Obstacle monitoring device using one-dimensional signal

• US 20030099400A1
• Filed: 11/25/2002
• Published: 05/29/2003
• Est. Priority Date: 11/26/2001
• Status: Active Grant

• Current Assignee: DENSO Corporation
• Original Assignee: DENSO Corporation
• Inventors: Ishikawa, Takahiro
• Granted Patent: US 7,170,548 B2
• Time in Patent Office: Days
• US Class Current: 382/190
• CPC Class Codes: G06V 10/255 Detecting or recognising potential candidate objects based on visual cues, e.g. shapes

Abstract

Luminous values of pixels of coordinates previously set as an obstacle presence line are inputted to a memory as a one-dimensional signal converted from a two-dimensional image signal taken by a camera. A portion of the line has a luminous value other than a luminous value predicted as a luminous value of a normal road. The portion of the line is then extracted as a vehicle candidate range. Further, based on characteristic of the vehicle such as a vehicle width or a relative speed, presence or absence of the vehicle is determined by removing noise from the extracted vehicle candidate range.

Claims

1. An obstacle monitoring device comprising:

imaging means for taking an image of a predetermined area;

one-dimension extracting means for extracting a one-dimensional signal from the image; and

obstacle detecting means for determining presence or absence of an obstacle based on the one-dimensional signal.

2. An obstacle monitoring device according to claim 1, wherein the one-dimension extracting means extracts the one-dimensional signal by inputting to a memory a signal at a predetermined portion, which is to be extracted as the one-dimensional signal, by using a synchronizing signal in an image signal of the image taken by the imaging means.

3. An obstacle monitoring device according to claim 1, wherein the one-dimension extracting means includes:

one-dimension setting means for setting a method of converting the image taken by the imaging means into the one-dimensional signal; and

one-dimension generating means for converting the image into the one-dimensional signal in accordance with the converting method set by the one-dimension setting means.

4. An obstacle monitoring device according to claim 3, further comprising:

acquiring means for acquiring data related to an object of taking the image,

wherein the one-dimension setting means changes the converting method based on data acquired by the acquiring means.

5. An obstacle monitoring device according to claim 3, wherein the one-dimension setting means includes:

line setting means for setting a suspected line on which an obstacle is suspected to be present; and

process setting means for setting, based on a pixel on the suspected line, the converting method to be adopted by the one-dimension generating means.

6. An obstacle monitoring device according to claim 5,

wherein the line setting means sets the suspected line based on a parameter of the imaging means and data indicating a characteristic of the object of taking the image by the imaging means.

7. An obstacle monitoring device according to claim 5, further comprising:

data detecting means for detecting a speed and a steering angle of an own vehicle where the obstacle monitoring device is provided,

wherein the line setting means adjusts the suspected line in accordance with the speed and the steering angle of the own vehicle detected by the data detecting means.

8. An obstacle monitoring device according to claim 5, further comprising:

navigation means for calculating a current position and map information around the current position,

wherein the line setting means sets the suspected line based on the current position and the map information around the current position calculated by the navigation means.

9. An obstacle monitoring device according to claim 5, further comprising:

lane recognizing means for recognizing a lane in the image taken by the imaging means,

wherein the line setting means sets the suspected line based on lane information detected by the lane recognizing means.

10. An obstacle monitoring device according to claim 5,

wherein the process setting means sets the converting method including projecting a pixel on the suspected line to a predetermined axis.

11. An obstacle monitoring device according to claim 5,

wherein the process setting means sets the converting method including aligning a pixel on the suspected line based on distance information from a predetermined position.

12. An obstacle monitoring device according to claim 1,

wherein the obstacle detecting means determines presence or absence of the obstacle by comparing the corresponding one-dimensional signal with a reference value of a one-dimensional signal which is obtained when no image of an obstacle is determined to be taken by the imaging means.

13. An obstacle monitoring device according to claim 12,

wherein the reference value is determined based on a frequency of the value of the one-dimensional signal.

14. An obstacle monitoring device according to claim 1,

wherein the obstacle detecting means determines presence or absence of the obstacle for a respective portion corresponding to the one-dimensional signal.

15. An obstacle monitoring device according to claim 14,

wherein the obstacle detecting means estimates, based on the presence or absence of the obstacle at the respective portion, an obstacle candidate range in the one-dimensional signal, the obstacle candidate range in which the obstacle is estimated to be present.

16. An obstacle monitoring device according to claim 15,

wherein the obstacle detecting means estimates a relative speed at a respective portion of the one-dimensional signal based on a correlation of two of the one-dimensional signals at different time points extracted by the one-dimension extracting means, and

wherein the obstacle detecting means determines whether the obstacle candidate range corresponds to the obstacle based on the estimated relative speed of the respective portion.

17. An obstacle monitoring device according to claim 1,

wherein the obstacle detecting means estimates a relative speed at a respective portion of the one-dimensional signal based on a correlation of two of the one-dimensional signals at different time points extracted by the one-dimension extracting means.

