Method and system for automatically determining the position and attitude of an object

• US 4,639,878 A
• Filed: 06/04/1985
• Issued: 01/27/1987
• Est. Priority Date: 06/04/1985
• Status: Expired due to Term

• Current Assignee: Fanuc Robotics America Incorporated (Fanuc Corporation)
• Original Assignee: GMF Robotics Corporation
• Inventors: Horn, Berthold K. P., Whitcomb, Louis L., Krause, Kenneth W., Day, Chia P.
• Primary Examiner(s): NOT, DEFINED
• Assistant Examiner(s): Lastova, John R.
• Application Number:
  US06/741,123
• Time in Patent Office: 602 Days
• Field of Search: 364/191-193, 364/513, 364/516, 364/571, 364/478, 901/9, 901/47, 356/2, 356/375, 356/376, 358/101, 358/107
• US Class Current: 700/259
• CPC Class Codes: G01B 11/002 for measuring two or more coordinates
• CPC Class Codes: G01B 2210/12 Method or fixture for calibrating the wheel aligner
• CPC Class Codes: G01B 2210/146 Two or more cameras imaging the same area
• CPC Class Codes: G05B 19/4083 Adapting programme, configuration
• CPC Class Codes: G05B 2219/36503 Adapt program to real coordinates, software orientation
• CPC Class Codes: G05B 2219/37572 Camera, tv, vision
• CPC Class Codes: G06V 10/147 Details of sensors, e.g. sensor lenses fingerprint or palmprint sensors G06V40/13; vascular sensors G06V40/145; eye sensors G06V40/19
• CPC Class Codes: G06V 2201/12 Acquisition of 3D measurements of objects

Abstract

A method and system are provided for automatically determining the position and attitude of a three-dimensional body at a work station by utilizing three cameras each of which generates a non-overlapping plane of image data, including a single target point of the body without the use of structured light. The locations of the target points are determined and processed within a programmed computer together with calibration data relating to the expected position of the body in the work station to provide data relating to the position and attitude of the body with respect to the work station. The resultant data relating to the position and attitude of the body are subsequently transformed into a first set of offset data relating to the difference between the actual and expected positions of the body. The first set of offset data is then transformed into a second set of offset data in the coordinate system of a peripheral device such as a robot, programmable controller, numerically controlled machine, etc. Finally, the programmed computer transfers the second set of offset data to the peripheral device which utilizes the data to modify its pre-programmed path.

Claims

1. Method for automatically determining the position and attitude of a three-dimensional body in space having at least three spaced apart target points at a vision station having a fixed coordinate system, said method comprising the steps of:

generating calibration data defining a nominal position of the body at the vision station;

generating substantially simultaneously a single plane of image data for each of said target points, each plane of image data containing its respective target point and each plane of image data being non-overlapping with respect to each of the other planes of image data;

determining the locations of the target points within their respective image planes to obtain positional data; and

processing the calibration data and the positional data together to provide data defining the position and attitude of the body with respect to the fixed coordinate system.

2. The method as claimed in claim 1 wherein said step of generating calibration data includes the steps of positioning a graduated calibration fixture at said vision station;

and

illuminating the fixture with light to form at least three fixture images in three-dimensional space.

3. The method as claimed in claim 2 wherein said step of generating calibration data includes the step of transforming the fixture images in the three-dimensional space to three planes of calibration data, each of said planes of calibration data including data containing at least four target points on the calibration fixture.

4. The method as claimed in claim 1 wherein said step of generating three planes of image data includes the step of obtaining an intermediate set of data containing a linear projection of each of said target points.

5. The method as claimed in claim 4 wherein said step of generating three planes of image data includes the step of modifying the intermediate set of data to provide perspective compensation of said image data.

6. Method for automatically determining the position and attitude of a three-dimensional body located at a vision station having a fixed coordinate system without the use of structured lighting, the body having at least three spaced apart target points, said method comprising the steps of:

generating calibration data defining a nominal position of the body at the vision station;

illuminating the body with unstructured light to form at least three images in three-dimensional space;

transforming substantially simultaneously the images in three-dimensional space to three planes of image data, each plane of image data containing its respective target point and each plane of image data being non-overlapping with respect to each of the other planes of image data;

determining the locations of the target points within their respective image planes to obtain positional data; and

processing the calibration data and the positional data together to provide data defining the position and attitude of the body with respect to the fixed coordinate system.

7. A method of automatically providing path compensation data to a programmed robot controller, to enable a robot controlled by the controller to move along a new path different from the path originally programmed in the controller to work on a three-dimensional body at a work station having a fixed coordinate system, the compensation data defining the difference between actual and nominal positions of the three-dimensional body at the work station, the method comprising the steps of:

generating calibration data defining the nominal position of the three-dimensional body at the work station;

generating substantially simultaneously three planes of image data, each of said planes of image data including image data containing a single target point on said body and each plane of image data being non-overlapping with respect to each of the other planes of image data;

determining the position of each of said target points within each of said image planes to obtain positional data;

processing the calibration data and the positional data of the target points together to obtain location data defining the position and attitude of the body with respect to the fixed coordinate system;

transforming the location data into a first set of offset data defining the difference between the actual and nominal positions of the body in said fixed coordinate system;

transforming the first set of offset data into a second set of offset data in the coordinate system of the robot; and

transferring the second set of offset data to the controller.

