US 11,676,391 B2
Robust correlation of vehicle extents and locations when given noisy detections and limited field-of-view image frames
Robert F. Cromp, Clifton, VA (US); and Jonathan Goldstein, Fredericksburg, VA (US)
Assigned to Raytheon Company, Waltham, MA (US)
Filed by Raytheon Company, Waltham, MA (US)
Filed on Feb. 19, 2021, as Appl. No. 17/180,111.
Claims priority of provisional application 63/010,979, filed on Apr. 16, 2020.
Prior Publication US 2021/0326645 A1, Oct. 21, 2021
Int. Cl. G06F 17/10 (2006.01); G06F 18/214 (2023.01); G06T 7/70 (2017.01); G06N 3/04 (2023.01); G06T 7/20 (2017.01); G06V 20/13 (2022.01)
CPC G06F 18/214 (2023.01) [G06N 3/04 (2013.01); G06T 7/20 (2013.01); G06T 7/70 (2017.01); G06V 20/13 (2022.01); G06T 2207/30244 (2013.01); G06T 2207/30248 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
accessing a plurality of image frames from one or more lenses of a camera system;
identifying, within the plurality of image frames, a plurality of vehicle front and a plurality of vehicle back detections;
pairing at least a subset of the plurality of vehicle back detections with vehicle front detections, wherein a given vehicle back detection is paired with a given vehicle front detection based on a camera angle relative to a predefined axis;
assigning, using each of a plurality of pools, a score to each vehicle front detection—vehicle back detection pair, each non-paired vehicle front detection, and each non-paired vehicle back detection, wherein each pool comprises a data structure representing a scoring mechanism and a set of detections, wherein the score is assigned using a scoring function that is based on an alignment between a camera orientation and a direction of travel;
assigning each detection to a pool that assigned a highest score to that detection; and
upon determining that a given pool comprises at least n detections: labeling the given pool as representing a specific vehicle, wherein n is a predetermined positive integer.