US 12,254,041 B2
Position recognition method and system based on visual information processing
Dongcheol Hur, Seongnam-si (KR); Yeong-Ho Jeong, Seongnam-si (KR); and Sangwook Kim, Seongnam-si (KR)
Assigned to NAVER CORPORATION, Seongnam-si (KR)
Filed by NAVER LABS CORPORATION, Seongnam-si (KR)
Filed on Apr. 4, 2022, as Appl. No. 17/712,849.
Application 17/712,849 is a continuation of application No. PCT/KR2020/016271, filed on Nov. 18, 2020.
Claims priority of application No. 10-2019-0168100 (KR), filed on Dec. 16, 2019.
Prior Publication US 2022/0230350 A1, Jul. 21, 2022
Int. Cl. G06F 16/587 (2019.01); G06F 16/583 (2019.01); G06T 7/246 (2017.01); G06T 7/70 (2017.01)
CPC G06F 16/587 (2019.01) [G06F 16/583 (2019.01); G06T 7/248 (2017.01); G06T 7/70 (2017.01)] 9 Claims
OG exemplary drawing
 
1. A position recognition method of a computer apparatus comprising at least one processor, the position recognition method, which uses the at least one processor, comprising:
generating a frame image through a camera;
transmitting a first global pose of the camera and the generated frame image to a server; and
receiving, from the server, a second global pose of the camera that is estimated based on a pose of an object included in the transmitted frame image,
wherein the receiving of the second global pose of the camera comprises:
receiving, from the server, the second global pose, a first confidence value corresponding to the second global pose, a third global pose as a global pose of the computer apparatus that is estimated based on a global localization at the server, and a second confidence value corresponding to the third global pose; and
determining a fourth global pose by assigning different weights based on the first confidence value and the second confidence value to the second global pose and the third global pose.