Intersection recognizing apparatus and computer-readable storage medium

Patent metadata

Field	Value
Publication number	US-9715632-B2
Application number	US-201414202406-A
Country	US
Kind code	B2
Filing date	Mar 10, 2014
Priority date	Mar 15, 2013
Publication date	Jul 25, 2017
Grant date	Jul 25, 2017

Abstract

An intersection recognizing apparatus includes a stereo image obtaining unit configured to obtain a stereo image by capturing a forward image in a street by a stereo camera; a parallax image generator configured to generate a parallax image based on the stereo image obtained by the stereo image obtaining unit; a parallax map generator configured to generate a parallax map based on the parallax image; a feature data storage unit configured to store feature data of an intersection road shoulder width regarding a road surface; and a recognition processing calculation unit configured to recognize an intersection condition based on the parallax map and the feature data of the intersection road shoulder width.

First claim

What is claimed is:

1. An intersection recognizing apparatus, comprising:
a stereo image obtaining unit configured to obtain a stereo image by capturing a forward image in a street by a stereo camera; and
a processor including:
a parallax image generator configured to generate a parallax image based on the stereo image obtained by the stereo image obtaining unit;
a parallax map generator configured to generate a parallax map based on the parallax image;
a feature data storage unit configured to store feature data of an intersection road shoulder width regarding a road surface; and
a recognition processing calculation unit configured to recognize an intersection condition based on the parallax map and the feature data of the intersection road shoulder width to generate a real map,
wherein the parallax map includes a first map that shows a frequency of parallax data in a horizontal direction of the parallax image, a second map that shows a maximum height of the parallax data in the horizontal direction of the parallax image, and a third map that shows a minimum height of the parallax data in the horizontal direction of the parallax image.

2. The intersection recognizing apparatus according to claim 1 , wherein
the stereo camera includes two stereo cameras that are attached in parallel at right and left sides to obtain a right stereo image and a left stereo image, respectively, and
the stereo image obtaining unit is configured to obtain the right stereo image and the left stereo image.

3. The intersection recognizing apparatus according to claim 1 , wherein the parallax image generator is configured to detect sectional images corresponding to each other between right and left images constituting the stereo image obtained by the stereo image obtaining unit and generate the parallax image formed of a parallax between the detected sectional images.

4. The intersection recognizing apparatus according to claim 1 , wherein the recognition processing calculation unit is configured to:
recognize a road shoulder on the third map based on lines extracted on the first map,
eliminate an object recognized on the second map from the road shoulder recognized on the third map,
extract end points on the real map, and
recognize an intersection by comparing a distance between the extracted end points and the feature data of the intersection road shoulder width.

5. The intersection recognizing apparatus according to claim 1 , wherein the recognition processing calculation unit is configured to compare 14 difference in height on the second map with the feature data of difference in road surface height stored in the feature data storage unit and configured to recognize an object on the second map.

6. The intersection recognizing apparatus according to claim 5 , wherein a difference between the second map and the third map is calculated as a difference in height with respect to every road shoulder recognized.

7. The intersection recognizing apparatus according to claim 5 , wherein a triangular relation is present with distance and an actual difference in height,
wherein the actual difference in height is directly proportional to the difference in height on the second map and the difference in height on the third map.

8. The intersection recognizing apparatus according to claim 7 , wherein the recognition processing calculation unit is further configured to:
calculate a distance from parallax data,
utilize difference in height from the distance information, and
calculate actual difference in height.

9. The intersection recognizing apparatus according to claim 8 , wherein when the actual difference in height is larger than the feature data of difference in road surface height, the difference in height is recognized as an object.

10. The intersection recognizing apparatus according to claim 8 , wherein when the actual difference in height is smaller than the feature data of difference in road surface height, the difference in height is recognized as a road shoulder.

11. The intersection recognizing apparatus according to claim 1 , further comprising a recognition result image display unit configured to display the intersection condition.

12. A non-transitory computer-readable storage medium with an executable program stored thereon and executed by a computer, wherein the program instructs the computer to perform:
generating a parallax image based on a stereo image obtained by capturing a forward image in a street by a stereo image obtaining unit;
generating a parallax map based on the parallax image; and
recognizing an intersection condition based on the parallax map and feature data of an intersection road shoulder width regarding a road surface to generate a real map, the feature data being stored in a storage unit,
wherein the parallax map includes a first map that shows a frequency of parallax data in a horizontal direction of the parallax image, a second map that shows a maximum height of the parallax data in the horizontal direction of the parallax image, and a third map that shows a minimum height of the parallax data in the horizontal direction of the parallax image.

13. An intersection recognizing apparatus comprising:
a stereo image obtaining unit configured to obtain a stereo image by capturing a forward image in a street by a stereo camera; and
a processor including:
a parallax image generator configured to generate a parallax image based on the stereo image obtained by the stereo image obtaining unit;
a parallax map generator configured to generate a parallax map based on the parallax image;
a feature data storage unit configured to store feature data of an intersection road shoulder width regarding a road surface; and
a recognition processing calculation unit configured to recognize an intersection condition based on the parallax map and the feature data of the intersection road shoulder width,
wherein the recognition processing calculation unit is configured to eliminate, from the parallax map, an object that has been recognized using a difference in height between a map that shows a minimum height of the parallax data in the parallax image and a map that shows a maximum height of the parallax data in the parallax image, configured to extract a plurality of end points from the parallax map from which the recognized object has been eliminated, and configured to compare a distance between the end points with the feature data to recognize an intersection, and
wherein the parallax map includes a first map that shows a frequency of parallax data in a horizontal direction of the parallax image, a second map that shows a maximum height of the parallax data in the horizontal direction of the parallax image, and a third map that shows a minimum height of the parallax data in the horizontal direction of the parallax image.

Assignees

• Zhong Wei
• Ricoh Co Ltd

Inventors

• Zhong Wei

Classifications

• G06K9/00798 Primary

  Physics · mapped topic

• H04N13/0203

  Electricity · mapped topic

• G06V20/588 Primary

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

• H04N13/204

  using stereoscopic image cameras (stereoscopic photography G03B35/00) · CPC title