Method for identifying an object in a surrounding region of a motor vehicle, driver assistance system and motor vehicle

Patent metadata

Field	Value
Publication number	US-10635896-B2
Application number	US-201616062978-A
Country	US
Kind code	B2
Filing date	Dec 16, 2016
Priority date	Dec 16, 2015
Publication date	Apr 28, 2020
Grant date	Apr 28, 2020

Abstract

The invention relates to a method for identifying an object (3) in a surrounding region (4) of a motor vehicle (1) as a pedestrian (5), in which the surrounding region (4) is captured in images (B) by means of at least one vehicle-side camera (6), wherein a shadow (11) that is projected from the object (3) onto a ground (12) in the surrounding region (4) and a contrast pattern (13) that is formed on the ground (12) and depends on the shadow (11) are detected in at least one of the captured images (B) and the object (3) is identified on the basis of a comparison of the contrast pattern (13) that was formed with a predetermined contrast pattern that characterizes a pedestrian (5). The invention moreover relates to a driver assistance system (2) and a motor vehicle (1).

First claim

The invention claimed is:

1. A method for identifying an object in a surrounding region of a motor vehicle as a pedestrian, the method comprising:
capturing the surrounding region in images by at least one vehicle-side camera;
detecting a shadow that is projected from the object onto a ground in the surrounding region and a contrast pattern that is formed on the ground and depends on the shadow in at least one of the captured images; and
identifying the object on the basis of a comparison of the contrast pattern that was formed with a predetermined contrast pattern that characterizes a pedestrian,
wherein the object is identified as a pedestrian when two dark stripes that extend substantially parallel to one another and a brighter stripe lying therebetween are detected as the predetermined contrast pattern in the contrast pattern that depends on the shadow of the object.

2. The method according to claim 1 , wherein the object is identified as a pedestrian if an orientation of at least one of the stripes is captured as deviating by no more than a predetermined angle from a detection axis of the camera.

3. The method according to claim 1 , wherein a respective length and/or a respective width of the stripes is captured and the object is identified as a pedestrian if the respective length and/or other respective width deviates by no more than a specified value from a length and/or width that is predetermined for the specified contrast pattern.

4. The method according to claim 1 , wherein a current direction of light incidence of sunlight is determined in relation to the motor vehicle and the predetermined contrast pattern is determined depending on the determined direction of light incidence.

5. The method according to claim 1 , wherein a position of the object identified as a pedestrian is determined in a surrounding map representing the surrounding area on the basis of the at least one image and the object is identified as a pedestrian on the basis of the shadow of the object and of the contrast pattern that is formed in at least one further image for the purposes of tracking the pedestrian and a further position of the object is determined in the surrounding map on the basis of the at least one further image.

6. The method according to claim 1 , wherein the object itself is additionally detected in at least one of the images and the object is only identified and confirmed as a pedestrian when the object was identified as a pedestrian on the basis of the object that was detected in at least one of the images and on the basis of the contrast pattern that was identified in at least one of the images.

7. The method according to claim 6 , wherein the object itself and the contrast pattern are detected in the same image that was captured by the at least one camera and the object is identified as a pedestrian on the basis of the same image.

8. The method according to claim 6 , wherein a process of identifying the object as a pedestrian on the basis of the identified object and a process of identifying the object as a pedestrian on the basis of the detected contrast pattern are carried out simultaneously, at least some of the time.

9. The method according to claim 6 , wherein a position of the detected object and a position of the detected contrast pattern are determined in a surrounding map and the detected object and the contrast pattern are plotted in the surrounding map.

10. The method according to claim 9 , wherein a projection of the detected object on the ground is determined in the surrounding map and the object is identified as a pedestrian if the projection of the object and the contrast pattern of the object that was plotted in the surrounding map overlap at least in a predetermined overlap region.

11. A driver assistance system for identifying an object in a surrounding region of a motor vehicle as a pedestrian, the driver assistance system comprising:
at least one vehicle-side camera for capturing the surrounding region in images; and
an image processing device configured to:
detect a shadow that is projected from the object onto a ground in the surrounding region,
detect a contrast pattern that is formed on the ground and depends on the shadow in at least one of the captured images, and

identify the object on the basis of a comparison of the contrast pattern that was formed with a predetermined contrast pattern that characterizes a pedestrian,
wherein the object is identified as a pedestrian when two dark stripes that extend substantially parallel to one another and a brighter stripe lying therebetween are detected as the predetermined contrast pattern in the contrast pattern that depends on the shadow of the object.

12. A motor vehicle having a driver assistance system according to claim 11 .

Assignees

• Valeo Schalter & Sensoren Gmbh

Inventors

• Heimberger Markus
• Rammos Perikles

Classifications

• G06T7/246

  using feature-based methods, e.g. the tracking of corners or segments · CPC title

• G06T7/73

  using feature-based methods · CPC title

• G06T2207/30252

  Vehicle exterior; Vicinity of vehicle · CPC title

• G06T7/60

  Analysis of geometric attributes · CPC title

• G06T2207/30196

  Human being; Person · CPC title