18. An obstacle monitoring device according to claim 1,

wherein the obstacle detecting means includes:

speed modeling means for estimating a relative speed at a respective point of the one-dimensional signal based on a correlation of two of the one-dimensional signals at different time points extracted by the one-dimension extracting means; and

obstacle determining means for determining the obstacle by using the estimated relative speed.

19. An obstacle monitoring device according to claim 18,

wherein the obstacle determining means is composed of:

obstacle presence determining means which groups points of the estimated relative speed in a direction of approaching the imaging means to determine a range as an obstacle when the range formed by the grouped points is equal or greater than a predetermined size.

20. An obstacle monitoring device according to claim 19,

wherein the obstacle determining means is further composed of:

obstacle relative position calculating means for calculating a relative position of the obstacle based on a range end point of the range determined as the obstacle by the obstacle presence determining means.

21. An obstacle monitoring device according to claim 19,

wherein the obstacle determining means is further composed of:

obstacle relative speed calculating means for calculating a relative speed of the obstacle based on a relative speed of the range determined as the obstacle by the obstacle presence determining means.

22. An obstacle monitoring device according to claim 19,

wherein the obstacle determining means is further composed of:

obstacle relative speed calculating means for calculating a relative speed of the obstacle based on a relative speed of the range determined as the obstacle by the obstacle presence determining means and peripheral ranges of the range determined as the obstacle by the obstacle presence determining means.

23. An obstacle monitoring device according to claim 1,

wherein the obstacle detecting means includes:

range modeling means for estimating an obstacle candidate range in which the obstacle is estimated to be present based on a change of a signal value of the one-dimensional signal extracted by the one-dimension extracting means;

speed modeling means for estimating a relative speed at a respective point of the one-dimensional signal based on a correlation of two of the one-dimensional signals at different time points extracted by the one-dimension extracting means; and

obstacle determining means for determining the obstacle by using the obstacle candidate range estimated by the range modeling means and the estimated relative speed by the speed modeling means.

24. An obstacle monitoring device according to claim 23,

wherein the speed modeling means is composed of:

speed converting means for converting two of the one-dimensional signals at the different time points extracted by the one-dimension extracting means into the one-dimensional signals emphasizing a difference between the obstacle candidate range and other range; and

speed model estimating means for estimating the relative speed at the respective point of the one-dimensional signal based on a correlation of two of the one-dimensional signals at different time points converted by the speed converting means.

25. An obstacle monitoring device according to claim 24,

wherein the speed converting means converts the one-dimensional signal extracted by the one-dimension extracting means into a one-dimensional signal expressing only the obstacle candidate range by masking a background range.

26. An obstacle monitoring device according to claim 23,

wherein the range modeling means is composed of:

range converting means for converting the one-dimensional signal extracted by the one-dimension extracting means into a one-dimensional signal emphasizing a difference between the obstacle candidate range and the other range based on the change of the signal value; and

range model estimating means for estimating the obstacle candidate range based on the change of the signal value of the one-dimensional signal converted by the range converting means.

27. An obstacle monitoring device according to claim 26,

wherein the range converting means converts the one-dimensional signal extracted by the one-dimension extracting means into a one-dimensional signal expressing only the obstacle candidate range by masking a background range.

28. An obstacle monitoring device according to claim 23,

wherein the obstacle determining means is composed of:

obstacle presence determining means,

wherein the speed modeling means calculates a relative speed in the obstacle candidate range estimated by the range modeling means and a peripheral range of the obstacle candidate range, and

wherein the obstacle presence determining means determines whether the obstacle candidate range is the obstacle by comparing a number of points of the relative speed in a direction of approaching the imaging means and a number of points of the relative speed in a direction of separating from the imaging means.

29. An obstacle monitoring device according to claim 1, further comprising:

frame synthesizing means for promoting a detection accuracy of the obstacle by synthesizing results of the obstacle detecting means based on a plurality of frames taken by the imaging means.

30. An obstacle monitoring device according to claim 1, further comprising:

line synthesizing means,

wherein the one-dimension extracting means extracts a plurality of one-dimension signals from the image taken by the imaging means,

wherein the obstacle detection means determines presence of the obstacle with respect to each of the one-dimensional signals extracted by the one-dimension extracting means, and

wherein the line synthesizing means synthesizes results of the obstacle detection means with respect to each of the one-dimensional signals to promote a detection accuracy of the obstacle.

31. An obstacle monitoring device according to claim 1, further comprising:

notifying means for issuing an alarm based on a detection result of the obstacle detecting means.

32. An obstacle monitoring device according to claim 1,

wherein the imaging means is installed at an own vehicle where the obstacle monitoring device is provided and the obstacle is another vehicle.

33. A method of monitoring an obstacle comprising steps of:

taking an image of a predetermined area;

extracting a one-dimensional signal from the image; and

determining presence or absence of an obstacle based on the one-dimensional signal.