US 12,444,004 B2
Systems and methods for road infrastructure management
Kanishk Varshney, Gurgaon (IN); Mihir Rajendra Kulkarni, Bangalore (IN); and Shrijit Basak, Bangalore (IN)
Assigned to AIDash Inc., Palo Alto, CA (US)
Filed by AIDash Inc., Palo Alto, CA (US)
Filed on May 24, 2024, as Appl. No. 18/674,746.
Claims priority of provisional application 63/559,846, filed on Feb. 29, 2024.
Claims priority of provisional application 63/549,422, filed on Feb. 2, 2024.
Prior Publication US 2025/0252514 A1, Aug. 7, 2025
Int. Cl. G06Q 50/08 (2012.01); G06T 7/11 (2017.01); G06V 20/00 (2022.01); G06V 30/19 (2022.01)
CPC G06Q 50/08 (2013.01) [G06T 7/11 (2017.01); G06V 20/00 (2022.01); G06V 30/19 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A non-transitory computer-readable medium comprising executable instructions, the executable instructions being executable by one or more processors to perform a method, the method comprising:
receiving multiple 360-degree images, each 360-degree image including a portion of a road, a 360-degree image captured by one or more 360-degree camera systems and having a timestamp and a location at which the one or more 360-degree camera systems captured the 360-degree image;
generating multiple images based on the multiple 360-degree images, an image including the portion of the road in a corresponding 360-degree image, the multiple images ordered by timestamps of the corresponding 360-degree images;
applying one or more road defect detection models trained to detect defects in a road to the multiple images, the road defect detection models employing similarity metrics and crack geometry to identify cracks detected across the multiple images;
detecting, based on applying the one or more road defect detection models to the multiple images, that there is a first defect in a first portion of the road in a first image and that there is a second defect in a second portion of the road in a second image, the second image ordered immediately after the first image;
receiving a first confidence score for the first defect;
receiving a second confidence score for the second defect;
determining that an intersection over union for the first defect and the second defect exceeds a threshold;
determining that the first defect and the second defect are part of a single defect based on the first confidence score, the second confidence score, and the intersection over union;
receiving at least one of a first location for a first 360-degree image corresponding to the first image and a second location for a second 360-degree image corresponding to the second image;
estimating, based on at least one of the first location and the second location, one or more estimated locations of the single defect;
generating defect location data that indicates the one or more estimated locations of the single defect;
generating a defect report image that includes the single defect and identifies the single defect; and
providing the defect location data and the defect report image, thereby allowing defects in the road to be identified.