US 12,354,380 B2
Optical traffic lane recognition
Tamas Kapelner, Hildesheim (DE)
Assigned to Robert Bosch GmbH, Stuttgart (DE)
Filed by Robert Bosch GmbH, Stuttgart (DE)
Filed on Jul. 22, 2022, as Appl. No. 17/814,416.
Claims priority of application No. 10 2021 208 204.1 (DE), filed on Jul. 29, 2021.
Prior Publication US 2023/0029833 A1, Feb. 2, 2023
Int. Cl. G06V 20/56 (2022.01); G06V 10/26 (2022.01); G06V 10/44 (2022.01); G06V 10/774 (2022.01); G06V 10/82 (2022.01); G06V 20/58 (2022.01)
CPC G06V 20/588 (2022.01) [G06V 10/26 (2022.01); G06V 10/44 (2022.01); G06V 10/7747 (2022.01); G06V 10/82 (2022.01); G06V 20/584 (2022.01)] 15 Claims
OG exemplary drawing
 
1. A method for detecting at least one linear object in an input image, the method comprising:
feeding at least one of the input image and an extract of the input image to an image classifier configured to classify specified regions of the at least one of the input image and the extract of the input image, in each case at least into (i) relevant regions, a center point of which is located within a threshold distance from a center point of at least one linear object passing through the respective region, and (ii) background regions, a center point of which is not located within the threshold distance from the center point of the at least one linear object passing through the respective region;
acquiring, for the relevant regions, coordinates from a regression, which indicate at least one local course of the at least one linear object in the relevant regions; and
evaluating by extrapolation, based on the coordinates, a course of the at least one linear object in an entirety of the input image.