Object detection device and vehicle using same

• US 9,886,649 B2
• Filed: 09/29/2014
• Issued: 02/06/2018
• Est. Priority Date: 10/07/2013
• Status: Active Grant

• Current Assignee: Hitachi Astemo Ltd. (Hitachi, Ltd.)
• Original Assignee: Hitachi Automotive Systems Limited (Hitachi, Ltd.)
• Inventors: Kakegawa, Shinji, Matono, Haruki, Otsuka, Yuji
• Primary Examiner(s): Black, Thomas G
• Assistant Examiner(s): Smith-Stewart, Demetra R
• Application Number:
  US15/026,713
• Publication Number:
  US 20160253575A1
• Time in Patent Office: 1,226 Days
• Field of Search: None
• CPC Class Codes: B60T 7/22 initiated by contact of vehicle, e.g. bumper, with an external object, e.g. another vehicle , or by means of contactless obstacle detectors mounted on the vehicle
• CPC Class Codes: G06T 2207/30256 Lane; Road marking
• CPC Class Codes: G06T 7/593 from stereo images
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle
• 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: G06V 20/588 Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
• CPC Class Codes: G06V 20/647 by matching two-dimensional images to three-dimensional objects
• CPC Class Codes: G08G 1/165 for passive traffic, e.g. including static obstacles, trees
• CPC Class Codes: G08G 1/166 for active traffic, e.g. moving vehicles, pedestrians, bikes

Abstract

Provided is an object detection device that makes it possible to more accurately detect a road surface even when an area that is not the road surface occupies the majority of an image and the gradient of the road surface is large, and to detect an obstacle on a road surface with greater accuracy than in the prior art. The object detection device includes: a parallax calculation unit that generates a parallax image on the basis of a plurality of images; a parallax gradient calculation unit that, for each pixel in the parallax image, calculates a parallax gradient for each pixel on the basis of the parallax of a predetermined pixel and a pixel that is adjacent to the predetermined pixel; a detection candidate determination unit that determines whether a detection candidate is present on the basis of the parallax gradient of each pixel; a voting unit that carries out a histogram vote using each of the pixels that are determined to be detection candidates and generates a histogram frequency for each of the pixels when the detection candidate is determined to be present; and an object detection unit that detects presence of an object and a position of the object on the basis of the histogram frequency.

Claims

1. An object detection device comprising:

a parallax calculation unit that generates a parallax image based on a plurality of images;

a parallax gradient calculation unit that, for each pixel in the parallax image, calculates a parallax gradient for each pixel based on the parallax of a predetermined pixel and a pixel that is adjacent to the predetermined pixel;

a detection candidate determination unit that determines whether a detection candidate is present based on the parallax gradient of each pixel;

a voting unit that carries out a histogram vote using each of the pixels that are determined to be detection candidates and generates a histogram frequency for each of the pixels when the detection candidate is determined to be present; and

an object detection unit that detects presence of an object and a position of the object based on the histogram frequency;

wherein the detection candidate determination unit compares the parallax gradient of the pixel and a parallax gradient threshold corresponding to the acquired parallax of the pixel for each pixel in the parallax image, and determines whether the detection candidate is present from each of the pixels;

wherein the parallax gradient calculation unit calculates and outputs the parallax gradient α

i=Δ

Disparity/Δ

V, where V is an image ordinate; and

wherein the parallax gradient threshold is determined based on a maximum gradient of the road surface.

2. The object detection device according to claim 1, wherein

the object to be detected by the object detection unit is a road surface or an obstacle.

3. The object detection device according to claim 1, wherein

the detection candidate of the detection candidate determination unit is a road surface candidate, and

the parallax gradient threshold is a threshold to determine whether the road surface is present for each parallax in a certain fixed range, and is stored in a parallax gradient threshold table in advance.

4. The object detection device according to claim 1, wherein the detection candidate determination unit determines whether a data point of each of the pixels, indicating a point on a three-dimensional space which is represented by an image coordinate of the pixel and the parallax stored in the pixel, is a road surface candidate.

5. The object detection device according to claim 1, wherein a V-disparity image is a two-dimensional image having an image coordinate in the vertical direction and parallax disparity as its axes, and the histogram vote counts up the histogram frequency of the pixel with which the spatial position of the data point is projected onto the V-disparity image, one by one, for each data point of target pixels.

6. The object detection device according to claim 1, wherein the parallax gradient calculation unit executes processing at a timing delayed by one line plus one pixel after the parallax calculation unit generates the parallax image.

7. The object detection device according to claim 1, wherein the application to determine whether a data point of the pixel is an object candidate in air including a traffic light or a road sign and detect an object in the air using the object detection unit on a subsequent stage is made by adjusting Δ

Yup and Δ

Ydown, where Y represents a height of the data point having a position of a camera as an origin point.

8. A vehicle comprising:

the object detection device according to claim 1 which is provided with a traveling control unit that generates an instruction to control a braking force or a driving force based on the presence of the object and the position of the object which are detected by the object detection unit.