US 12,087,004 B2
Multi-collect fusion
Luca Del Pero, London (GB); and Karim Tarek Mahmoud Elsayed Ahmed Shaban, London (GB)
Assigned to Lyft, Inc., San Francisco, CA (US)
Filed by Lyft, Inc., San Francisco, CA (US)
Filed on Jun. 30, 2020, as Appl. No. 16/917,793.
Prior Publication US 2021/0407186 A1, Dec. 30, 2021
Int. Cl. G06T 7/579 (2017.01); G01C 21/36 (2006.01); G06T 3/14 (2024.01); G06T 7/13 (2017.01); G06T 7/33 (2017.01); G06T 7/73 (2017.01); G06T 17/05 (2011.01); G06V 20/56 (2022.01)
CPC G06T 7/579 (2017.01) [G01C 21/3602 (2013.01); G06T 3/14 (2024.01); G06T 7/13 (2017.01); G06T 7/33 (2017.01); G06T 7/74 (2017.01); G06T 17/05 (2013.01); G06V 20/56 (2022.01); G06T 2200/04 (2013.01); G06T 2207/20221 (2013.01); G06T 2207/30252 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
generating first structure data by processing first image data that was captured during a first collect using a photogrammetric imaging technique, wherein the first structure data comprises (i) a first three-dimensional (3D) representation of a first set of visible features captured in the first image data and (ii) first sequential pose data associated with the first image data;
generating further structure data by processing further image data that was captured during a further collect using a photogrammetric imaging technique, wherein the further structure data comprises (i) a further 3D representation of a further set of visible features captured in the further image data and (ii) further sequential pose data associated with the further image data, and wherein the further set of visible features overlaps at least in part with the first set of visible features and the further sequential pose data overlaps at least in part with the first sequential pose data;
identifying pose correlations between poses included in the first sequential pose data and poses included in the further sequential pose data;
identifying feature correlations between visible features included in the first set of visible features and visible features included in the further set of visible features;
based on the identified pose correlations and the identified feature correlations, determining a transformation of the further structure data relative to the first structure data; and
generating combined structure data by (i) using the determined transformation to align and warp the further structure data relative to the first structure data and (ii) fusing the further structure data and the first structure data.