Method and apparatus for early warning of vehicle offset

• US 10,759,448 B2
• Filed: 04/12/2019
• Issued: 09/01/2020
• Est. Priority Date: 09/30/2018
• Status: Active Grant

• Current Assignee: BOE Technology Group Co., Ltd.
• Original Assignee: BOE Technology Group Co., Ltd.
• Inventors: Lin, Hangdong
• Primary Examiner(s): Nguyen, Tai T
• Application Number:
  US16/383,554
• Publication Number:
  US 20200101983A1
• Time in Patent Office: 508 Days
• Field of Search: 340463, 340435, 340436, 340438, 340903, 340907, 340933, 340937, 701 28, 701 41, 701301, 701302
• CPC Class Codes: B60W 2050/143 Alarm means
• CPC Class Codes: B60W 2420/403 Image sensing, e.g. optical camera
• CPC Class Codes: B60W 2552/53 Road markings, e.g. lane marker or crosswalk
• CPC Class Codes: B60W 50/14 Means for informing the driver, warning the driver or prompting a driver intervention
• CPC Class Codes: G06T 2207/30256 Lane; Road marking
• CPC Class Codes: G06T 7/13 Edge detection
• CPC Class Codes: G06T 7/70 Determining position or orientation of objects or cameras camera calibration G06T7/80
• CPC Class Codes: G06V 20/588 Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
• CPC Class Codes: G08G 1/167 Driving aids for lane monitoring, lane changing, e.g. blind spot detection

Abstract

The present disclosure provides a method and apparatus for early warning of vehicle offset. The method includes: acquiring collected road image to be detected; obtaining a corresponding grayscale image by performing grayscale processing on the road image to be detected; determining a target lane line within the grayscale image; and determining whether to issue an early warning of vehicle offset according to the target lane line and an early warning trigger condition.

Claims

1. A method for early warning of vehicle offset, the method comprising:

acquiring collected road image to be detected;

obtaining a corresponding grayscale image by performing grayscale processing on the road image to be detected;

determining a target lane line within the grayscale image; and

determining whether to issue an early warning of vehicle offset according to the target lane line and an early warning trigger condition,

wherein the step of determining a target lane line within the grayscale image comprises;

determining a near field-of-view region of the grayscale image;

extracting a lane line feature image of the near field-of-view region;

obtaining the target lane line by fitting lane line features in the lane line feature image; and

acquiring a target lane line angle,

wherein the step of determining whether to issue an early warning of vehicle offset according to the target lane line and an early warning trigger condition comprises;

determining a first distance between the vehicle and the target lane line according to the target lane line; and

issuing the early warning of vehicle offset when the first distance satisfies a first threshold distance condition, and

wherein after the step of determining a target lane line within the grayscale image, the method further comprises;

extracting multiple feature points in the grayscale image;

inputting the multiple feature points into a preset detector to determine entity object information; and

determining a second distance between the vehicle and an entity object according to the entity object information, and issuing the early warning of vehicle offset when the second distance satisfies a second threshold distance condition.

2. The method according to claim 1, wherein the step of extracting a lane line feature image in the near field-of-view region comprises:

obtaining a near field-of-view image by intercepting the near field-of-view region of the grayscale image; and

extracting, through a steerable filter, the lane line feature image in the near field-of-view image.

3. The method according to claim 2, wherein the step of extracting a lane line feature image in the near field-of-view image when the road image to be detected is a first frame image comprises:

processing the near field-of-view image to generate a gradient image;

acquiring multiple feature angles within the gradient image; and

inputting the multiple feature angles into the steerable filter, and processing the near field-of-view image by the steerable filter to obtain the lane line feature image.

4. The method according to claim 3, wherein the step of extracting a lane line feature image in the near field-of-view image when the road image to be detected is not the first frame image comprises:

acquiring a target lane line angle acquired according to a road image to be detected in a previous frame; and

inputting the target lane line angle of the road image to be detected in the previous frame into the steerable filter, and processing the near field-of-view image by the steerable filter to obtain the lane line feature image.

