US 12,437,557 B2
Method, apparatus, and computer program for extracting traffic signal information through image data analysis including traffic light
Hawook Jeong, Jeju-si (KR); Hee Kon Kim, Jeju-si (KR); and Jung Hee Park, Seoul (KR)
Assigned to RideFlux Inc., Jeju-si (KR)
Filed by RideFlux Inc., Jeju-si (KR)
Filed on Sep. 5, 2022, as Appl. No. 17/929,716.
Claims priority of application No. 10-2021-0118204 (KR), filed on Sep. 6, 2021.
Prior Publication US 2023/0071709 A1, Mar. 9, 2023
Int. Cl. G06V 20/58 (2022.01); G06V 10/25 (2022.01)
CPC G06V 20/584 (2022.01) [G06V 10/25 (2022.01)] 8 Claims
OG exemplary drawing
 
1. A method of extracting, by a computing apparatus, signal information through image data analysis including a traffic light, the method comprising:
capturing, by a camera module provided in a vehicle, an image in front of the vehicle, and collecting, by the computing apparatus connected to the camera module, image data of the captured image;
identifying a traffic light by analyzing the collected image data;
extracting signal information from the identified traffic light; and
providing, by the computing apparatus, to the vehicle that is waiting for a signal, a departure notification, based on the extracted signal information and movement information of the vehicle, or determining, by the computing apparatus, a driving control command of the vehicle, based on the extracted signal information,
wherein the identifying of the traffic light includes: selecting any one of a plurality of identified traffic lights by comparing an operation pattern of the vehicle for a predetermined period and a change pattern of the signal information, output from each of the plurality of identified traffic lights, for the predetermined period, when the plurality of traffic lights are identified by analyzing the collected image data; and setting a region corresponding to any one of the selected traffic lights as a region of interest (ROI) in the collected image data,
wherein the extracting of the signal information includes extracting the signal information on the vehicle by analyzing only the set ROI, and
wherein the identifying of the traffic light comprises: selecting any one of the plurality of identified traffic lights based on a traveling route of the vehicle when the plurality of traffic lights are identified by analyzing the collected image data; when two or more traffic lights are selected as being corresponding to the traveling route of the vehicle, selecting one traffic light of the two or more traffic lights based on an output form of the signal information and also based on road guide signs installed at locations adjacent to each of the two or more traffic lights; and setting a region corresponding to the selected one traffic light in the collected image data as the ROI.