Lane recognizing image processing system

• US 7,295,683 B2
• Filed: 05/06/2004
• Issued: 11/13/2007
• Est. Priority Date: 12/17/2003
• 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
• Primary Examiner(s): Werner; Brian P.
• Assistant Examiner(s): YEH, EUENG NAN
• Application Number:
  US10/839,247
• Publication Number:
  US 20050135658A1
• Time in Patent Office: 1,286 Days
• Field of Search: None
• US Class Current: 382/104
• CPC Class Codes: G06T 7/73 using feature-based methods
• CPC Class Codes: G06V 10/44 Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
• 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 recognizing image processing system for recognizing lane markings laid on a driving road for motor vehicles includes a lane marking candidate point extraction means (20) for extracting lane marking candidate points (C) from an image (G) taken by an image pickup means (10) mounted on the motor vehicle, and a lane recognizing means (30) for recognizing a lane defined on the road on the basis of set of the lane marking candidate points (C). The lane recognizing means (30) includes a lane marking candidate point position converting means (31) for recognizing as lane marking candidate point positions the positions of the lane marking candidate points (C) transformed onto a straight line extending toward a null point.

Claims

1. A lane recognizing image processing system installed on a motor vehicle, comprising:

image pickup means for picking up scenes appearing in front of said motor vehicle;

lane marking candidate point extraction means for extracting candidate points of lane markings from an image taken by said image pickup means as lane marking candidate points; and

lane recognition means for recognizing, on the basis of a set of said lane marking candidate points, a lane on a road on which said motor vehicle is traveling,

wherein said lane recognition means includes;

lane marking candidate point position converting means for converting said lane marking candidate points to lane marking candidate point positions,

wherein said lane marking candidate point position converting means recognizes as said lane marking candidate point positions the positions on a straight line extending in a direction toward a null point onto and along which said lane marking candidate points are converted, and recognizes a midpoint of a lane marking candidate region including said lane marking candidate points as said lane marking candidate point position;

threshold value setting means for setting a threshold value serving as reference for decision as to validity of said set of the lane marking candidate points; and

lane marking candidate decision means for extracting paired points from said set of the lane marking candidate points along a direction corresponding to traveling direction of said motor vehicle to thereby make decision as to whether or not the paired points whose inter-point distance exceeds said threshold value exist among said extracted paired points.

2. A lane recognizing image processing system according to claim 1,

wherein said lane marking candidate decision means is so designed as to decide whether or not the inter-point distance between two points located most distantly from each other as viewed in the traveling direction of said motor vehicle and belonging to said set of the lane marking candidate points exceeds said threshold value.

3. A lane recognizing image processing system according to claim 1,

wherein said threshold value is so set as to be equal to or greater than a lane marking width of said lane marking laid on the road of said motor vehicle and equal to or smaller than an inter-set distance of said lane markings in the traveling direction of said motor vehicle.

4. A lane recognizing image processing system according to claim 1,

wherein said lane recognition means further includes:

invalidating means for invalidating information derived from said lane marking candidate points in the case where said inter-point distance which exceeds said threshold value does not exist.

5. A lane recognizing image processing system according to claim 4,

wherein said lane recognition means includes:

lane recognition result storing means for storing a result of lane recognition processing,

wherein when information is invalidated by said invalidating means, information stored in said lane recognition result storing means is validated.

6. A lane recognizing image processing system according to claim 5,

wherein said lane recognition means is designed such that when invalidation processing executed by said invalidating means continues over or beyond a predetermined time, a validation processing of the stored information held in said lane recognition result storing means is continued for a predetermined period corresponding to a time taken for said motor vehicle to cover said inter-set distance of the lane markings while inhibiting the validation processing of said stored information after lapse of said predetermined period.

7. A lane recognizing image processing system according to claim 4,

wherein said lane recognition means includes:

time-based average processing means for performing a time-based average processing on the result of the lane recognition; and

time-based average result storing means for storing result of time-based average processing delivered from said time-based average processing means,

wherein said information stored in said time-based average result storing means is validated upon invalidation of information by said invalidating means.

8. A lane recognizing image processing system according to claim 7,

wherein said lane recognition means is designed such that when the invalidation processing executed by said invalidating means continues over or beyond a predetermined time, a validation processing of the stored information held in said time-based average result storing means is continued for a predetermined period corresponding to a time taken for said motor vehicle to cover said inter-set distance of the lane markings while inhibiting said validation processing of said stored information after lapse of said predetermined period.