Unmanned vehicle control system with guide line detection

• US 5,087,969 A
• Filed: 07/19/1990
• Issued: 02/11/1992
• Est. Priority Date: 07/20/1989
• Status: Expired due to Term

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Kamada, Hiroshi, Ozaki, Tohru
• Primary Examiner(s): Kostak, Victor R.
• Application Number:
  US07/554,365
• Time in Patent Office: 572 Days
• Field of Search: 358/108, 358/103, 358/100, 180/167-169, 364/424.01, 364/424.02, 364/424.03, 901/1
• US Class Current: 348/119
• CPC Class Codes: G05D 1/0221 involving a learning process
• CPC Class Codes: G05D 1/0234 using optical markers or beacons optical beacons per se G01S1/70
• CPC Class Codes: G05D 1/0246 using a video camera in combination with image processing means
• CPC Class Codes: Y10S 706/90 Fuzzy logic

Abstract

A system for controlling an unmanned vehicle by detection of a guide line based on a picture image picked up by a camera on the vehicle, including a parameter (ρ, θ) deriving unit for deriving parameters (ρ, θ) of point extracted from a picture image picked up by a camera, and a maximum density parameter detection unit for integrating the densities of parameters derived by the parameter deriving unit and determining the maximum density parameter from the integrated densities of the parameters. The maximum density parameter is determined from the parameter derived by the parameter deriving unit, for the parameter area having at the center thereof the parameter estimated from the maximum density parameter, and the detected maximum density parameter is decided as the parameter of said guide line. The determination of steered amount by a fuzzy inference can be used.

Claims

1. A system for controlling an unmanned vehicle by detection of a guide line based on a picture image picked up by a camera on said vehicle, said system comprising:

parameter (ρ

, θ

) deriving means for deriving parameters (ρ

, θ

) of a point extracted from a picture image picked up by a camera; and

maximum density parameter detection means for integrating the densities of parameters derived by said parameter deriving means and determining the maximum density parameter from the integrated densities of the parameters;

the maximum density parameter being determined from the parameter derived by said parameter deriving means for the parameter area having at the center thereof the parameter estimated from the maximum density parameter;

said detected maximum density parameter being decided as the parameter of said guide line.

2. A system according to claim 1, wherein the parameter area is expanded when the number of parameters in the parameter area becomes less than a predetermined number, and the parameter area is reduced when the number of parameters in the parameter area becomes larger than the predetermined number.

3. A system for controlling an unmanned vehicle by detection of a guide line from a picture image picked up by a camera on said vehicle, said system comprising:

parameter (ρ

, θ

) deriving means for deriving parameters (ρ

, θ

) of a point extracted from a picture image picked up by a camera;

maximum density parameter detection means for integrating the densities of parameters derived by said parameter deriving means and deciding the maximum density parameter from the integrated densities of the parameters;

estimated parameter area calculating means for estimating a possible parameter area from past parameters of the maximum density detected by said maximum density parameter detection means; and

attention region table storing means for storing a table for converting an estimated parameter area estimated by said estimated parameter area calculating means into the corresponding attention region on the picture image;

parameters being derived by said parameter deriving means with regard to extracted points in the attention region converted by said attention region table storing means;

the parameter of the maximum density being detected from the derived parameters by said maximum density parameter detection means to be delivered as the parameter of said guide line.

4. A system according to claim 3, wherein the attention region is expanded when the number of the extracted points in the attention region becomes less than a predetermined number, and the attention region is reduced when the number of the extracted points in the entire region or the attention region becomes larger than the predetermined number.

5. A method for controlling an unmanned vehicle by detection of a guide line from a picture image picked up by a camera on said vehicle in which a straight line approximation process is used for calculating an approximate straight line from received point data, said straight line approximation process comprising the steps of:

selecting a combination in which a parallel component of a line segment connecting two points from the received point data with respect to an estimated straight line is longer than a predetermined threshold length and a perpendicular component of the above-indicated line segment with respect to the estimated straight line is shorter than a predetermined threshold length by using a parameter calculation device;

calculating parameters (ρ

, θ

) of the line segment connecting the selected two points by using the parameter calculation device;

integrating the parameters (ρ

, θ

) calculated by the parameter calculation device; and

determining parameter coordinates (ρ

, θ

) having the maximum parameter density as the parameter of the approximate straight line.

6. A system for controlling an unmanned vehicle by detection of a guide line from a picture image picked up by a camera on said vehicle in which a straight line approximation process is used for calculating an approximate straight line from received point data, comprising the steps of:

selecting points each having a distance less than a predetermined threshold length from an estimated straight line by using parameter calculation means; and

calculating parameters (ρ

, θ

) of straight lines passing through one of the selected points or parameters (ρ

, θ

) of a line segment connecting two of the selected points by using the parameter calculation means;

the parameters of the estimated straight line being determined from the result of said calculation.

7. A method according to claim 6, wherein integration of parameters (ρ

, θ

) calculated by said parameter calculation means is repeated for each of grid divided squares and the parameters for a square having the maximum density parameter are determined as the parameters of the approximate straight line.

8. A system for controlling an unmanned vehicle by detection of a guide line from a picture image picked up by a camera on said vehicle, said system comprising:

picture picking-up means for picking up a picture image of a road on which said vehicle is running;

picture processing means for deriving color data from the picked-up picture image, and deriving a predetermined attention pixel having a predetermined picture image distribution around the same as an extracted point;

straight line approximation means for deriving a straight line data of the guide line on the road on which said vehicle is running from said extracted point data; and

steered amount determination means for determining a steered amount of said vehicle based on the straight line data by said straight line approximation means, and steering control means for controlling a steering of said vehicle based on a steered amount determined by said steered amount determination means and a current steering status.

9. A system according to claim 8, wherein said picture processing means comprises:

color obtaining means for extracting only a picture image signal having an intensity greater than a predetermined intensity from a received color picture image;

pixel detection means for detecting a pixel coordinate and pixel value from the picture image signal extracted by said color obtaining means;

attention coordinate determination means for determining whether or not the detected pixel coordinate is a stored attention pixel coordinate;

binarization processing means for transforming the detected pixel value into a binary data;

video speed delay storing means for transmitting the binary data in accordance with an output timing of said binarization processing means;

pixel distribution detection means for receiving a predetermined plurality of pixel values out of the binary data transmitted through said video speed delay storing means and determining whether or not the received pixel values are a predetermined combination of values; and

extracted point determination means for determining a pixel coordinate as the extracted point coordinate and supplying the determined pixel coordinate to said straight line approximation means when the detected pixel coordinate is determined as the stored attention signal by said attention coordinate determination means and the received pixel values are determined as the predetermined combination of values by said pixel distribution detection means.

12. A system according to claim 8, wherein a fuzzy inference is used for determining a steered amount.

10. A system according to claim 10, wherein a steered amount is determined based on a parameter containing an X-directional component of a perpendicular line from a reference point on the picture image to a guide line picture image.

11. A system according to claim 10, wherein a steered amount is determined based on change with time of a parameter containing an X-directional component of a perpendicular line from a reference point on the picture image to a guide line picture image.