Vehicle detecting method and system

• US 9,524,557 B2
• Filed: 11/10/2014
• Issued: 12/20/2016
• Est. Priority Date: 12/24/2013
• Status: Active Grant

• Current Assignee: Hyundai Motor Company (Hyundai Motor Group), Industry Academic Cooperation Foundation, Kia Motors Corporation (Kia Corp.)
• Original Assignee: Hyundai Motor Company (Hyundai Motor Group), Industry Academic Cooperation Foundation, Kia Motors Corporation (Kia Corp.)
• Inventors: Hwang, Jin Kyu, Choi, Jin Ha, Kim, Dae Hwan, Kim, Jae Sung, Lee, Kyung Jae, Lee, Sang Youn
• Primary Examiner(s): Koziol, Stephen R
• Assistant Examiner(s): Lee, Jonathan S
• Application Number:
  US14/537,331
• Publication Number:
  US 20150178911A1
• Time in Patent Office: 771 Days
• US Class Current: 1/1
• CPC Class Codes: G06T 2200/04 involving 3D image data
• CPC Class Codes: G06T 2207/10021 Stereoscopic video; Stereoscopic image sequence
• CPC Class Codes: G06T 2207/30256 Lane; Road marking
• CPC Class Codes: G06T 2207/30261 Obstacle
• CPC Class Codes: G06T 7/593 from stereo images
• CPC Class Codes: G06T 7/73 using feature-based methods

Abstract

A vehicle detecting system and method are provided. The method includes creating, by a controller, a disparity map using images photographed by a stereoscopic imaging device and detecting a road area on the disparity map. In addition, the controller is configured to detect a vehicle present in the road area in a shape of a stereoscopic object.

Claims

1. A vehicle detecting method, comprising:

creating, by a controller, a disparity map using images photographed by a stereoscopic imaging device;

detecting, by the controller, a road area on the disparity map;

detecting, by the controller, a vehicle present in the road area in a shape of a stereoscopic object;

converting, by the controller, the disparity map into a top view image;

detecting, by the controller, a vehicle area having a rectangular shape in the top view image;

converting, by the controller, the top view image into a second disparity map; and

detecting, by the controller, a height of the vehicle area the second disparity map.

2. The vehicle detecting method according to claim 1, wherein the detection of the road area includes:

creating, by the controller, a v-disparity map using the disparity map;

detecting, by the controller, a road component having a straight line form in the v-disparity map; and

detecting, by the controller, the road area in the disparity map using the road component.

3. The vehicle detecting method according to claim 2, wherein in the detecting of the road component, a plurality of road components are detected in the v-disparity map.

4. The vehicle detecting method according to claim 1, wherein the detection of the vehicle includes:

creating, by the controller, a virtual three-dimensional structure having a preset shape on the disparity map;

deforming, by the controller, the three-dimensional structure to coincide a bottom surface of the three-dimensional structure with the road area; and

detecting, by the controller, the vehicle present in the deformed three-dimensional structure in the shape of the stereoscopic object.

5. The vehicle detecting method according to claim 1, wherein in the converting of the disparity map into the top view image, an area included in the three-dimensional structure is converted into the top view image.

6. The vehicle detecting method according to claim 1, wherein in the detecting of the vehicle area, obstacles having a height equal to or greater than a preset height among obstacles present in the top view image are detected as the vehicle area.

7. The vehicle detecting method according to claim 1, wherein in the detecting of the vehicle area, the vehicle area is detected in consideration of the number of pixels occupied by the vehicle based on a distance from the imaging device to the vehicle.

8. The vehicle detecting method according to claim 1, further comprising:

creating, by the controller, a virtual stereoscopic object that corresponds to the vehicle in the images photographed by the stereoscopic imaging device;

detecting, by the controller, feature points in an area of the stereoscopic object;

creating, by the controller, a motion vector by comparing the feature points with feature points detected in an image temporally continuous to the image; and

moving, by the controller, a position of the stereoscopic object based on the motion vector.

9. A vehicle detecting system, comprising:

a memory configured to store program instructions; and

a processor configured to execute the program instructions, wherein the program instructions when executed cause the processor to;

create a disparity map using images photographed by a stereoscopic imaging device;

detect a road area on the disparity map;

detect a vehicle present in the road area in a shape of a stereoscopic object;

convert the disparity map into a top view image;

detect a vehicle area having a rectangular shape in the top view image;

convert the top view image into a second disparity map; and

convert a height of the vehicle area using the second disparity map.

10. The vehicle detecting system of claim 9, wherein the program instructions when executed are further configured to:

create a v-disparity map using the disparity map;

detect a road component having a straight line form in the v-disparity map; and

detect the road area in the disparity map using the road component.

11. The vehicle detecting system of claim 10, wherein in the detecting of the road component, a plurality of road components are detected in the v-disparity map.

12. The vehicle detecting system of claim 9, wherein the program instructions when executed are further configured to:

create a virtual three-dimensional structure having a preset shape on the disparity map;

deform the three-dimensional structure to coincide a bottom surface of the three-dimensional structure with the road area; and

detect the vehicle present in the deformed three-dimensional structure in the shape of the stereoscopic object.

13. The vehicle detecting system of claim 12, wherein the program instructions when executed are further configured to:

create a virtual stereoscopic object that corresponds to the vehicle in the images photographed by the stereoscopic imaging device;

detect feature points in an area of the stereoscopic object;

create a motion vector by comparing the feature points with feature points detected in an image temporally continuous to the image; and

move a position of the stereoscopic object based on the motion vector.

14. A non-transitory computer readable medium containing program instructions executed by a controller, the computer readable medium comprising:

program instructions that create a disparity map using images photographed by a stereoscopic imaging device;

program instructions that detect a road area on the disparity map;

program instructions that detect a vehicle present in the road area in a shape of a stereoscopic object;

program instruction that converts the disparity map into a top view image;

program instruction that detects a vehicle area having a rectangular shape in the top view image;

program instructions that convert the top view image into a second disparity map; and

program instructions that convert a height of the vehicle area using the second disparity map.

15. The non-transitory computer readable medium of claim 14, further comprising:

program instructions that create a v-disparity map using the disparity map;

program instructions that detect a road component having a straight line form in the v-disparity map; and

program instructions that detect the road area in the disparity map using the road component.

16. The non-transitory computer readable medium of claim 14, further comprising:

program instructions that create a virtual three-dimensional structure having a preset shape on the disparity map;

program instructions that deform the three-dimensional structure to coincide a bottom surface of the three-dimensional structure with the road area; and

program instructions that detect the vehicle present in the deformed three-dimensional structure in the shape of the stereoscopic object.

17. The non-transitory computer readable medium of claim 14, further comprising:

program instructions that create a virtual stereoscopic object that corresponds to the vehicle in the images photographed by the stereoscopic imaging device;

program instructions that detect feature points in an area of the stereoscopic object;

program instructions that create a motion vector by comparing the feature points with feature points detected in an image temporally continuous to the image; and

program instructions that move a position of the stereoscopic object based on the motion vector.