US 12,080,060 B2
Method and system for indoor multipath ghosts recognition
Ruoyu Feng, Leuven (BE); Andre Bourdoux, Theux (BE); Hichem Sahli, Etterbeek (BE); and Sofie Pollin, Rotselaar (BE)
Assigned to Imec vzw, Leuven (BE); and Katholieke Universiteit Leuven, KU LEUVEN R&D, Leuven (BE)
Filed by Imec vzw, Leuven (BE); and Katholieke Universiteit Leuven, KU LEUVEN R&D, Leuven (BE)
Filed on Sep. 21, 2021, as Appl. No. 17/448,250.
Claims priority of application No. 20200713 (EP), filed on Oct. 8, 2020.
Prior Publication US 2022/0114363 A1, Apr. 14, 2022
Int. Cl. G06K 9/00 (2022.01); G01S 13/72 (2006.01); G01S 13/89 (2006.01); G06T 7/277 (2017.01); G06V 20/00 (2022.01)
CPC G06V 20/00 (2022.01) [G01S 13/723 (2013.01); G01S 13/89 (2013.01); G06T 7/277 (2017.01); G06T 2207/10044 (2013.01); G06T 2207/20061 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for multipath ghosts recognition for a multiple-input-multiple-output radar with collocated antennas, the method comprising:
generating a two-dimensional Range-Doppler map for consecutive radar data frames,
applying a temporal clustering algorithm to the consecutive radar data frames, and
applying a linear pattern extraction algorithm on the two-dimensional Range-Doppler map, wherein the two-dimensional Range-Doppler map comprises detections from a target, a first-order ghost, and a second-order ghost with respect to one wall reflector,
wherein the method further comprises recognizing, via the linear pattern extraction algorithm, a set of three points in the two-dimensional Range-Doppler map corresponding to the detections from the target, the first-order ghost, and the second-order ghost.