Method and apparatus for calibrating a non-contact gauging sensor with respect to an external coordinate system

• US 4,841,460 A
• Filed: 09/08/1987
• Issued: 06/20/1989
• Est. Priority Date: 09/08/1987
• Status: Expired due to Term

• Current Assignee: Perceptron Incorporated (Atlas Copco AB)
• Original Assignee: Perceptron Incorporated (Atlas Copco AB)
• Inventors: Dewar, Robert, Greer, Dale R.
• Primary Examiner(s): Smith, Jerry
• Assistant Examiner(s): Gordon, Paul
• Application Number:
  US07/094,329
• Time in Patent Office: 651 Days
• Field of Search: 364/167, 364/571, 364/559, 364/516, 364/513, 364/558, 364/525, 356/3, 356/375, 356/376, 901/9, 901/47
• US Class Current: 702/95
• CPC Class Codes: G01B 11/002 for measuring two or more coordinates

Abstract

One or more non-contacting gauging sensors are calibrated to a common external coordinate system by providing each sensor to be calibrated with a target which can be simultaneously viewed by the sensor and an external measuring system for determining coordinates of the target in terms of the external coordinate system and in terms of the sensor coordinate system. With the resulting spatial coordinates of the target expressed in both frames of reference, transform matrix coefficients can be calculated, for example with a programmable computer, to calibrate each sensor to the preselected common external coordinate system.

Claims

1. A method for calibrating an arbitrarily oriented non-contact measurement sensor that is pre-calibrated and includes a structural light source and optical sensing means, with respect to an external coordinate system, the method comprising:

using the structured light from the sensor to create a target lying simultaneously within a measuring zone of the sensor and a measuring zone of measuring means external to the sensor;

using the sensor to determine first spatial coordinates of the target expressed in terms of the sensor'"'"'s coordinate system;

using the external measuring means to determine second spatial coordinates of the target expressed in terms of the external coordinate system; and

transforming the first spatial coordinates into the coordinates expressed in terms of the external system using the first and second spatial coordinates of the target.

2. The method of claim 1 wherein the sensor is coupled to a tactile element of a robot means for movement relative to the target.

3. The method of claim 1 wherein the target is coupled to a tactile element of a robot means for movement relative to the sensor.

4. The method of claim 1 wherein the structured light source from the non-contact gauging sensor emits at least one plane of light defining the sensor measuring zone.

5. The method of claim 4 wherein the external measuring means comprises at least two theodolites, each coupled to means for generating measurement data expressed in terms of the external coordinate system.

6. The method of claim 5 wherein the target comprises at least three points simultaneously viewable by both the optical sensing means of the sensor and the at least two theodolites.

7. The method of claim 5 wherein the target comprises four non-symetrically positioned points simultaneously viewable by the optical sensing means of the sensor and the at least two theodolites.

8. The method of claim 1 wherein the step of transforming further comprises:

(a) creating sensor vectors by subtracting the mean of all sensor coordinates from the coordinates of the target as measured by the sensor;

(b) creating external coordinate vectors by subtracting the mean of all exterior coordinates from the coordinates of the target as measured by the external measuring means;

(c) ordering pairs of sensor and external coordinates for all possible permutations thereof;

(d) for each permutation

(i) determining a pair of sensor vectors yielding the largest cross product vector therebetween;

(ii) generating a corresponding cross product for a corresponding pair of external coordinate vectors;

(iii) determining a least squares mapping of all sensor vectors and the largest sensor vector cross product to each of the coordinate axes for the external coordinate vectors and the corresponding external coordinate cross product; and

(iv) determining whether the resulting transformation results in orthogonality within a predetermined degree, and, if so, calculating transformation errors between the known external coordinates and the transformed sensor coordinates; and

(e) selecting the permutation resulting in the smallest transformation error.

9. A method for calibrating an arbitrarily oriented optical gauging sensor with respect to an external coordinate system, the method comprising;

providing a target lying simultaneously within a measuring zone of the sensor and a measuring zone of measuring means external to the sensor;

illuminating the target with a measuring light plane of a sensor optical source to generate at least three non-symetrical illuminated points on the target;

measuring coordinates of the at least three points with the sensor, the measured coordinates expressed in terms of the sensor coordinate system;

measuring coordinates of the at least three points with the external measuring means, the measured coordinates expressed in terms of the external coordinate system;

determining the mean value of each measured coordinate in the sensor coordinate system;

determining the mean value of each measured coordinate in the external coordinate system;

creating a sensor vector for each measured sensor coordinate point, each sensor vector extending from the sensor coordinate mean value point to a corresponding target point coordinate;

creating an external vector for each measured external coordinate point, each external vector extending from the external coordinate mean value point to a corresponding target point coordinate;

ordering pairs of sensor and external coordinates for all possible permutations thereof;

for each permutation, using a least squares approximation to determine a mapping of sensor vectors into external coordinate vectors with acceptable orthogonality; and

selecting the permutation resulting in the smallest transformation error.

