PROCEDE DE DETERMINATION D'UNE REGION D'INTERET DANS UNE IMAGE

• US 20110013839A1
• Filed: 07/07/2010
• Published: 01/20/2011
• Est. Priority Date: 07/08/2009
• Status: Active Grant

• Current Assignee: Valeo Vision (Valeo SE)
• Original Assignee: Valeo Vision (Valeo SE)
• Inventors: HUE, David, BENSRHAIR, Abdelaziz, AHIAD, Samia
• Granted Patent: US 8,520,898 B2
• Time in Patent Office: Days
• US Class Current: 382/173
• CPC Class Codes: G01N 21/538 for determining atmospheric attenuation and visibility
• CPC Class Codes: G06T 2207/30252 Vehicle exterior; Vicinity of vehicle
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/187 involving region growing; involving region merging; involving connected component labelling
• 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

A system and method for determining a region of interest in an image, characterized in that it comprises the following steps:

• • a plurality of lines L_{i i=1, . . . , n} formed by pixels is defined in the image;
  • in each line L_i, a plurality of segments S_{j j=1, . . . , m} constituted by pixels is defined;
  • a luminance level NVG(S_j) is calculated for each segment S_j based on the luminance of the pixels constituting this segment S_j;
  • depending on this luminance level NVG(S_j) the segment S_j is selected or excluded from the region of interest; and
  • the region of interest is obtained by amalgamating all the segments selected for each line.

Claims

1. A method for determining a region of interest in an image, this method being implemented in a road vehicle, said method comprising the following steps:

a plurality of lines L_{i i=1, . . . , n} formed by pixels is defined in the image;

in each line L_i, a plurality of segments S_{j j=1, . . . , m} constituted by pixels is defined;

a luminance level NVG(S_j) is calculated for each segment S_j based on the luminance of the pixels constituting this segment S_j;

depending on this luminance level NVG(Sj) the segment S_j is selected or excluded from the region of interest; and

the region of interest is obtained by amalgamating all the segments selected for each line.

2. The method according to claim 1, wherein the luminance level NVG(S_j) is an average luminance level NVGaverage(S_j) corresponding to the average luminance of each pixel of this segment S_j.

3. The method according to claim 1, wherein the following steps are carried out to determine if the S_j segment is selected or not:

it is determined if the luminance level NVG(S_j) of this segment is comprised in the interval [NVGaverage(L_i-1)−

S;

NVGaverage(L_i-1)+S′

] with S and S′

being preset thresholds and NVGaverage(L_i-1) being the luminance level of one or more other adjacent lines;

if NVG(S_j) is comprised in this interval, then the segment S_j is selected; and

if NVG(S_j) is not comprised in this interval, then this segment S_j is excluded from the region of interest.

4. The method according to claim 1, wherein the definition of at least one segment S_j of a line L_i comprises the following steps:

a pixel of the line L_i is considered, the segment S_j is constituted by incrementing this pixel of the pixels that are consecutive and pertaining to this same line L_i until a predetermined maximum number of pixels is reached.

5. The method according to claim 4 wherein, prior to defining the segment S_j, contours are searched by a contour detecting method, and wherein incrementing of the segment S_j is stopped and this segment S_j is closed if the pixel incremented forms a contour.

6. The method according to claim 1, wherein the segment S_j is excluded from the region of interest if the length of this segment S_j is less than a predetermined minimum length.

7. The method according to claim 1, wherein a search window, inside which the plurality of lines L_i is defined, is isolated in the image.

8. The method according to claim 7, wherein incrementing of the segment S_j is stopped and this segment S_j is closed if the pixel incremented corresponds to an edge of the search window.

9. The method according to claim 7, wherein the search window is segmented into a plurality of areas and specific rules concerning the definition of the segments S_j and/or of the lines L_i are allotted to each area.

10. The method according to claim 9, wherein the search window comprises at least one first area and one second area respectively positioned horizontally around a first and a second point of interest in the image.

11. The method according to claim 10, wherein the method is implemented in a road vehicle and at least one point of interest is determined as being a vanishing point from a system for detecting a vanishing point based on a steering wheel angle of the vehicle and/or from data generated by a navigation system and/or from a system for detecting road marking lines.

12. The method according to claim 10, wherein the search window comprises a third area defined so as to ensure continuity between the first and second areas.

13. The method according to claim 12, wherein the search window is segmented so that the third area corresponds in the image to a transition between the road and the sky.

14. The method according to claim 12, wherein the limits between the areas of the search window are vertically positioned according to data relating to pitching of the vehicle or so that the third area is vertically centered around a vanishing point of the image.

15. A system for determining a region of interest in an image, said system comprising a device for acquiring an image and means of processing arranged to implement the method according to claim 1.

16. The method according to claim 8, wherein the search window is segmented into a plurality of areas and specific rules concerning the definition of the segments S_j and/or of the lines L_i are allotted to each area.

17. The method according to claim 11, wherein the search window comprises a third area defined so as to ensure continuity between the first and second areas.

18. The method according to claim 13, wherein the limits between the areas of the search window are vertically positioned according to data relating to pitching of the vehicle or so that the third area is vertically centered around a vanishing point of the image.

19. The method according to claim 2, wherein the following steps are carried out to determine if the S_j segment is selected or not:

it is determined if the luminance level NVG(S_j) of this segment is comprised in the interval [NVGaverage(L_i-1)−

S;

NVGaverage(L_i-1)+S′

] with S and S′

being preset thresholds and NVGaverage(L_i-1) being the luminance level of one or more other adjacent lines;

if NVG(S_j) is comprised in this interval, then the segment S_j is selected; and

if NVG(S_j) is not comprised in this interval, then this segment S_j is excluded from the region of interest

20. The method according to claim 2, wherein the definition of at least one segment S_j of a line L_i comprises the following steps:

a pixel of the line L_i is considered, the segment S_j is constituted by incrementing this pixel of the pixels that are consecutive and pertaining to this same line L_i until a predetermined maximum number of pixels is reached.

21. The method according to claim 3, wherein the definition of at least one segment S_j of a line L_i comprises the following steps:

a pixel of the line L_i is considered, the segment S_j is constituted by incrementing this pixel of the pixels that are consecutive and pertaining to this same line L_i until a predetermined maximum number of pixels is reached.

22. A system for aiding a driver to drive a road vehicle comprising:

means for defining a plurality of lines L_{i i=1, . . . , n} formed by pixels is defined in the image;

means for defining in each line L_i, a plurality of segments S_{j j=1, . . . , m} constituted by pixels;

means for collecting a luminance level NVG(S_j) for each segment S_j based on the luminance of the pixels constituting this segment S_j;

means for selecting or excluding the segment S_j from the region of interest in response to this luminance level NVG(Sj); and

means for amalgamating all the segments selected for each line to obtain the region of interest.

23. The system according to claim 22, wherein the luminance level NVG(S_j) is an average luminance level NVGaverage(S_j) corresponding to the average luminance of each pixel of this segment S_j.

24. The system according to claim 22, wherein the following steps are carried out to determine if the S_j segment is selected or not:

it is determined if the luminance level NVG(S_j) of this segment is comprised in the interval [NVGaverage(L_i-1)−

S;

NVGaverage(L_i-1)+S′

] with S and S′

being preset thresholds and NVGaverage(L_i-1) being the luminance level of one or more other adjacent lines;

if NVG(S_j) is comprised in this interval, then the segment S_j is selected; and

if NVG(S_j) is not comprised in this interval, then this segment S_j is excluded from the region of interest.