Image processing device for dividing an image into a plurality of regions

• US 8,498,479 B2
• Filed: 02/25/2010
• Issued: 07/30/2013
• Est. Priority Date: 03/24/2009
• Status: Active Grant

• Current Assignee: Subaru Corp.
• Original Assignee: Fuji Jukogyo KK (Subaru Corp.)
• Inventors: Saito, Toru
• Primary Examiner(s): GORADIA, SHEFALI DINESH
• Application Number:
  US12/659,115
• Publication Number:
  US 20100246896A1
• Time in Patent Office: 1,251 Days
• Field of Search: 382/106, 382/173, 382/100
• US Class Current: 382/173
• CPC Class Codes: G06T 2207/30252 Vehicle exterior; Vicinity of vehicle
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06V 20/584 of vehicle lights or traffic lights
• 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

Abstract

An image processing device for dividing an image imaged by imaging means into multiple regions includes: processing means for grouping, when difference of the pieces of image data between a single pixel within the image and a pixel adjacent thereto is less than a predetermined threshold, the single pixel and the adjacent pixel, and dividing the image into multiple regions with finally obtained each group as each region of the image; and average-value calculating means for calculating the average value of the image data within the group including the single pixel; with the processing means comparing the image data of the single pixel, and the average value calculated at the average-value calculating means regarding the group to which the adjacent pixel belongs; and when the difference thereof is equal to or greater than a predetermined second threshold, doing not group the single pixel and the adjacent pixel.

Claims

1. An image processing device configured to divide an image imaged by imaging means into a plurality of regions, said image processing device comprising:

processing means configured to group, in a case that a difference of pieces of image data between a single pixel within said image and a pixel adjacent thereto is less than a predetermined first threshold, said single pixel and said adjacent pixel, and to divide said image into a plurality of regions with finally obtained each group as each region of said image; and

average-value calculating means configured to calculate an average value of said image data within said group including said single pixel;

wherein said processing means compares the image data of said single pixel, and an average value calculated at said average-value calculating means regarding said group to which said adjacent pixel belongs to, and

wherein in a case that the difference thereof is equal to or greater than a predetermined second threshold, do not group said single pixel and said adjacent pixel.

2. The image processing device according to claim 1, wherein, with said processing means, upon the image data of said single pixel being transmitted from said imaging means, said image data of a pixel adjacent to said single pixel that comprises a pixel transmitted prior to said single pixel, of a same image as said single pixel, is compared with the image data of said single pixel, and the image data of said single pixel is compared with said average value of said group to which said adjacent pixel of which said image data was transmitted prior to said single pixel belongs to.

3. The image processing device according to claim 1, wherein, with said processing means, in a case that one of pieces of image data of adjacent said pixels are within a predetermined numeric value range, said adjacent pixels of which said image data is within said numeric value range are grouped in a preferred manner regardless of a comparison of pieces of image data between said single pixel and said adjacent pixels, and a comparison between the image data of said single pixel and said average value of said group.

4. The image processing device according to claim 1, wherein, with said processing means, in a case that, of pieces of the image data of adjacent said pixels, one piece of image data is within a predetermined numeric value range, and another piece of image data is out of said predetermined numeric value range, said adjacent pixels are not grouped regardless of a comparison of pieces of image data between said single pixel and said adjacent pixels, and a comparison between the image data of said single pixel and said average value of said group.

5. The image processing device according to claim 1, wherein, with said processing means, in a case that a minimum value of pixel intervals of two said groups that are not adjacent to each other on an image is equal to or smaller than a predetermined threshold, and a difference of average values of the two groups is equal to or smaller than a predetermined threshold, the two groups are integrated into a single group.

6. The image processing device according to claim 1, wherein said imaging means configured to output a pair of images comprises:

distance detecting means configured to subject said pair of images output from said imaging means to stereo matching processing to detect distance on a real space corresponding to at least a pixel of one image of said pair of images;

wherein, with said processing means, in a case that difference between said distance on a real space corresponding to said single pixel within said one image, and said distance on real space corresponding to a pixel adjacent to said single pixel is equal to or greater than a predetermined threshold, said single pixel and said adjacent pixel are not grouped.

7. The image processing device according to claim 6, wherein said processing means determines whether a difference between said distance on a real space closest to said distance on a real space corresponding to said single pixel of said distance on real space corresponding to each pixel belonging to said group to which said adjacent pixel belongs, and said distance on a real space corresponding to said single pixel is equal to or greater than said threshold.

8. The image processing device according to claim 6, wherein said processing means calculates coordinates of each point on real space corresponding to said each pixel based on said each distance on a real space corresponding to each pixel belonging to said group that said adjacent pixel belongs, calculates the inclination of the distribution of said each point, and determines whether a difference between a distance on a real space corresponding to said adjacent pixel estimated from said inclination, and said distance on real space corresponding to said single pixel is equal to or greater than said threshold.

9. The image processing device according to claim 6, wherein said average-value calculating means calculates one of an average value, a maximum value, and a minimum value of said each distance on a real space corresponding to each pixel belonging to said group;

and wherein said processing means compares one of an average value, a maximum value, and a minimum value of said distance on a real space calculated at said average-value calculating means regarding said group to which said adjacent pixel belongs to with said distance on a real space corresponding to said single pixel, and in a case that a difference thereof is equal to or greater than a predetermined threshold, do not group said single pixel and said adjacent pixel.

10. The image processing device according to claim 6, further comprising:

road surface detecting means configured to detect a road surface from an image imaged by said imaging means;

wherein said processing means calculates a height of each point on a real space corresponding to said each pixel based on said each distance on a real space corresponding to each pixel belonging to said group, and in a case that a height of said each point from said road surface that said road surface detecting means detected is equal to or higher than a predetermined threshold, recognizes said group as a solid object existing on a road surface, and in a case that a height of said each point from said road surface that said road surface detecting means detected is lower than said predetermined threshold, recognizes said group as a pattern indicated on a road surface.

11. The image processing device according to claim 10, wherein, with said processing means, in a case that of a height from said road surface on a point on a real space corresponding to said single pixel, and a height from said road surface of said each point on a real space corresponding to each pixel belonging to said group to which said adjacent pixel belongs, one height is lower than said predetermined threshold, and another height is equal to or higher than said predetermined threshold, said single pixel and said adjacent pixel are not grouped.

12. The image processing device according to claim 6, wherein, with said processing means, in a case that, with two distributions of each point on a real space corresponding to each pixel belonging to two said groups that are not adjacent to each other on an image, a minimum value of an interval of two points on a real space belonging to a different distribution is equal to or smaller than a predetermined threshold, and in a case that a maximum value is equal to or smaller than a predetermined threshold that is a greater value than said threshold, the two groups are integrated into a single group.

13. The image processing device according to claim 6, wherein, with said processing means, in a case that, of two said groups that are not adjacent to each other on an image, a height from said road surface of said each point on real space corresponding to each pixel belonging to one of said groups is equal to or greater than said predetermined threshold, and in a case that a height from said road surface of said each point on real space corresponding to each pixel belonging to another of said groups is lower than said predetermined threshold, the two groups are not integrated.