Method, apparatus, and computing device for lane recognition

1. A method for lane recognition comprising:

obtaining, from a monitoring device on a road, a video that records a plurality of vehicles travelling on the road; and

recognizing at least one lane in the video based on first vehicle trajectories of the vehicles in the video,

wherein the first vehicle trajectories are based on positions of each of the vehicles in a plurality of video frames of the video,

wherein each trajectory of the first vehicle trajectories corresponds to a respective vehicle of the vehicles,

wherein a lane type of each of the at least one lane is based on types of the vehicles and the first vehicle trajectories,

wherein the types are based on a vehicle type detection model, and

wherein the lane type indicates a type of a vehicle that can travel on the at least one lane.

2. The method of claim 1 , wherein the road comprises an intersection, wherein the method further comprises obtaining, from the monitoring device, an extension video that is recorded in a time period after the video, wherein an attribute of each of the at least one lane is based on the first vehicle trajectories and second vehicle trajectories of the vehicles in the extension video, and wherein the attribute comprises one or more of:

a right-turn lane;

a left-turn lane;

a straight-through lane;

a right-turn and straight-through lane;

a left-turn and straight-through lane; or

an all-purpose lane.

3. The method of claim 1 , further comprising:

calculating a distance between a vehicle trajectory of a first vehicle of the vehicles and an adjacent lane in the video, wherein the adjacent lane is in the at least one lane and is nearest to the vehicle trajectory of the first vehicle;

comparing the distance and a pixel width of the first vehicle;

and

recognizing, based on the vehicle trajectory of the first vehicle, a new lane that is in the at least one lane.

4. The method of claim 1 , further comprising calculating a traffic flow on each of the at least one lane.

5. The method of claim 1 , further comprising:

identifying that a quantity of recognized lanes is not equal to a preset quantity, wherein the preset quantity is an actual quantity of detected lanes in the video; and

obtaining, in response to identifying that the quantity of recognized lanes is not equal to the preset quantity, at least one corrected lane in the video based on the preset quantity and the first vehicle trajectories.

6. The method of claim 1 , further comprising:

receiving, from the monitoring device and in real time, a video stream comprising the video; or

periodically obtaining, from the monitoring device, the video.

7. A computing device for lane recognition and comprising:

a memory configured to store computer instructions; and

a processor coupled to the memory, wherein the computer instructions cause the processor to be configured to:

obtain, from a monitoring device on a road, a video that records a plurality of vehicles travelling on the road; and

recognize at least one lane in the video based on first vehicle trajectories of the vehicles in the video,

wherein the first vehicle trajectories are based on positions of each of the vehicles in a plurality of video frames of the video,

wherein each trajectory of the first vehicle trajectories corresponds to a respective vehicle of the vehicles,

wherein a lane type of each of the at least one lane is based on types of the vehicles and the first vehicle trajectories,

wherein the types are based on a vehicle type detection model, and

wherein the lane type indicates a type of a vehicle that can travel on the at least one lane.

8. The computing device of claim 7 , wherein the road comprises an intersection, wherein the computer instructions further cause the processor to be configured to obtain, from the monitoring device, an extension video that is recorded in a time period after the video, wherein an attribute of each of the at least one lane is based on the first vehicle trajectories and second vehicle trajectories of the vehicles in the extension video, and wherein the attribute comprises one or more of:

a right-turn lane;

a left-turn lane;

a straight-through lane;

a right-turn and straight-through lane;

a left-turn and straight-through lane; or

an all-purpose lane.

9. The computing device of claim 7 , wherein the computer instructions further cause the processor to be configured to:

calculate a distance between a vehicle trajectory of a first vehicle of the vehicles and an adjacent lane in the video, wherein the adjacent lane is in the at least one lane and is nearest to the vehicle trajectory of the first vehicle;

compare the distance and a pixel width of the first vehicle;

and

recognize, based on the vehicle trajectory of the first vehicle, a new lane that is in the at least one lane.

10. The computing device of claim 7 , wherein the computer instructions further cause the processor to be configured to calculate a traffic flow on each of the at least one lane.

11. The computing device of claim 7 , wherein the computer instructions further cause the processor to be configured to:

identify that a quantity of recognized lanes is not equal to a preset quantity, wherein the preset quantity is an actual quantity of detected lanes in the video; and

obtain at least one corrected lane in the video based on the preset quantity and the first vehicle trajectories.

12. The computing device of claim 7 , wherein the computer instructions further cause the processor to be configured to:

receive, from the monitoring device and in real time, a video stream comprising the video; or

periodically obtain, from the monitoring device, the video.

13. A computer program product comprising computer-executable instructions that are stored on a non-transitory computer-readable storage medium and that, when executed by a processor, cause a computing device to:

obtain, from a monitoring device on a road, a video that records a plurality of vehicles travelling on the road; and

recognize at least one lane in the video based on the first vehicle trajectories of the vehicles in the video, wherein the first vehicle trajectories are based on positions of each of the vehicles in a plurality of video frames of the video,

wherein each trajectory of the first vehicle trajectories corresponds to a respective vehicle of the vehicles,

wherein a lane type of each of the at least one lane based on types of the vehicles and the first vehicle trajectories,

wherein the types are based on a vehicle type detection model, and

wherein the lane type indicates a type of a vehicle that can travel on the at least one lane.

14. The computer program product of claim 13 , wherein the road comprises an intersection, wherein the computer-executable instructions further cause the computing device to obtain, from the monitoring device, an extension video that is recorded in a time period after the video, wherein an attribute of each of the at least one lane is based on the first vehicle trajectories and second vehicle trajectories of the vehicles in the extension video, and wherein the attribute comprises one or more of:

a right-turn lane;

a left-turn lane;

a straight-through lane;

a right-turn and straight-through lane;

a left-turn and straight-through lane; or

an all-purpose lane.

15. The computer program product of claim 13 , wherein the computer-executable instructions further cause the computing device to:

calculate a distance between a vehicle trajectory of a first vehicle and an adjacent lane in the video, wherein the adjacent lane is in the at least one lane and is nearest to the vehicle trajectory of the first vehicle;

compare the distance and a pixel width of the first vehicle;

and

recognize, based on the vehicle trajectory of the first vehicle, a new lane that is in the at least one lane.

16. The computer program product of claim 13 , wherein the computer-executable instructions further cause the computing device to calculate a traffic flow on each of the at least one lane.

17. The computer program product of claim 13 , wherein the computer-executable instructions further cause the computing device to:

identify that a quantity of recognized lanes is not equal to a preset quantity, wherein the preset quantity is an actual quantity of detected lanes in the video; and

obtain at least one corrected lane in the video based on the preset quantity and the first vehicle trajectories.

18. The computer program product of claim 13 , wherein the computer-executable instructions further cause the computing device to receive, from the monitoring device and in real time, a video stream comprising the video.

19. The computer program product of claim 13 , wherein the computer-executable instructions further cause the computing device to periodically obtain, from the monitoring device, the video.

20. The computer program product of claim 13 , wherein the vehicle type detection model is an artificial intelligence (AI) model.