Apparatus and method for controlling a moving vehicle utilizing a digital differential analysis circuit

• US 5,379,353 A
• Filed: 09/22/1993
• Issued: 01/03/1995
• Est. Priority Date: 05/09/1988
• Status: Expired due to Fees

• Current Assignee: Honda Giken Kogyo Kabushiki Kaisha (Honda Motor Company)
• Original Assignee: Honda Giken Kogyo Kabushiki Kaisha (Honda Motor Company)
• Inventors: Yshushi, Okada, Hasegawa, Hiroshi
• Primary Examiner(s): Couso, Jose L.
• Application Number:
  US08/125,350
• Time in Patent Office: 468 Days
• Field of Search: 382/28, 382/41, 382/43, 382/48, 364/461, 364/424.02, 364/424.07, 358/103
• US Class Current: 382/104
• CPC Class Codes: G05D 1/0246 using a video camera in combination with image processing means
• CPC Class Codes: G06V 10/48 by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
• 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

This invention relates to a picture processing device in which picture data of a road or lines marking off a lane taken by a television camera installed on a vehicle is processed for the recognition of the contour of the road or the lane. Picture data is analog/digital converted, then edited and is subjected to the binary processing to be supplied to a digital differential analysis computing circuit. The digital differential analysis circuit comprises circuits serially connected each of which computes a rotary motion recurrence formula every rotation by a rotation angle. The recurrence formula is computed in the pipe-line system. In this computation the Hough transform is performed to give sine curves in a ρ-θ coordinate system. The points of intersection of these sine curves (Hough curves) are given, and those of them having more occurrences are selected in the ρ-θ coordinate system. Resultantly, the contour of the shoulder lines and the center line of the road, or the contour of the lane are given.

Claims

1. A device for detecting a boundary of a roadway upon which a vehicle travels, the device comprising:

camera means for generating electronic signals representing an image of the roadway, the image comprising a view of the roadway extending in a direction of travel of the vehicle;

a signal inputting unit for digitizing the signals representing the image of the roadway;

an edge detecting unit for processing the digitized signals to produce edged picture element signals differentiated with respect to pixel to pixel brightness variations;

a preprocessing unit for receiving the edged picture element signals from the edge detecting unit and storing in a multivalue memory circuit edged picture data corresponding to the edged picture element signals, the edged picture data corresponding to each edged picture element signal comprising an address in an X-Y coordinate system and an edged brightness value;

digital differential analysis computing means including a plurality of digital differential analysis computing circuits which are serially connected, each for converting the addresses of the edged picture data from the X-Y coordinate system to an α

-β

coordinate system and computing a rotary motion recurrence formula to generate a plurality of approximate HOUGH curves in a ρ

-θ

coordinate system, the approximate HOUGH curves intersecting at a plurality of points in the ρ

-θ

coordinate system, and said rotary motion recurrence formula comprising the equation

space="preserve" listing-type="equation">α

.sub.i+1 =α

.sub.i -2.sup.-m β

.sub.i ;

and

space="preserve" listing-type="equation">β

.sub.i+1 =2.sup.-m α

.sub.i+1 +β

.sub.i

wherein an angle of rotation comprises

space="preserve" listing-type="equation">ε

=2.sup.-m (rad) (m=0, 1, 2, . . . );

a histogram memory circuit for storing address data corresponding to the intersection points between the generated approximate HOUGH curves; and

a processor for generating a signal representing the boundary of the roadway based on the stored address data corresponding to the intersection points of the approximate HOUGH curves.

2. The device of claim 1, wherein the digital differential analysis computing means further comprises an initial value computing circuit for computing coordinate (α

_i, β

_i) of an initial value in the α

-β

coordinate system, and the digital differential analysis computing circuits perform the rotary motion recurrence formula computation based on the coordinates (α

_i, β

_i) of the initial value.

3. The device of claim 1, wherein the processor comprises a peak filtering circuit for identifying intersection points having relative maximum frequencies of intersection by the approximate Hough curves, and the processor generates the signal indicative of the boundary of the roadway based on the stored address data corresponding to the intersection points having relative maximum frequencies of intersection.

