METHOD AND APPARATUS FOR TUNNEL DECISION

• US 20160162741A1
• Filed: 07/16/2015
• Published: 06/09/2016
• Est. Priority Date: 12/05/2014
• Status: Active Grant

• Current Assignee: Hyundai Mobis Co., Ltd.
• Original Assignee: Hyundai Mobis Co., Ltd.
• Inventors: Shin, Chang Mok
• Granted Patent: US 9,785,843 B2
• Time in Patent Office: Days
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle

Abstract

The present invention provides a tunnel decision apparatus, including: a camera which outputs an image in front of a vehicle; a road area classifying unit which detects a vanishing point from the image in front of a vehicle to output at least one of a road area and a non-road area; a pattern detecting unit which detects at least one of a lamp pattern from the non-road area and a lane pattern from the road area; and a tunnel decision unit which determines whether the vehicle enters a tunnel in consideration of at least one of the lamp pattern and the lane pattern.

Claims

1. A tunnel decision apparatus, comprising:

a camera which outputs an image in front of a vehicle;

a road area classifying unit which detects a vanishing point from the image in front of a vehicle to output at least one of a road area and a non-road area;

a pattern detecting unit which detects at least one of a lamp pattern from the non-road area and a lane pattern from the road area; and

a tunnel decision unit which determines whether the vehicle enters a tunnel in consideration of at least one of the lamp pattern and the lane pattern.

2. The tunnel decision apparatus of claim 1, wherein the road area classifying unit outputs a lower portion of a horizontal line including the vanishing point in the image in front of a vehicle as the road area.

3. The tunnel decision apparatus of claim 1, wherein the road area classifying unit outputs a upper portion of a horizontal line including the vanishing point in the image in front of a vehicle as the non-road area.

4. The tunnel decision apparatus of claim 1, wherein the road area classifying unit detects a horizontal line from a coordinate system set in the camera to output the road area and the non-road area.

5. The tunnel decision apparatus of claim 1, wherein the pattern detecting unit detects the lamp pattern in consideration of a brightness of a pixel in the non-road area.

6. The tunnel decision apparatus of claim 1, wherein the tunnel decision unit determines whether the vehicle enters the tunnel in consideration of the lamp pattern and the lane pattern and at least one of a curvature, a heading direction, and a gradient of the lane pattern in the image in front of a vehicle.

7. The tunnel decision apparatus of claim 1, wherein the tunnel decision unit determines whether the vehicle escapes the tunnel in consideration of at least one of a duration time of detection and a detection frequency of the lamp pattern.

8. A tunnel decision method, comprising:

outputting an image in front of a vehicle;

detecting a vanishing point from the image in front of a vehicle to output at least one of a road area and a non-road area;

detecting at least one of a lamp pattern from the non-road area and a lane pattern from the road area; and

determining whether the vehicle enters a tunnel in consideration of at least one of the lamp pattern and the lane pattern.

9. The tunnel decision method of claim 8, wherein in the outputting of the road area, a lower portion of a horizontal line including the vanishing point in the image in front of a vehicle is output as the road area.

10. The tunnel decision method of claim 8, wherein in the outputting of the road area, an upper portion of a horizontal line including the vanishing point in the image in front of a vehicle is output as the non-road area.

11. The tunnel decision method of claim 8, wherein in the detecting of a pattern, the lamp pattern is detected in the non-road area in consideration of a brightness of a pixel.

12. The tunnel decision method of claim 8, wherein in the determining of whether the vehicle enters or escapes the tunnel, it is determined whether the vehicle enters the tunnel in consideration of the lamp pattern and the lane pattern and at least one of a curvature, a heading direction, and a gradient of the lane pattern in the image in front of a vehicle.

13. The tunnel decision method of claim 8, wherein in the determining of whether the vehicle enters or escapes the tunnel, it is determined whether the vehicle escapes the tunnel in consideration of at least one of a duration time of detection and a detection frequency of the lamp pattern.

14. A tunnel decision method, comprising:

outputting an image in front of a vehicle;

detecting a vanishing point from the image in front of a vehicle to output at least one of a road area and a non-road area; and

detecting a lamp pattern from the non-road area and detecting a lane pattern from the road area.

15. The tunnel decision method of claim 14, wherein in the detecting of a lamp pattern and a lane pattern, the lamp pattern is detected from the non-road area and the lane pattern is detected from the road area, simultaneously, in the image in front of a vehicle.

16. The tunnel decision method of claim 15, further comprising:

simultaneously detecting the lamp pattern and the lane pattern and determining whether the vehicle enters the tunnel in consideration of the lamp pattern and the lane pattern and at least one of a curvature, a heading direction, and a gradient of the lane pattern.

17. The tunnel decision method of claim 16, wherein determining whether the vehicle escapes the tunnel in consideration of a duration time of detection and a detection frequency of the lamp pattern when a vehicle feature similarity of the lane pattern is equal to or larger than a set value.