8. A method of automatically providing path compensation data to a programmed robot controller, to enable a robot controlled by the controller to move along a new path different from the path originally programmed in the controller to work on a three-dimensional body at a work station having a fixed coordinate system, the compensation data defining the difference between actual and nominal positions of the three-dimensional body at the work station, the method comprising the steps of:

generating calibration data defining the nominal position of the three-dimensional body at the work station;

illuminating the body with unstructured light to form at least three non-overlapping images in three-dimensional space;

transforming substantially simultaneously the images in three-dimensional space to three planes of image data, each of said planes of image data including image data containing a single target point on said body and each plane of image data being non-overlapping with respect to each of the other planes of image data;

determining the position of each of said target points within each of said image planes to obtain positional data;

processing the calibration data and the positional data of the target points together to obtain location data defining the position and attitude of the body with respect to the fixed coordinate system;

transforming the resultant data into a first set of offset data defining the difference between the actual and nominal positions of the body in said fixed coordinate system;

transforming the first set of offset data into a second set of offset data in the coordinate system of the robot; and

transferring the second set of offset data to the controller.

9. A system for automatically determining the position and attitude of a three-dimensional body in space having at least three spaced apart target points at a vision station having a fixed coordinate system, said system comprising:

means for storing calibration data defining a nominal position of the body at the vision station;

at least three spaced apart cameras, each of said cameras operating substantially simultaneously with each of the other cameras and generating a plane of image data containing a single target point and each plane of image data being non-overlapping with respect to each of the other planes of image data;

means for providing positional data defining the location of the target points within their respective image planes; and

means for processing the calibration data and the positional data together to provide data defining the position and attitude of the body with respect to the fixed coordinate system.

10. A system for automatically determining the position and attitude of a three-dimensional body in space having at least three spaced apart target points at a vision station having a fixed coordinate system without the use of structured light, said system comprising:

means for storing calibration data defining a nominal position of the body at the vision station;

at least three spaced apart cameras mounted at predetermined positions at the vision station, each of said cameras having a field of view;

means for illuminating the body with unstructured light to form at least three images in three-dimensional space;

means for moving the body relative to said cameras in the vision station so that each of the target points is positioned within the field of view of its respective camera, each of said cameras operating substantially simultaneously with each of the other cameras to transform the images in three-dimensional space to three planes of image data, each plane of image data containing its respective target point and each plane of image data being non-overlapping with respect to each of the other planes of image data;

means for providing positional data defining the location of the target points within their respective image planes; and

means for processing the calibration data and the positional data together to provide data defining the position and attitude of the body with respect to the fixed coordinate system.

11. The system as claimed in claim 9 or 10 including means for generating the calibration data.

12. The system as claimed in claim 11 wherein said means for generating calibration data includes a graduated calibration fixture adapted to be positioned at said vision station within the field of view of each of said cameras.

13. The system as claimed in claim 12 wherein said fixture includes a cubic solid.

14. A system for automatically providing path compensation data to a programmed robot controller, to enable a robot controlled by the controller to move along a new path different from the path originally programmed in the controller to work on a three-dimensional body at a work station having a fixed coordinate system, the compensation data defining the difference between the actual and nominal positions of the three-dimensional body at the work station, the system comprising:

means for storing calibration data defining the nominal position of the three-dimensional body at the work station;

at least three spaced apart cameras mounted at predetermined positions in the work station, each of said cameras having a field of view;

means for moving the body at the work station so that each of the target points is positioned within the field of view of its respective camera, each of said cameras operating substantially simultaneously with each of the other cameras and generating a plane of image data containing its respective target point and each plane of image data being non-overlapping with respect to each of the other planes of image data;

means for providing positional data defining the location of the target points within their respective image planes;

means for processing the calibration data and the positional data of the target points together to obtain location data defining the position and attitude of the body with respect to the fixed coordinate system;

means for transforming the location data into a first set of offset data defining the difference between the actual and nominal positions of the body in said fixed coordinate system;

means for transforming the first set of offset data into a second set of offset data in the coordinate system of the robot; and

means for transferring the second set of offset data to the controller.

16. The system as claimed in claim 14 or 15 including means for generating calibration data.

17. The system as claimed in claim 16 wherein said means for generating calibration data includes a graduated calibration fixture adapted to be positioned at said vision station within the field of view of each of said cameras.

18. The system as claimed in claim 17 wherein said fixture includes a cubic solid.

15. A system for automatically providing path compensation data to a programmed robot controller, to enable a robot controlled by the controller to move along a new path different from the path originally programmed in the controller to work on a three-dimensional body at a work station having a fixed coordinate system, the compensation data defining the difference between actual and nominal positions of the three-dimensional body at the work station, the system comprising:

means for storing calibration data defining the nominal position of the three-dimensional body at the work station;

at least three spaced apart cameras mounted at predetermined positions in the work station, each of said cameras having a field of view;

means for illuminating the body with unstructured light to form at least three images in three-dimensional space;

means for moving the body in the vision station so that each of the target points is positioned within the field of view of its respective camera, each of said cameras operating substantially simultaneously with each of the other cameras to transform the images in three-dimensional space to three planes of image data, each of said planes of image data containing its respective target point and each plane of image data being non-overlapping with respect to each of the other planes of image data;

means for providing positional data defining the location of the target points within their respective image planes;

means for processing the calibration data and the positional data of the target points together to obtain location data defining the position and attitude of the body with respect to the fixed coordinate system;

means for transforming the location data into a first set of offset data defining the difference between the actual and nominal positions of the body in said fixed coordinate system;

means for transforming the first set of offset data into a second set of offset data in the coordinate system of the robot; and

means for transferring the second set of offset data to the controller.