Method and apparatus for warning an obstacle of a vehicle

Patent metadata

Field	Value
Publication number	US-9704062-B2
Application number	US-201414562217-A
Country	US
Kind code	B2
Filing date	Dec 5, 2014
Priority date	Aug 19, 2014
Publication date	Jul 11, 2017
Grant date	Jul 11, 2017

Abstract

An apparatus for warning an obstacle of a vehicle includes an image obtaining device configured to take an image around the vehicle through image sensors. An image segmentation device is configured to segment a current reference image obtained in blocks of an object. A motion estimator is configured to estimate two dimensional motion information of the object. A variation estimator is configured to estimate distance information between the object and the vehicle. A relative motion estimator is configured to estimate three dimensional relative motion information of the object by combining the two dimensional motion information and the distance information. A collision time estimator is configured to estimate a collision time between the object and the vehicle by using the three dimensional relative motion information. An image processor is configured to adjust a brightness of a corresponding segmented block. A display is configured to display the image.

First claim

What is claimed is:

1. An apparatus for warning an obstacle of a vehicle comprising:
an image obtaining device configured to take an image around the vehicle through at least two image sensors;
an image segmentation device configured to segment a current reference image obtained through one image sensor among the at least two image sensors in blocks of an object;
a motion estimator configured to estimate two dimensional motion information of the object in each of segmented blocks segmented by the image segmentation device;
a variation estimator configured to estimate distance information between the object in a segmented block and the vehicle;
a relative motion estimator configured to estimate three dimensional relative motion information of the object in the segmented block by combining the two dimensional motion information and the distance information;
a collision time estimator configured to estimate a collision time between the object in the segmented block and the vehicle by using the three dimensional relative motion information;
an image processor configured to adjust a brightness of a corresponding segmented block in a pre-defined ratio according to the collision time; and
a display configured to display an image processed by the image processor,
wherein the image processor adds a shadow effect to an outline of the corresponding segmented block if the collision time is less than a threshold time.

2. The apparatus according to claim 1 , wherein the at least two image sensors are mounted on the vehicle in a horizontal arrangement.

3. The apparatus according to claim 1 , wherein the object is an obstacle.

4. The apparatus according to claim 1 , wherein the motion estimator estimates the two dimensional motion information by comparing the current reference image and a previous reference image.

5. The apparatus according to claim 1 , wherein the variation estimator estimates the distance information by estimating a variation using a correlation between the current reference image and a comparison image inputted through another image sensor.

6. The apparatus according to claim 1 , wherein the collision time estimator estimates the collision time between the object in the segmented block and the vehicle by using a time difference between the current reference image and the previous reference image and the three dimensional relative motion information.

7. The apparatus according to claim 1 , wherein the apparatus further comprises a warning output configured to output a warning signal if the collision time is less than the threshold time.

8. A method for warning an obstacle of a vehicle comprising steps of:
taking an image around the vehicle through at least two image sensors;
segmenting a current reference image obtained through one image sensor among the at least two image sensors in blocks of an object;
estimating two dimensional motion information of the object in each of segmented blocks and distance information;
estimating three dimensional relative motion information of the object in a segmented block by combining the two dimensional motion information and the distance information;
estimating a collision time between the object in the segmented block and the vehicle by using the three dimensional relative motion information;
adding a shadow effect to an outline of a corresponding segmented block if the collision time is less than a threshold time; and
displaying the image by adjusting a brightness of the corresponding segmented block in a pre-defined ratio according to the collision time.

9. The method according to claim 8 , wherein the two dimensional motion information is estimated by comparing the current reference image and a previous reference image.

10. The method according to claim 8 , wherein the distance information is estimated by estimating a variation using a correlation between the current reference image and a comparison image inputted through another image sensor.

11. The method according to claim 8 , wherein the step of estimating the collision time estimates the collision time between the object in the segmented block and the vehicle by using a time difference between the current reference image and a previous reference image and the three dimensional relative motion information.

12. The method according to claim 8 , further comprising: a step of outputting a warning signal if the collision time is less than the threshold time.

Assignees

• Hyundai Motor Co Ltd

Inventors

• Won Ji Eun
• Jung Gyo Young
• Chang Eu Gene
• Yang Uk Il

Classifications

• G06V20/64

  Three-dimensional [3D] objects · CPC title

• G06V20/58

  Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads · CPC title

• G06V10/50

  by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis · CPC title

• B60R2300/105

  using multiple cameras · CPC title

• G06T7/223

  using block-matching · CPC title