US 12,080,071 B2
Strong laser scanner matching method and apparatus considering movement of ground robot
Hyeon Myeong, Daejeon (KR); Hyungjin Kim, Daejeon (KR); and Seungwon Song, Daejeon (KR)
Assigned to Korea Advanced Institute of Science and Technology, Daejeon (KR)
Appl. No. 17/266,520
Filed by KOREA ADVANCED INSTITUTE OF SCIENCE AND TECHNOLOGY, Daejeon (KR)
PCT Filed Nov. 15, 2019, PCT No. PCT/KR2019/015640
§ 371(c)(1), (2) Date Feb. 5, 2021,
PCT Pub. No. WO2020/101424, PCT Pub. Date May 22, 2020.
Claims priority of application No. 10-2018-0141636 (KR), filed on Nov. 16, 2018; and application No. 10-2019-0062980 (KR), filed on May 29, 2019.
Prior Publication US 2021/0303908 A1, Sep. 30, 2021
Int. Cl. G06V 20/56 (2022.01); G01S 17/89 (2020.01)
CPC G06V 20/56 (2022.01) [G01S 17/89 (2013.01)] 8 Claims
OG exemplary drawing
 
1. A scan matching method comprising:
receiving two point clouds generated by a 3D light detection and ranging (LiDAR) sensor of an autonomous ground vehicle or a mobile ground robot;
sequentially inspecting height values until a point is within a specific height range to search for correspondence points between the two point clouds; and
searching for the correspondence points, wherein the searching includes:
dividing the height range into layers along a z-axis;
searching for a correspondence point between the two point clouds in three of the layers based on the height range of each of the three layers, wherein the three layers include a relevant layer, an upper layer above the relevant layer along the z-axis, and a lower layer below the relevant layer along the z-axis;
searching for nearest points respectively in the relevant layer, the upper layer, and the lower layer; and
selecting the nearest point among the searched three points as the correspondence point; and
performing scan matching between the two point clouds based on the search result of the correspondence points.