	US 12,482,278 B2
	Method and apparatus for determining and classifying at least one object in a sensing area of a sensor
Martin Schumann, Berlin (DE)
Assigned to CARIAD SE, Wolfsburg (DE)
Appl. No. 17/915,932
Filed by CARIAD SE, Wolfsburg (DE)
PCT Filed Apr. 1, 2021, PCT No. PCT/EP2021/058667 § 371(c)(1), (2) Date Sep. 29, 2022, PCT Pub. No. WO2021/198446, PCT Pub. Date Oct. 7, 2021.
Claims priority of application No. 10 2020 109 364.0 (DE), filed on Apr. 3, 2020.
Prior Publication US 2023/0148091 A1, May 11, 2023
Int. Cl. G06V 20/64 (2022.01); G06V 10/764 (2022.01); G06V 10/82 (2022.01); G06V 20/56 (2022.01); G06V 20/58 (2022.01)

CPC G06V 20/64 (2022.01) [G06V 10/764 (2022.01); G06V 10/82 (2022.01); G06V 20/56 (2022.01); G06V 20/58 (2022.01)]

12 Claims

1. A method for determining and classifying at least one object in a sensing area of at least one sensor, the method comprising:

capturing an image of the sensing area using the at least one sensor, wherein the image is a two-dimensional image or a three-dimensional image,

generating sensor data corresponding to the image, and

determining the object in the image using the sensor data and first template data of at least one first template object using a neural network,

wherein the neural network has been trained using the first template data,

wherein the first template data corresponds to an image of the first template object,

wherein at least a first object class is assigned to the first template object,

wherein the object is classified by the neural network determining whether the object is to be assigned to the first object class,

wherein the object in the image is determined using the sensor data and second template data of a second template object using a neural network,

wherein the second template data corresponds to an image of the second template object to which at least a second object class is assigned, and

wherein the object is classified by the neural network determining whether the object is to be assigned to the second object class.