This disclosure relates to methods and tools for calibrating a camera-based three-dimensional optical metrology system such as that used for non-contact measurement of strain in a materials test system.
A three-dimensional optical metrology system, in order to perform non-contact measurement of strain of a sample during materials testing, typically combines multiple camera images from different perspective views of the same object in order to obtain the depth information necessary to provide a single three-dimensional view of the object. Each camera viewpoint has to be precisely calibrated in three dimensions in order to produce accurate three-dimensional measurements of the object under test. For each viewpoint, or camera, the calibration must perform the following actions:
1. Locate the pointing vector of the camera viewpoint in three-dimensional space. This establishes the position of the camera and its direction of view.
2. Define the orientation of the two-dimensional image sensor within the camera relative to its direction of view (pointing vector). This establishes any tilts or rotations of the image sensor.
3. Calibrate the distortions of the two-dimensional camera image produced by its optical system including lens distortion, alignment of lens relative to image sensor and any non-uniformity effects of the image sensor.
4. Calibrate changes to the two-dimensional camera image produced by object position changes in depth axis.
When each camera is thus calibrated, the combined calibration data sets define a common three-dimensional measurement volume in object space. The metrology system can then make calibrated three-dimensional measurements anywhere within this common volume.
This calibration process is typically accomplished by taking a series of pictures from each camera of a flat calibration plate that is held at multiple locations and orientations within the object space. The flat calibration plate provides an array of marks such as, for example, dots with known pre-calibrated diameter and spacing. The process is as follows:
1. An image of the calibration plate from each camera viewpoint is taken as the plate is held stationary at a given position in the three-dimensional measurement volume.
2. The calibration plate is then moved to new three-dimensional position within the measurement volume and an additional set of images is taken, one from each camera. The new location may typically include tilting and/or rotating the plate as well as translating it in the depth axis.
3. This process is repeated, progressively moving the calibration plate through the intended measurement volume. The calibration plate may alternatively be moved to random locations and orientations within the volume, being careful to sample the whole volume with the calibration plate.
4. The process ends when sufficient sampling of the volume is made—typically requiring 6 to 12 different calibration plate positions within the volume.
Any apparatus used for calibration must be easy to insert and remove from the measurement volume. It must be removed after calibration is complete to enable the user make three-dimensional measurements of an arbitrary specimen in the same volume. Therefore, it is a challenge to have a motorized means of moving the calibration plate that is simple, compact and self-contained. For this reason, the typical method of providing the necessary position changes of the calibration plate is for the operator to hold the plate by hand and move it about manually while the cameras take snapshots at intervals thereby capturing the plate at many unique locations and orientations within the target volume.
Generally, the exact locations/orientations of the calibration plate in the measurement volume do not have to be known. Instead, the calibration method relies on the known geometry of the features on the calibration plate to provide the necessary information for calibrating the array of images from each camera.
Prior art methods typically use a hand-held calibration plate 200 that is moved about the target calibration volume randomly as shown in
The following references provide examples of mathematical algorithms commonly used to produce a three-dimensional calibration map from similar sets of calibration plate images as described in this disclosure.
It is the object of this disclosure to provide a method of calibrating a three-dimensional optical metrology system that eliminates need for manual movements of a calibration plate and provides more repeatable and systematic calibration results. This provides greater ease of use, reduces level of operator training for calibration and reduces operator variability as a factor in final results.
This and other objects are attained by providing a mechanical device, which is typically engaged or clamped by a grip within a materials testing device, which supports a spindle with periodic detent rotational positions. The spindle further includes supports upon which a planar visual target (such as a calibration plate) pivots to desired angular positions. The rotation of the spindle provides for rotation about an axis to pre-determined detent positions, thereby providing a first degree of freedom of movement in the positioning of the visual target while the pivoting of the visual target provides a second degree of freedom of movement. Typically, two cameras are used to capture images of the visual target throughout a range of positions of the visual target thereby defining a common three-dimensional measurement volume in object space for subsequent non-contact measurement of strain of a sample in materials testing.
Further objects and advantages of the disclosure will become apparent from the following description and from the accompanying drawings, wherein:
Referring now to the drawings in detail, wherein like numerals indicate like elements throughout the several views, one sees the disclosed embodiment of present method and apparatus. As shown in
The wedge 10 that rotates about a fixed rotational axis 12, as shown in
Typically, the key element is that both the rotational angle 17 about axis 14 and tilt angle 16 can be kept to calibrated values, independent of operator manual input to affect rotational angle 17 or tilt angle 16, thereby providing repeatability and predictability to the process and method. All camera views of the calibration plate 14 can optimally fill the camera field of view without risk of skewing the calibration plate 14 to one side of the camera image by uncontrolled manual movements of the calibration plate 14.
