US 12,340,524 B2
Digital video computing system for vehicle
Igor Janjic, Newport News, VA (US)
Assigned to AURORA FLIGHT SCIENCES CORPORATION, A SUBSIDIARY OF THE BOEING COMPANY, Manassas, VA (US)
Filed by Aurora Flight Sciences Corporation, Manassas, VA (US)
Filed on Apr. 7, 2022, as Appl. No. 17/658,369.
Claims priority of provisional application 63/202,144, filed on May 28, 2021.
Prior Publication US 2022/0383515 A1, Dec. 1, 2022
Int. Cl. G06T 7/73 (2017.01); G06T 7/246 (2017.01); G06T 7/269 (2017.01)
CPC G06T 7/248 (2017.01) [G06T 7/269 (2017.01); G06T 7/73 (2017.01); G06T 2207/10021 (2013.01); G06T 2207/30252 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A digital video computing system, comprising:
a logic subsystem; and
a storage subsystem holding instructions executable by the logic subsystem to:
receive, from a camera system of a vehicle, two or more frames depicting an environment;
identify a salient image feature in the two or more frames;
calculate a global motion vector indicative of movement of the salient image feature between the two or more frames that is at least partially attributable to a movement of the vehicle between capture of each frame of the two or more frames;
calculate a local motion vector indicative of movement of the salient image feature between the two or more frames that is independent from the movement of the vehicle between capture of each frame of the two or more frames;
determine that the salient image feature has an apparent motion relative to the environment that is independent from the movement of the vehicle between capture of each frame of the two or more frames, based at least on the local motion vector for the salient image feature;
identify a candidate image patch within the two or more frames, the candidate image patch including the salient image feature, wherein the candidate image patch is a two-channel image including a greyscale channel and a background-subtracted channel, such that the candidate image patch includes a first plurality of pixel values corresponding to the greyscale channel and a second plurality of pixel values corresponding to the background-subtracted channel; and
analyze the candidate image patch to output a likelihood that the candidate image patch depicts a second vehicle moving through the environment.