Rear obstruction detection

• US 10,827,151 B2
• Filed: 07/29/2019
• Issued: 11/03/2020
• Est. Priority Date: 04/30/2007
• Status: Active Grant

• Current Assignee: MobilEye Vision Technologies Ltd. (Intel Corporation)
• Original Assignee: MobilEye Vision Technologies Ltd. (Intel Corporation)
• Inventors: Stein, Gideon
• Primary Examiner(s): Kim, Hee-Yong
• Application Number:
  US16/524,875
• Publication Number:
  US 20190356886A1
• Time in Patent Office: 463 Days
• Field of Search: 348148
• CPC Class Codes: B60R 2300/30 characterised by the type of image processing
• CPC Class Codes: B60R 2300/404 triggering from stand-by mode to operation mode
• CPC Class Codes: B60R 2300/806 for aiding parking
• CPC Class Codes: B60R 2300/8093 for obstacle warning
• 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
• CPC Class Codes: H04N 7/183 for receiving images from a single remote source

Abstract

A method is provided using a system mounted in a vehicle. The system includes a rear-viewing camera and a processor attached to the rear-viewing camera. When the driver shifts the vehicle into reverse gear, and while the vehicle is still stationary, image frames from the immediate vicinity behind the vehicle are captured. The immediate vicinity behind the vehicle is in a field of view of the rear-viewing camera. The image frames are processed and thereby the object is detected which if present in the immediate vicinity behind the vehicle would obstruct the motion of the vehicle. The processing is preferably performed in parallel for a plurality of classes of obstructing objects using a single image frame of the image frames.

Claims

1. An obstacle detection method for a system mountable in a vehicle, the system including a camera and a processor, the method comprising:

capturing an image frame from a vicinity of the vehicle;

detecting imaged object edges of an object in the image frame;

illuminating the object positioned in a moving direction of the vehicle using a vehicle light;

detecting one or more shadow edges of a shadow of the object responsive to the illumination from the vehicle light and one or more object edges of the object;

generating a subset of pairs of the one or more imaged shadow edges and the one or more object edges, wherein each pair in the subset of pairs comprises a shadow edge and an object edge where;

at least one of the shadow edge and the object edge or a first extension of the shadow edge and a second extension of the object edge meet at a point;

an angle between the shadow edge and the object edge is less than 90 degrees; and

for each pair in the subset;

determining a location of the meeting point in three-dimensional space;

projecting an expected shadow edge based on the object edge; and

determining a distance between at least one point on the expected shadow edge and at least one point on the shadow edge; and

determining the object is at a three-dimensional location based on a comparison of the distance to a predetermined threshold.

2. The obstacle detection method of claim 1, wherein the point is at a terminus of at least one of the shadow or the object edge.

3. The obstacle detection method of claim 1, wherein the object edge in each of the subset of pairs is a vertical line.

4. The obstacle detection method of claim 3, wherein the meeting point is at a bottom terminus of the object edge.

5. The obstacle detection method of claim 4, wherein the bottom terminus of the object edge is at a ground plane.

6. The obstacle detection method of claim 1, wherein at least one of the object edge or the shadow edge is a curved line.

7. The obstacle detection method of claim 1, wherein determining the location of the meeting point is based on one or more image space coordinates of the meeting point responsive to a geometry of the camera and the vehicle light providing the illumination and a focal length of the camera.

8. The obstacle detection method according to claim 1, wherein the vehicle light is illuminated in response to the vehicle shifting into a reverse gear.

9. A system mountable in a vehicle for obstacle detection, the system including a camera and a processor attached to the camera, the processor configured to:

capture an image frame from a vicinity of the vehicle;

detect imaged object edges of an object in the image frame;

illuminate the object positioned in a moving direction of the vehicle using a vehicle light;

detect one or more shadow edges of a shadow of the object responsive to the illumination from the vehicle light and one or more object edges of the object;

generate a subset of pairs of the one or more imaged shadow edges and the one or more object edges, wherein each pair in the subset of pairs comprises a shadow edge and an object edge where;

at least one of the shadow edge and the object edge or a first extension of the shadow edge and a second extension of the object edge meet at a point;

an angle between the shadow edge and the object edge is less than 90 degrees; and

for each pair in the subset;

determine a location of the meeting point in three-dimensional space;

project an expected shadow edge based on the object edge; and

determine a distance between at least one point on the expected shadow edge and at least one point on the shadow edge; and

determine the object is at a three-dimensional location based on a comparison of the distance to a predetermined threshold.

10. The system according to claim 9, wherein the point is at a terminus of at least one of the shadow edge or the object edge.

11. The system according to claim 9, wherein the location of the meeting point is determined based on one or more image space coordinates of the meeting point responsive to a geometry of the camera and the vehicle light providing the illumination and a focal length of the camera.