Lane recognition image processing apparatus

• US 7,209,832 B2
• Filed: 12/06/2004
• Issued: 04/24/2007
• Est. Priority Date: 07/15/2004
• Status: Active Grant

• Current Assignee: Mitsubishi Denki Kabushiki Kaisha (Mitsubishi Electric Corporation)
• Original Assignee: Mitsubishi Denki Kabushiki Kaisha (Mitsubishi Electric Corporation)
• Inventors: Yamamoto, Takayuki, Ishikura, Hisashi, Fujie, Kenichi, Fujii, Yoshiyuki, Sugiyama, Akinobu
• Primary Examiner(s): To; Tuan C
• Application Number:
  US11/003,468
• Publication Number:
  US 20060015252A1
• Time in Patent Office: 869 Days
• Field of Search: 701/301, 701/96, 701/211, 701/300, 701/117, 382/260, 382/103, 382/268, 382/154, 382/263, 382/264, 382/266, 382/291, 382/104, 382/199, 382/168, 382/22, 340/907, 340/919, 340/435, 340/436, 340/937, 340/903, 340/933, 348/148, 348/113, 348/118, 706/22, 706/20
• US Class Current: 701/301
• CPC Class Codes: G01S 11/12 using electromagnetic waves other than radio waves
• CPC Class Codes: G01S 17/931 of land vehicles
• CPC Class Codes: G01S 5/16 using electromagnetic waves other than radio waves
• 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

Abstract

A lane recognition image processing apparatus can improve lane marking recognition performance by preventing false detection by addition of a condition without changing the basic principle of an one-dimensional image filter. A search area is set for each lane marking with respect to images stored in a image storage part through a window. A candidate point extraction part extracts candidate points of each lane marking from the search area thus set. A lane marking mathematical model equation is derived by approximating sets of extracted candidate points by a mathematical model equation. The candidate point extraction part includes a kernel size setting part, a filtering part that outputs, as a filtering result, the smaller of differences between the gray value of a pixel of interest and those of pixels forwardly and rearwardly apart a kernel size from the pixel of interest in a scanning direction, respectively, and a binarization part that binarizes the filtering result.

Claims

1. A lane recognition image processing apparatus installed on a vehicle for recognizing a lane based on a sensed image of at least one lane marking on the surface of a road, said apparatus comprising:

an image sensing part sensing a forward view of said vehicle;

an image storage part temporarily storing images obtained by said image sensing part;

a window setting part setting a search area for said at least one lane marking with respect to the images stored in said image storage part by means of a window;

a candidate point extraction part extracting candidate points for said at least one lane marking from said search area set by said window setting part; and

a lane recognition part deriving a lane marking mathematical model equation by approximating sets of candidate points extracted by said candidate point extraction part by a mathematical model equation,

wherein said candidate point extraction part comprises;

a kernel size setting part setting a kernel size Δ

h in accordance with a forward distance from said vehicle;

a one-dimensional image filtering part outputting, as a filtering result, the smaller one of the values that are obtained by two equations {g(h)−

g(h−

Δ

h)} and {g(h)−

g(h+Δ

h)} using the gray value g(h) of a pixel of interest and the gray values g(h−

Δ

h), g(h+Δ

h) of pixels forwardly and rearwardly apart said kernel size Δ

h from said pixel of interest in a scanning direction, respectively; and

a binarization part that binarizes said filtering result with a threshold.

2. The lane recognition image processing apparatus as set forth in claim 1, wherein said binarization part comprises:

an S/N ratio calculation part calculating an S/N ratio of said filtering result; and

a threshold lower limit setting part setting a lower limit of said threshold based on said S/N ratio;

wherein said S/N ratio calculation part counts the number of filter pass ranges having their range widths less than a specified value in said filtering result as the number of noise ranges, also counts the number of filter pass ranges having their range widths greater than or equal to the specified value in said filtering result as the number of signal ranges, and calculates said S/N ratio based on said number of signal ranges and said number of noise ranges.

3. The lane recognition image processing apparatus as set forth in claim 2, wherein said image sensing part comprises a CMOS image sensor.

4. The lane recognition image processing apparatus as set forth in claim 1, wherein said binarization part includes a multi-threshold setting part setting said threshold for each search scanning line of said one-dimensional image filtering part.

5. The lane recognition image processing apparatus as set forth in claim 1, wherein said window setting part includes a candidate point reference part setting, in the case of said candidate points being two or more, a window based on an intersection between a straight line connecting two points among said candidate points and the following search scanning line.

6. The lane recognition image processing apparatus as set forth in claim 1, wherein

said at least one lane marking comprises a pair of right and left lane markings, and said apparatus further comprises a vanishing point learning part obtaining a vanishing point learning position based on said right and left lane markings;

said vanishing point learning part approximates sets of candidate points extracted by said candidate point extraction part by straight lines, and obtains said vanishing point learning position from an intersection between approximate straight lines of said right and left lane markings; and

said window setting part includes a vanishing point reference part that sets said window on the basis of an intersection between a straight line connecting one of said candidate points and said vanishing point learning position and the following search scanning line.