10. Apparatus for calibrating an arbitrarily oriented non-contact measurement sensor that is pre-calibrated and includes a structured light source and optical sensing means, with respect to an external coordinate system comprising:

external measuring means;

target means lying simultaneously within a measuring zone of the sensor and a measuring zone of the external measuring means and including radiant target points created by the interaction of the target means and the structured light emitted by the sensor; and

transformation means coupled for receipt of radiant target point coordinate data relative to the target determined by both the sensor and the external measuring means, the transformation means operative to derive a transformation of sensor coordinate data into external coordinate data using the received target coordinate data from the sensor and the external measuring means.

11. The apparatus of claim 10 further comprising robot means coupled to the sensor for effecting movement of the sensor relative to the target means.

12. The apparatus of claim 10 further comprising robot means coupled to the target means for effecting movement of the target means relative to the sensor.

13. The apparatus of claim 10 wherein the structured light source from the non-contact measurement sensor emits at least one plane of light defining the sensor measuring zone.

14. The apparatus of claim 13 wherein the external measuring means comprises at least two theodolites, each coupled to means for generating measurement data expressed in terms of the external coordinate system.

15. The apparatus of claim 14 wherein the interaction between the structured light emitted by the sensor and target means produces at least three non-symmetrically positioned radiant target points simultaneously viewable by the sensor and the at least two theodolites.

16. The apparatus of claim 14 wherein the interaction between the structured light emitted by the sensor and target means produces four non-symmetrically positioned radiant target points simultaneously viewable by the sensor and the at least two theodolites.

17. The apparatus of claim 16 wherein the target means comprises means releasably coupled to the sensor and carrying four elongate members extending transversely to the at least one plane of light emitted by the structured light source from the sensor, each elongate member having characteristics enabling the generation of a radiant target point at the intersection of the light plane with the elongate member.

18. The apparatus of claim 17 wherein the elongate members extend substantially perpendicularly to the plane of light.

19. The apparatus of claim 17 wherein each elongate member comprises a fiber of white translucent material.

20. The apparatus of claim 10 wherein the transformation means comprises programmable computer means.

21. The apparatus of claim 20 wherein the computer means includes a program for performing the following steps:

(a) creating sensor vectors by subtracting the mean of all sensor coordinates from the coordinates of the target as measured by the sensor;

(b) creating external coordinate vectors by subtracting the mean of all exterior coordinates from the coordinates of the target as measured by the external measuring means;

(c) ordering pairs of sensor and external coordinates in all possible permutations thereof;

(d) for each permutation;

(i) determining a pair of sensor vectors yielding the largest cross product vector therebetween;

(ii) generating a corresponding cross product for a corresponding pair of external coordinate vectors;

(iii) determining a least squares mapping of all sensor vectors and the largest sensor vector cross product to each of the coordinate axes for the external coordinate vectors and the corresponding external coordinate cross product; and

(iv) determining whether the resulting transformation results in orthogonality within a predetermined degree, and if so, calculating the transformation errors between the known external coordinates and the transformed sensor coordinates; and

(e) selecting the permutation resulting in the smallest transformation error.

22. In a non-contact optical measuring system having a plurality of optical sensors prepositioned with respect to a workpiece to be measured, each optical sensor having means for determining coordinates of points located within the sensor field of view with respect to a sensor coordinate system, an arrangement for calibrating any of the plurality of sensors with respect to an external coordinate system comprising:

external measuring means;

target means releasably coupled to a sensor being calibrated, the target means carrying at least three non-symetrically positioned target point generating elements operative upon illumination by the sensor to generate at least three corresponding illuminated target points simultaneously viewable by the sensor and the external measuring means, such that the sensor can determine coordinates of each target point in terms of the sensor coordinate system and the external measuring means can determine coordinates of each target point in terms of the external coordinate system; and

transformation means coupled for receipt of sensor determined coordinate data and external measuring means determined coordinate data for the at least three target points, the transformation means operative to derive a matrix of transform coefficients effective to transform sensor coordinate measurement data into external coordinate measurement data, thereby calibrating a sensor with respect to the external coordinate system.

23. The arrangement of claim 22 wherein the target point generating elements each comprise elongate translucent members and the target means includes first and second side walls for coupling receipt of respective first and second ends of the elongate translucent members such that the elongate members extend generally transversely to a measuring plane of light emitted by the sensor to thereby produce the illuminated target points at intersections of each of the elongate members with the measuring plane of light.

24. The arrangement of claim 23 wherein each elongate translucent member is comprised of a substantially white fiber.

25. The arrangement of claim 23 wherein the elongate translucent members extend substantially perpendicularly to the measuring plane of light.

