US 12,452,537 B2
Automatic camera configuration systems and methods
Brian Kurz, Germantown, MD (US)
Assigned to Westinghouse Air Brake Technologies Corporation, Pittsburgh, PA (US)
Filed by Westinghouse Air Brake Technologies Corporation, Pittsburgh, PA (US)
Filed on Nov. 22, 2021, as Appl. No. 17/532,522.
Claims priority of provisional application 63/122,701, filed on Dec. 8, 2020.
Prior Publication US 2022/0182552 A1, Jun. 9, 2022
Int. Cl. H04N 23/695 (2023.01); H04N 23/661 (2023.01); H04N 23/698 (2023.01)
CPC H04N 23/695 (2023.01) [H04N 23/661 (2023.01); H04N 23/698 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising:
a camera having a field of view, the camera configured to acquire imaging information within the field of view;
a trained neural network model, wherein the trained neural network model is configured to compare identified objects or aspects within the acquired imaging information with expected results for one or more views corresponding to known locations and orientations associated with a vehicle; and
at least one processor coupled to the camera and configured to acquire, at a time of an installation process, the imaging information from the camera, wherein the installation process is a process of installing the camera onto the vehicle to position the camera in an installation location on the vehicle, wherein the at least one processor is configured to:
autonomously set, at the time of the installation process, vehicle configuration parameters using the trained neural network model based on the imaging information within the field of view;
determine that the field of view of the camera is not in compliance with a target specification during the installation process;
provide instructions to an operator to adjust the camera into compliance with the target specification in response to the processor determining that the field of view of the camera is not in compliance with the target specification during the installation process;
analyze a resulting field of view of the camera from the newly adjusted camera by using the trained neural network model after the operator follows the provided instructions during the installation process; and
provide additional instructions to the operator, during the installation process, to further adjust the camera until the field of view of the camera is in compliance with the target specification if the resulting field of view is not in compliance with the target specification.