As shown in
It is further typically important the calibration tool 30 be quickly and easily installed, be completely self-contained, require no critical alignments and be simple for the operator to use. As shown in
The vertical mounting tab 32 is attached to a lower horizontal support member 34, which, in turn, is attached to a vertical spindle support member 36, thereby providing a frame 39 (typically made of injected plastic or metal) for the rotatable mounting of spindle 38. The spindle 38 is journaled for rotation and rotatably mounted on the upper end of the vertical spindle support member 36. The rotatable mounting of the spindle 38 on vertical support member 36 includes a detent function so that the spindle 38 rotatably moves to repeatable discrete rotational positions (thereby implementing the changes in rotational angle 17 about axis 12 in
A view of the rotating wedge calibration tool 30 from roughly the viewpoint of one of the first and second cameras 20, 22 is shown in
In this embodiment, the spindle 38 may be rotated manually about its axis by the user. However, as shown in
Some embodiments may include a second pattern of calibrated images on the back or reverse side of the calibration plate 14. For example, black dots on a white background. The design of the calibration tool 30 allows for the calibration plate 14 to be easily removed from the spindle 34 and flipped over or reversed so that a single calibration plate 14 may provide two different sets of calibration images. This may be useful, for example, to change scale or type of features for calibration of different three-dimensional system cameras.
Typical advantages of the disclosed calibration tool and method may include:
1. Eliminates the need for operator to manually move a hand-held calibration plate about the three-dimensional image volume to be calibrated.
2. Provides repeatable sampling of the three-dimensional image volume for every calibration run.
3. Ease of use.
4. Improved calibration accuracy and repeatability.
5. Capable of reducing the number of unknown parameters in the calibration process.
6. Capable of providing automatic calibration plate rotation without additional motors or complexity.
7. Method is scalable upwards or downwards to calibration plate sizes that are impractical for hand-held methods.
Thus the several aforementioned objects and advantages are most effectively attained. Although preferred embodiments of the invention have been disclosed and described in detail herein, it should be understood that this invention is in no sense limited thereby.
This application is a National Phase application of PCT/US2017/032100 filed May 11, 2017 which claims priority under 35 U.S.C. § 119(e) of U.S. provisional application Ser. No. 62/340,766 filed on May 24, 2016, the contents of which is hereby incorporated by reference in its entirety and for all purposes.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/US2017/032100 | 5/11/2017 | WO | 00 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2017/205061 | 11/30/2017 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
5173679 | Mount | Dec 1992 | A |
7995098 | Pedeville | Aug 2011 | B2 |
8619144 | Chang et al. | Dec 2013 | B1 |
8662676 | Chang | Mar 2014 | B1 |
10445898 | Liu | Oct 2019 | B2 |
10453218 | Doganis | Oct 2019 | B2 |
10502552 | Hoehl | Dec 2019 | B2 |
10510162 | Doganis | Dec 2019 | B2 |
10552984 | Virodov | Feb 2020 | B2 |
20030190116 | Freeman | Oct 2003 | A1 |
20090249606 | Diez | Oct 2009 | A1 |
20110026014 | MacK | Feb 2011 | A1 |
20120105591 | Kim et al. | May 2012 | A1 |
20120176629 | Allen | Jul 2012 | A1 |
20120242839 | Shen | Sep 2012 | A1 |
20120287248 | Erdman, III | Nov 2012 | A1 |
20130063570 | Michopoulos | Mar 2013 | A1 |
20130070048 | Huang | Mar 2013 | A1 |
20130083168 | Kim | Apr 2013 | A1 |
20150273381 | Stoner, Jr. et al. | Oct 2015 | A1 |
20170221210 | Martinello | Aug 2017 | A1 |
20170228864 | Liu | Aug 2017 | A1 |
20190154554 | Iwakiri | May 2019 | A1 |
Number | Date | Country |
---|---|---|
102519710 | Jun 2012 | CN |
103048872 | Apr 2013 | CN |
203185190 | Sep 2013 | CN |
203426903 | Feb 2014 | CN |
104792297 | Jul 2015 | CN |
204887156 | Dec 2015 | CN |
1973109340 | Dec 1973 | JP |
2003172713 | Jun 2003 | JP |
2009145278 | Jul 2009 | JP |
2013169628 | Sep 2013 | JP |
20150094928 | Aug 2015 | KR |
2012135014 | Oct 2012 | WO |
2016179798 | Nov 2016 | WO |
Entry |
---|
International Search Report issued in PCT/US2017/032100 dated Jul. 6, 2017. |
Japanese Office Action issued in 2018-561650 dated Dec. 22, 2020. |
Number | Date | Country | |
---|---|---|---|
20190212139 A1 | Jul 2019 | US |
Number | Date | Country | |
---|---|---|---|
62340766 | May 2016 | US |