US 12,243,199 B2
Methods and systems for using trained generative adversarial networks to impute 3D data for construction and urban planning
Ryan Knuffman, Danvers, IL (US)
Assigned to STATE FARM MUTUAL AUTOMOBILE INSURANCE COMPANY, Bloomington, IL (US)
Filed by STATE FARM MUTUAL AUTOMOBILE INSURANCE COMPANY, Bloomington, IL (US)
Filed on Mar. 27, 2024, as Appl. No. 18/618,673.
Application 18/618,673 is a continuation of application No. 18/091,254, filed on Dec. 29, 2022, granted, now 11,972,541.
Application 18/091,254 is a continuation of application No. 17/982,174, filed on Nov. 7, 2022, granted, now 11,995,805.
Application 17/982,174 is a continuation of application No. 17/031,580, filed on Sep. 24, 2020, granted, now 11,508,042, issued on Nov. 22, 2022.
Claims priority of provisional application 62/967,315, filed on Jan. 29, 2020.
Prior Publication US 2024/0242316 A1, Jul. 18, 2024
Int. Cl. G06T 5/77 (2024.01); G06N 3/045 (2023.01); G06N 3/088 (2023.01); G06T 7/579 (2017.01)
CPC G06T 5/77 (2024.01) [G06N 3/045 (2023.01); G06N 3/088 (2013.01); G06T 7/579 (2017.01); G06T 2207/10028 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method for using a trained generative adversarial network to improve construction volumetrics, comprising:
obtaining a semantically segmented point cloud associated with a terrain of a site;
generating a terrain point cloud for the site using the trained generative adversarial network to fill one or more gaps within the semantically segmented point cloud; and
determining a volumetric soil measurement of a portion of the site using the terrain point cloud.