5. The method according to claim 1, wherein the entity object information is the distance information between the entity object and the vehicle.

6. An apparatus for early warning of vehicle offset, comprising:

a processor;

a memory having stored thereon instructions which, when executed by the processor, cause the processor to;

acquire collected road image to be detected;

obtain a corresponding grayscale image by performing grayscale processing on the road image to be detected;

determine a target lane line within the grayscale image; and

determine whether to issue an early warning of vehicle offset according to the target lane line and an early warning trigger condition,

wherein the instructions, when executed by the processor, further cause the processor to;

determine a near field-of-view region of the grayscale image;

extract a lane line feature image of the near field-of-view region;

obtain the target lane line by fitting lane line features in the lane line feature image; and

acquire a target lane line angle,

wherein the instructions, when executed by the processor, further cause the processor to;

determine a first distance between the vehicle and the target lane line according to the target lane line; and

issue the early warning of vehicle offset when the first distance satisfies a first threshold distance condition, and

wherein the instructions, when executed by the processor, further cause the processor to;

extract multiple feature points in the grayscale image;

input the multiple feature points into a preset detector to determine entity object information; and

determine a second distance between the vehicle and an entity object according to the entity object information, and issue the early warning of vehicle offset when the second distance satisfies a second threshold distance condition.

7. The apparatus according to claim 6, wherein the instructions, when executed by the processor, further cause the processor to:

obtain a near field-of-view image by intercepting the near field-of-view region of the grayscale image; and

extract, through a steerable filter, the lane line feature image in the near field-of-view image.

8. The apparatus according to claim 7, wherein the instructions, when executed by the processor, further cause the processor to:

process the near field-of-view image to generate a gradient image;

acquire multiple feature angles within the gradient image; and

input the multiple feature angles into the steerable filter, and process the near field-of-view image by the steerable filter to obtain the lane line feature image.

9. The apparatus according to claim 7, wherein the instructions, when executed by the processor, further cause the processor to:

acquire a target lane line angle acquired according to a road image to be detected in a previous frame; and

input the target lane line angle into the steerable filter, and process the near field-of-view image by the steerable filter to obtain the lane line feature image.

10. A vehicle, comprising:

a sensor configured to collect a road image to be detected;

a processor communicatively coupled to the sensor;

a memory having stored thereon instructions which, when executed by the processor, cause the processor to;

acquire collected road image to be detected from the sensor;

obtain a corresponding grayscale image by performing grayscale processing on the road image to be detected;

determine a target lane line within the grayscale image; and

determine whether to issue an early warning of vehicle offset according to the target lane line and an early warning trigger condition,

wherein the instructions, when executed by the processor, further cause the processor to;

determine a near field-of-view region of the grayscale image;

extract a lane line feature image of the near field-of-view region;

obtain the target lane line by fitting lane line features in the lane line feature image; and

acquire a target lane line angle,

wherein the instructions, when executed by the processor, further cause the processor to;

determine a first distance between the vehicle and the target lane line according to the target lane line; and

issue the early warning of vehicle offset when the first distance satisfies a first threshold distance condition, and

wherein the instructions, when executed by the processor, further cause the processor to;

extract multiple feature points in the grayscale image;

input the multiple feature points into a preset detector to determine entity object information; and

determine a second distance between the vehicle and an entity object according to the entity object information, and issue the early warning of vehicle offset when the second distance satisfies a second threshold distance condition.

11. The vehicle according to claim 10, wherein the instructions, when executed by the processor, further cause the processor to:

obtain a near field-of-view image by intercepting the near field-of-view region of the grayscale image; and

extract, through a steerable filter, the lane line feature image in the near field-of-view image.

12. The vehicle according to claim 10, wherein the instructions, when executed by the processor, further cause the processor to:

process the near field-of-view image to generate a gradient image;

acquire multiple feature angles within the gradient image; and

input the multiple feature angles into the steerable filter, and process the near field-of-view image by the steerable filter to obtain the lane line feature image.