26. The arrangement of claim 23 further comprising coupling means rigidly but releasably coupled to the sensor and to the target means and including means for adjusting a distance between the sensor and the target means.

27. The arrangement of claim 26 wherein the coupling means includes a first mounting block coupled to the sensor, a second mounting block coupled to one of the first and second side walls and a cylindrical rod-like coupler member positioned for sliding receipt in mating bores in the first and second blocks, the means for adjusting including first and second set screw means respectively received in threaded bores in the first and second mounting blocks, with the threaded bores intersecting respective mating bores.

28. The arrangement of claim 22 wherein the target point generating elements each comprise elongate translucent members and the target means includes a substantially open-faced, cage-like structure for coupling receipt of respective first and second ends of the elongate translucent members such that the elongate members extend generally transversely to a measuring plane of light emitted by the sensor to thereby produce the illuminated target points at intersections of each of the elongate members with the measuring plane of light, the substantially open-faced structure thereby enabling line-of-sight viewing of the target points by the external measuring means from substantially all directions.

29. The arrangement of claim 22 wherein the transformation means comprises a programmable computer means.

30. The arrangement of claim 29 wherein the computer means is programmed to perform the following steps:

(a) determining the mean value of each measured coordinate in the sensor coordinate system;

(b) determining the mean value of each measured coordinate in the external coordinate system;

(c) creating a sensor vector for each measured sensor coordinate point, each sensor vector extending from the sensor coordinate mean value point to a corresponding target point coordinate;

(d) creating an external vector for each measured external coordinate point, each external vector extending from the external coordinate mean value point to a corresponding target point coordinate;

(e) ordering pairs of sensor and external coordinates for all possible permutations thereof;

(f) for each permutation

(i) determining a pair of sensor vectors yielding the largest cross product vector therebetween;

(ii) generating a corresponding cross product for a corresponding pair of external coordinate vectors;

(iii) determining a least squares mapping of all sensor vectors and the largest sensor vector cross product to each of the coordinate axes for the external coordinate vectors and the corresponding external coordinate cross product; and

(iv) determining whether the resulting transformation results in orthogonality in a predetermined degree and, if so, calculating the transformation errors between the known external coordinates and the transformed sensor coordinates; and

(g) selecting the permutation resulting in the smallest transformation error.

31. A method of calibrating an arbitrarily oriented, pre-calibrated, non-contact measurement sensor with respect to an external coordinate system, the method comprising:

calibrating an independent external measuring system including a pair of theodolites relative to said external coordinate system;

providing a target lying simultaneously within the measuring zone of the sensor and the measuring zone of the theodolites;

using the sensor to determine first spatial coordinates of the target expressed in terms of the sensor'"'"'s coordinate system;

using the theodolites to determine second spatial coordinates of the target expressed in terms of the external coordinate system; and

transforming the first spatial coordinates into coordinates expressed in terms of the external coordinate system using the first and second spatial coordinates of the target.

32. The method of claim 31 wherein said sensor is an optical sensor including a structured light source and optical sensing means and further wherein the structured light source from the optical sensor is used to create the target.

33. Apparatus for calibrating an arbitrarily oriented, pre-calibrated, non-contact measurement sensor having a measuring zone, with respect to an external coordinate system, comprising:

an independent external measuring system including a pair of theodolites calibrated relative to said external coordinate system;

target means lying simultaneously within the measuring zone of the sensor and the measuring zone of said theodolites; and

transformation means coupled for receipt of coordinate data relative to the target determined by both the sensor and the external measuring means, the transformation means operative to derive a transformation of sensor coordinate data into external coordinate data using the received target coordinate data from the sensor and the external measuring means.

34. The apparatus of claim 33 wherein the non-contact measurement sensor comprises an optical sensor including a structured light source and optical sensing means and further wherein the interaction of the structured light emitted from the optical sensor and said target means creates a predetermined target pattern.

35. The apparatus of claim 34 wherein said predetermined target pattern comprises a plurality of non-symmetrically positioned radiant target points.

36. The apparatus of claim 35 wherein said target means includes elongated fibers of translucent material.

37. A method of calibrating an arbitrarily oriented, pre-calibrated, non-contact measurement sensor with respect to an external coordinate system, comprising the steps of:

calibrating an independent external measuring system relative to the external coordinate system;

providing a target lying simultaneously within the measurement zone of the sensor and the measurement zone of the independent external measuring system;

using the sensor to determine the spatial coordinates of the target in the sensor'"'"'s coordinate system;

using the independent external measuring system to determine the translational position and rotational orientation of the target relative to the external coordinate system;

using the data from the previous step to transform the spatial coordinates from the sensor'"'"'s coordinate system to the external coordinate system.

38. The method of claim 37 wherein the sensor includes a structured light source and optical sensing means and further wherein the target is created with the structured light source from the sensor.