4. A machine implemented process for detecting a boundary of a roadway upon which a vehicle travels, the process comprising the steps of:

generating electronic signals representing an image of the roadway, the image comprising a view of the roadway in a direction of travel of the vehicle;

digitizing the signals representing the image of the roadway;

generating edged picture element signals from the digitized signals, the edged picture element signals being differentiated with respect to pixel to pixel brightness;

storing in a memory circuit edged picture data corresponding to the edged picture element signals, the edges picture data comprising addresses in an X-Y coordinate system and edged brightness values;

converting the addresses of the stored edged picture data from the X-Y coordinate system to an α

-β

coordinate system and computing a rotary motion recurrence formula to generate a plurality of approximate HOUGH curves in a ρ

-θ

coordinate system, the approximate HOUGH curves intersecting at a plurality of points in the ρ

-θ

coordinate system, using a digital differential analysis computing means including a plurality of serially connected differential analysis computing circuits, said rotary motion recurrence formula comprising the equations

space="preserve" listing-type="equation">α

.sub.i+1 =α

.sub.i -2.sup.-m β

.sub.i ;

and

space="preserve" listing-type="equation">β

.sub.i+1 =2.sup.-m α

.sub.i+1 +β

.sub.i

wherein an angle of rotation comprises

space="preserve" listing-type="equation">ε

=.sup.- m (rad) (m=0, 1, 2, . . . );

storing address data corresponding to the intersection points between the generated approximate HOUGH curves and a histogram memory circuit; and

generating a signal representing the boundary of the roadway based on the stored address data corresponding to the intersection points of the approximate HOUGH curves.

5. A machine implemented process for generating a plurality of approximate HOUGH curves in a ρ

-θ

coordinate system, said process comprising the steps of;

generating pixel data representative of an image, said pixel data comprising addresses in an X-Y coordinate system and pixel brightness values;

storing said pixel data in a memory circuit;

retrieving said pixel data from said memory circuit; and

using a digital differential analysis computing means including an initial value computing circuit and a plurality of serially connected digital differential analysis circuits, converting the addresses of the retrieved pixel data from the X-Y coordinate system to an α

-β

coordinate system and processing said converted pixel data using a rotary motion recurrence formula to generate a plurality of approximate HOUGH curves in a ρ

-θ

coordinate system, said rotary motion recurrence formula being defined by the following equations;

space="preserve" listing-type="equation">α

.sub.i+1 =f.sub.α

(α

.sub.i,β

.sub.i,ε

)

space="preserve" listing-type="equation">β

.sub.i+1 =f.sub.β

(α

.sub.i,β

.sub.i,ε

).

6. The machine implemented process of claim 5, wherein said rotary motion recurrence formula comprises the equations

7. The machine implemented process of claim 5, wherein said rotary motion recurrence formula comprises the equations

8. The machine implemented process of claim 5, wherein said rotary motion recurrence formula comprises the equations

9. A device for detecting a boundary of a roadway upon which a vehicle travels, the device comprising:

camera means for generating electronic signals representing an image of the roadway, the image comprising a view of the roadway extending in a direction of travel of the vehicle;

a signal inputting unit for digitizing the signals representing the image of the roadway;

an edge detecting unit for processing the digitized signals to produce edged picture element signals differentiated with respect to pixel to pixel brightness variations;

a preprocessing unit for receiving the edged picture element signals from the edge detecting unit and storing in a multivalue memory circuit edged picture data corresponding to the edged picture element signals, the edged picture data corresponding to each edged picture element signal comprising an address in an X-Y coordinate system and an edged brightness value;

digital differential analysis computing means including a plurality of digital differential analysis computing circuits which are serially connected, each for converting the addresses of the edged picture data from the X-Y coordinate system to an α

-β

coordinate system and computing a rotary motion recurrence formula to generate a plurality of approximate HOUGH curves in a ρ

-θ

coordinate system, the approximate HOUGH curves intersecting at a plurality of points in the ρ

-θ

coordinate system, and said rotary motion recurrence formula comprising the equation

space="preserve" listing-type="equation">α

.sub.i+1 =α

.sub.i (1-2.sup.-2m-1)-2.sup.-m β

.sub.i ;

and

space="preserve" listing-type="equation">β

.sub.i+1 =2.sup.-m α

.sub.i +β

.sub.i (1-2.sup.-2m-1)

wherein an angle of rotation comprises

ε

=^- m (rad) (m=0, 1, 2, . . . );

a histogram memory circuit for storing address data corresponding to the intersection points between the generated approximate HOUGH curves; and

a processor for generating a signal representing the boundary of the roadway based on the stored address data corresponding to the intersection points of the approximate HOUGH curves.

