobile robot changes.
Fig 11.
(a)The change of the plane region to the view field region (VFR). (b)Size of point cloud inside filter ROI when the supposed initial y-coordinate of the mobile robot changes.
Fig 12.
(a)The change of the plane region to the view field region (VFR). (b)Size of point cloud inside filter ROI when the supposed initial pose angle of the mobile robot changes.
Fig 13.
(a)The change of the plane region to the view field region (VFR). (b)Size of point cloud inside filter ROI when the supposed initial x-coordinate of the mobile robot changes.
Fig 14.
(a)The change of the plane region to the view field region (VFR). (b)Size of point cloud inside filter ROI when the supposed initial y-coordinate of the mobile robot changes.
Fig 15.
(a)The change of the plane region to the view field region (VFR). (b)Size of point cloud inside filter ROI when the supposed initial pose angle of the mobile robot changes.
Fig 16.
Path sequence of tracking experiment.
Fig 17.
Point cloud segmentation in the process of path tracing.
Fig 18.
Comparison before and after using proposed method.
Table 3.
Results of the comparative experiment.