Method, apparatus, and system for detecting degraded ground paint in an image

Patent metadata

Field	Value
Publication number	US-11023747-B2
Application number	US-201916293328-A
Country	US
Kind code	B2
Filing date	Mar 5, 2019
Priority date	Mar 5, 2019
Publication date	Jun 1, 2021
Grant date	Jun 1, 2021

Abstract

An approach is provided for detecting degraded ground paint in an image. The approach, for example, involves performing semantic segmentation on the image to determine one or more pixels of the image that are classified in a ground paint category. The approach also involves generating a binary image that contains the one or more pixels of the image that are classified in the ground paint category. The approach further involves generating a hole-filled binary image by filling in the binary image to generate one or more curvilinear structures from the one or more pixels. The approach further involves determining a difference between the image and the hole-filled binary image to identify one or more degraded ground paint pixels of the image and providing the one or more degraded ground paint pixels as an output.

First claim

What is claimed is:

1. A computer-implemented method for detecting degraded ground paint in an image comprising:
performing semantic segmentation on the image to determine one or more pixels of the image that are classified in a ground paint category;
generating a binary image that contains the one or more pixels of the image that are classified in the ground paint category;
generating a hole-filled binary image by filling in the binary image to generate one or more curvilinear structures from the one or more pixels;
determining a difference between the image and the hole-filled binary image to identify one or more degraded ground paint pixels of the image; and
providing the one or more degraded ground paint pixels as an output.

2. The method of claim 1 , further comprising:
replacing a color value of the one or more degraded ground paint pixels in the image with a ground paint color value to create an output image.

3. The method of claim 2 , further comprising:
determining the ground paint color value from a respective color value of the one or more pixels of the image that are classified in the ground paint category.

4. The method of claim 3 , wherein the one or more pixels of the image from which the ground value is determined are within a proximity threshold of the one or more degraded ground paint pixels.

5. The method of claim 2 , wherein the ground paint color value is an aggregate value based on the respective color value of the one or more pixels of the image that are classified in the ground paint category.

6. The method of claim 2 , further comprising:
providing the output image for feature labeling.

7. The method of claim 6 , wherein the feature labeling is used from determining a ground control point.

8. The method of claim 1 , further comprising:
identifying the image as a faded ground paint image based on the one or more degraded ground paint pixels.

9. The method of claim 8 , wherein the faded ground paint image is excluded from feature labeling.

10. An apparatus for detecting degraded ground paint in an image comprising:
at least one processor; and
at least one memory including computer program code for one or more programs,
the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform at least the following,
perform semantic segmentation on the image to determine one or more pixels of the image that are classified in a ground paint category;
generate a binary image that contains the one or more pixels of the image that are classified in the ground paint category;
generate a hole-filled binary image by filling in the binary image to generate one or more curvilinear structures from the one or more pixels;
determine a difference between the image and the hole-filled binary image to identify one or more degraded ground paint pixels of the image; and
provide the one or more degraded ground paint pixels as an output.

11. The apparatus of claim 10 , wherein the apparatus is further caused to:
replace a color value of the one or more degraded ground paint pixels in the image with a ground paint color value to create an output image.

12. The apparatus of claim 11 , wherein the apparatus is further caused to:
determine the ground paint color value from a respective color value of the one or more pixels of the image that are classified in the ground paint category.

13. The apparatus of claim 12 , wherein the one or more pixels of the image from which the ground value is determined are within a proximity threshold of the one or more degraded ground paint pixels.

14. The apparatus of claim 11 , wherein the ground paint color value is an aggregate value based on the respective color value of the one or more pixels of the image that are classified in the ground paint category.

15. The apparatus of claim 11 , wherein the apparatus is further caused to:
provide the output image for feature labeling.

16. A non-transitory computer-readable storage medium for detecting degraded ground paint in an image, carrying one or more sequences of one or more instructions which, when executed by one or more processors, cause an apparatus to perform:
performing semantic segmentation on the image to determine one or more pixels of the image that are classified in a ground paint category;
generating a binary image that contains the one or more pixels of the image that are classified in the ground paint category;
generating a hole-filled binary image by filling in the binary image to generate one or more curvilinear structures from the one or more pixels;
determining a difference between the image and the hole-filled binary image to identify one or more degraded ground paint pixels of the image; and
providing the one or more degraded ground paint pixels as an output.

17. The non-transitory computer-readable storage medium of claim 16 , wherein the apparatus is caused to further perform:
replacing a color value of the one or more degraded ground paint pixels in the image with a ground paint color value to create an output image.

18. The non-transitory computer-readable storage medium of claim 17 , wherein the apparatus is caused to further perform:
determining the ground paint color value from a respective color value of the one or more pixels of the image that are classified in the ground paint category.

19. The non-transitory computer-readable storage medium of claim 18 , wherein the one or more pixels of the image from which the ground value is determined are within a proximity threshold of the one or more degraded ground paint pixels.

20. The non-transitory computer-readable storage medium of claim 17 , wherein the ground paint color value is an aggregate value based on the respective color value of the one or more pixels of the image that are classified in the ground paint category.

Assignees

• Here Global Bv

Inventors

• Pojman Nicholas
• Mittal Anish
• Lawlor David
• Chen Zhanwei

Classifications

• G06T7/10 Primary

  Segmentation; Edge detection (motion-based segmentation G06T7/215) · CPC title

• G06V20/653

  by matching three-dimensional models, e.g. conformal mapping of Riemann surfaces · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V20/588

  Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road · CPC title

• G06F18/24137

  Distances to cluster centroïds · CPC title