10. A device for detecting a boundary of a roadway upon which a vehicle travels, the device comprising:

camera means for generating electronic signals representing an image of the roadway, the image comprising a view of the roadway extending in a direction of travel of the vehicle;

a signal inputting unit for digitizing the signals representing the image of the roadway;

an edge detecting unit for processing the digitized signals to produced edged picture element signals differentiated with respect to pixel to pixel brightness variations;

a preprocessing unit for receiving the edged picture element signals from the edge detecting unit and storing in a multivalue memory circuit edged picture data corresponding to the edged picture element signals, the edged picture data corresponding to each edged picture element signal comprising an address in an X-Y coordinate system and an edged brightness value;

digital differential analysis computing means including a plurality of digital differential analysis computing circuits which are serially connected, each for converting the addresses of the edged picture data from the X-Y coordinate system to an α

-β

coordinate system and computing a rotary motion recurrence formula to generate a plurality of approximate HOUGH curves in a ρ

-θ

coordinate system, the approximate HOUGH curves intersecting at a plurality of points in the ρ

-θ

coordinate system, and said rotary motion recurrence formula comprising the equations

space="preserve" listing-type="equation">α

.sub.i+1 =α

.sub.i (1-2-2m-1)+β

.sub.i (-2.sup.-m +ε

.sup.-m /6);

and

space="preserve" listing-type="equation">β

.sub.i+1 =α

.sub.i (2.sup.-m +ε

.sup.-m /6)+β

.sub.i (1-2.sup.-2m-1)

wherein an angle of rotation comprises

space="preserve" listing-type="equation">ε

=.sup.- m (rad) (m=0, 1, 2, . . . );

a histogram memory circuit for storing address data corresponding to the intersection points between the generated approximate HOUGH curves; and

a processor for generating a signal representing the boundary of the roadway based on the stored address data corresponding to the intersection points of the approximate HOUGH curves.

11. A machine implemented process for detecting a boundary of a roadway upon which a vehicle travels, the process comprising the steps of:

generating electronic signals representing an image of the roadway, the image comprising a view of the roadway in a direction of travel of the vehicle;

digitizing the signals representing the image of the roadway;

generating edged picture element signals from the digitized signals, the edged picture element signals being differentiated with respect to pixel to pixel brightness;

storing in a memory circuit edged picture data corresponding to the edged picture element signals, the edges picture data comprising addresses in an X-Y coordinate system and edged brightness values;

converting the addresses of the stored edged picture data from the X-Y coordinate system to an α

-β

coordinate system and computing a rotary motion recurrence formula to generate a plurality of approximate HOUGH curves in a ρ

-θ

coordinate system, the approximate HOUGH curves intersecting at a plurality of points in the ρ

-θ

coordinate system, using a digital differential analysis computing means including a plurality of serially connected differential analysis computing circuits, said rotary motion recurrence formula comprising the equations

space="preserve" listing-type="equation">α

.sub.i+1 =α

.sub.i (1-2.sup.-2m-1)-2.sup.-m β

.sub.i ;

and

space="preserve" listing-type="equation">β

.sub.i+1 =2.sup.-m α

.sub.i+β

.sub.i (1-2.sup.-2m-1)

wherein an angle of rotation comprises

space="preserve" listing-type="equation">ε

=2.sup.-m (rad) (m=0, 1, 2, . . . );

storing address data corresponding to the intersection points between the generated approximate HOUGH curves and a histogram memory circuit; and

generating a signal representing the boundary of the roadway based on the stored address data corresponding to the intersection points of the approximate HOUGH curves.

12. A machine implemented process for detecting a boundary of a roadway upon which a vehicle travels, the process comprising the steps of:

generating electronic signals representing an image of the roadway, the image comprising a view of the roadway in a direction of travel of the vehicle;

digitizing the signals representing the image of the roadway;

generating edged picture element signals from the digitized signals, the edged picture element signals being differentiated with respect to pixel to pixel brightness;

storing in a memory circuit edged picture data corresponding to the edged picture element signals, the edges picture data comprising addresses in an X-Y coordinate system and edged brightness values;

converting the addresses of the stored edged picture data from the X-Y coordinate system to an α

-β

coordinate system and computing a rotary motion recurrence formula to generate a plurality of approximate HOUGH curves in a ρ

-θ

coordinate system, the approximate HOUGH curves intersecting at a plurality of points in the ρ

-θ

coordinate system, using a digital differential analysis computing means including a plurality of serially connected differential analysis computing circuits, said rotary motion recurrence formula comprising the equations

space="preserve" listing-type="equation">α

.sub.i+1 =α

.sub.i (1-2-2m-1)+β

.sub.i (-2.sup.-m +ε

.sup.-3m /6);

and

space="preserve" listing-type="equation">β

.sub.i+1 =α

.sub.i (2.sup.-m +ε

.sup.-3m /6)+β

.sub.i (1-2.sup.-2m-1)

wherein an angle of rotation comprises

space="preserve" listing-type="equation">ε

=2.sup.-m (rad) (m=0, 1, 2, . . . );

storing address data corresponding to the intersection points between the generated approximate HOUGH curves and a histogram memory circuit; and

generating a signal representing the boundary of the roadway based on the stored address data corresponding to the intersection points of the approximate HOUGH curves.