US 12,073,723 B2
Intelligent park assist system with sign interpretation
Lawrence Andrew Bush, Shelby Township, MI (US); Prabhjot Kaur, Auburn Hills, MI (US); and Alireza Esna Ashari Esfahani, Novi, MI (US)
Assigned to GM GLOBAL TECHNOLOGY OPERATIONS LLC, Detroit, MI (US)
Filed by GM Global Technology Operations LLC, Detroit, MI (US)
Filed on Jan. 5, 2023, as Appl. No. 18/150,507.
Prior Publication US 2024/0233537 A1, Jul. 11, 2024
Int. Cl. G08G 1/14 (2006.01); G06V 10/84 (2022.01); G06V 20/58 (2022.01); H04W 4/44 (2018.01)
CPC G08G 1/143 (2013.01) [G06V 10/85 (2022.01); G06V 20/582 (2022.01); G06V 20/586 (2022.01); H04W 4/44 (2018.02)] 15 Claims
OG exemplary drawing
 
1. A method for providing parking assistance, comprising:
receiving, by a controller of a vehicle, sign information of one or more traffic signs in an area surrounding the vehicle, wherein the sign information is a time series of data acquired by a set of cameras as the vehicle moves, using the vehicle as a frame of reference;
identifying, by the controller, a potential parking spot in an area surrounding the vehicle;
determining, by the controller, using the sign information, whether the potential parking spot is valid or invalid for the vehicle by determining, for each of the one or more traffic signs, whether the traffic sign is relevant to the potential parking spot using a hidden Markov model and, for each traffic sign that is relevant to the potential parking spot, analyzing parking restriction parameters of the traffic sign by:
determining, for each time step of a plurality of time steps, whether the traffic sign is detected at each of a plurality of directional bins with respect to the vehicle;
classifying a detected traffic sign with a corresponding sign indicator;
clustering, for each time step of the plurality of time steps, corresponding sign indicators in appropriate directional bins;
determining a corresponding histogram over the plurality of time steps of traffic sign detections for each directional bin of the plurality of bins; and
predicting traffic sign detections at each directional bin using the histograms and a calculated vehicle movement to determine whether the traffic sign is relevant using the hidden Markov model; and
generating, by the controller, a notification that the potential parking spot is invalid.