Vehicle environment recognition system

• US 20090190800A1
• Filed: 01/23/2009
• Published: 07/30/2009
• Est. Priority Date: 01/25/2008
• Status: Active Grant

• Current Assignee: Subaru Corp.
• Original Assignee: Fuji Heavy Industries Limited (Subaru Corp.)
• Inventors: Takahashi, Yasushi
• Granted Patent: US 8,244,027 B2
• Time in Patent Office: Days
• US Class Current: 382/104
• CPC Class Codes: G06T 2207/10012 Stereo images
• CPC Class Codes: G06T 2207/30261 Obstacle
• CPC Class Codes: G06T 7/593 from stereo images
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads

Abstract

A vehicle environment recognition system includes stereo-image taking means for taking images of an environment around a subject vehicle and for outputting the images as a reference image and a comparative image, first stereo matching means for forming a first distance image on the basis of the reference image and the comparative image or on the basis of two images obtained by preprocessing the reference image and the comparative image, second stereo matching means for forming a second distance image on the basis of two images obtained by preprocessing the reference image and the comparative image in a different manner, detection means for detecting objects in the reference image on the basis of the first and second distance images, and selection means for selecting one of the results of detection based on the first and second distance images.

Claims

1. A vehicle environment recognition system comprising:

stereo-image taking means for simultaneously taking images of an environment around a subject vehicle with a pair of cameras and for outputting the images as a reference image and a comparative image;

first stereo matching means for calculating first parallaxes or first distances by stereo matching based on the reference image and the comparative image or based on two images formed by subjecting the reference image and the comparative image to a predetermined first preprocessing operation, and for forming a first distance image by assigning the calculated first parallaxes or first distances to pixel blocks in the reference image;

second matching means for calculating second parallaxes or second distances by stereo matching based on two images formed by subjecting the reference image and the comparative image to a predetermined second preprocessing operation different from the first preprocessing operation, and for forming a second distance image by assigning the calculated second parallaxes or second distances to the pixel blocks in the reference image;

detection means for detecting objects in the reference image on the basis of the first distance image and the second distance image; and

selection means for selecting one of a result of detection of the objects based on the first distance image and a result of detection of the objects based on the second distance image.

2. The vehicle environment recognition system according to claim 1, wherein the selection means performs the selection on the basis of a specific object of the objects detected in the reference image by the detection means.

3. The vehicle environment recognition system according to claim 2, wherein the selection means compares a current lateral width of the specific object detected in a current object detection result with a previous lateral width of the specific object calculated on the basis of the object detection result selected in previous object detection, and rejects the current detection result when the current lateral width is smaller by at least a first predetermined threshold value than the previous lateral width.

4. The vehicle environment recognition system according to claim 2, wherein the selection means compares a current distance between the specific object and the subject vehicle that is detected in a current object detection result with a previous distance between the specific object and the subject vehicle that is calculated on the basis of the object detection result selected in previous object detection, and rejects the current object detection result when the current distance is different by at least a predetermined second threshold value from the previous distance.

5. The vehicle environment recognition system according to claim 2, wherein the selection means rejects both of the object detection result based on the first distance image and the object detection result based on the second distance image when the specific object is detected in one of the object detection results, and is not detected in the other object detection result at a detecting position in the reference image where the object is detected in the one detection result, and when the number of data on the first or second parallaxes or the first or second distances existing at the detecting position of the specific object in the first or second distance image in the one object detection result is less than a predetermined third threshold value.

6. The vehicle environment recognition system according to claim 2,

wherein the first stereo matching means forms the first distance image by performing the stereo matching based on the reference image and the comparative image, and

wherein the second stereo matching means forms the second distance image by performing the stereo matching based on two images that are formed by subjecting the reference image and the comparative image to edge detection as the second preprocessing operation.

7. The vehicle environment recognition system according to claim 6, wherein the selection means rejects the object detection result based on the first distance image when a distance between the specific object and the subject vehicle calculated on the basis of the first distance image is less than a distance between the specific object and the subject vehicle calculated on the basis of the second distance image, when the number of data on the first parallaxes or first distances at a detecting position of the specific object in the first distance image is less than a predetermined fourth threshold value, and when the number of data on the second parallaxes or second distances at a detecting position of the specific object in the second distance image is more than a predetermined fifth threshold value.

8. The vehicle environment recognition system according to claim 3, wherein, when none of the object detection result based on the first distance image and the object detection result based on the second distance image are rejected, the selection means rejects one of the object detection results including a smaller number of data on the first or second parallaxes or the first or second distances at a position where the specific object is detected in the first distance image or the second distance image.