CN103234454B - A kind of self-calibrating method of image measurer - Google Patents
A kind of self-calibrating method of image measurer Download PDFInfo
- Publication number
- CN103234454B CN103234454B CN201310143160.7A CN201310143160A CN103234454B CN 103234454 B CN103234454 B CN 103234454B CN 201310143160 A CN201310143160 A CN 201310143160A CN 103234454 B CN103234454 B CN 103234454B
- Authority
- CN
- China
- Prior art keywords
- image
- subregion
- measured workpiece
- translation
- self
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
The invention discloses a kind of self-calibrating method for image measurer, it is based on mathematical image relative theory, with the two width adjacent images of measured workpiece in two dimensional surface before and after translation for object, calculate pixel displacement value, combine the reading value of grating scale when gathering image again, try to achieve actual physical size displacement value, finally try to achieve the scale factor of image measurer according to both ratio, realize the self-calibration of image measurer.The present invention, without the need to the high-precision calibrating plate by custom-made, has succinct, the efficient advantage of operation, can demarcate image measurer at any time rapidly, and save cost.
Description
Technical field
The present invention relates to image measurer scaling method field, be specially a kind of self-calibrating method of image measurer.
Background technology
Radiographic measurement technology is the class needing to carry out quantitative measurment in vision-based detection, primary study be the geometric measurement of object.Along with the progress of science and technology, the most typical product of this technology, namely utilizes noncontact image gauge head to carry out the image measurer measured, gradually accept by numerous industry.Through the development of decades, the range of application of image measurer constantly expands, and has possessed the ability of the workpiece profile of complexity and surface configuration being carried out to precision measurement.
For image measurer, the most crucial task in the preliminary work before measurement is demarcated exactly, and it is the bridge setting up corresponding relation between camera review location of pixels and scene point location.Demarcate image measurer and want simple compared to general nonlinear model camera calibration, but due to its towards be precision measurement, therefore higher to the accuracy requirement of demarcating.
At present, the conventional method demarcating image measurer is standard component method, namely the accurate dimension of standard component is passed to digital picture, exactly by measuring scaling board, physical size on calculating scaling board and the ratio of measured image pixel number, thus ask for scale factor α.Wherein, scaling board, also known as target, usually adopts the materials such as optical glass, pottery and metal to make substrate, then portrays various types of target pattern thereon.
Although it is comparatively ripe that standard component method has developed, and have higher precision, when the enlargement ratio of microlens needs frequent changes, the party's rule has highlighted operating process shortcoming loaded down with trivial details, consuming time.In addition, because the visual field of image measurer is less, the scaling board of excellent performance is made to be difficult to make, and expensive, and in use, these high-precision scaling boards are very easily polluted and are worn and torn, thus the precision of demarcating is impacted, finally affect the measurement result of measuring instrument.
Summary of the invention
For prior art Problems existing, the object of this invention is to provide a kind of self-calibrating method of image measurer, the method without the need to scaling board, make operator can more succinctly, the scale factor α of more efficiently setting of image measuring instrument, and saved cost.
In order to achieve the above object, the technical solution adopted in the present invention is:
A kind of self-calibrating method of image measurer, it is characterized in that: with measured workpiece in two dimensional surface before and after translation two width adjacent images for object, based on mathematical image relative theory, calculate the pixel displacement value in X-axis and Y-axis of two width adjacent images before and after measured workpiece translation according to zero-mean normalized crosscorrelation criterion; Combine the reading value of the grating scale of image measurer self configuration when gathering image before and after measured workpiece translation again, try to achieve the actual physical size displacement value of measured workpiece in X-axis and Y-axis; Finally according to the ratio of the pixel displacement value of two width adjacent images before and after measured workpiece actual physical size displacement value and measured workpiece translation, try to achieve the scale factor α of image measurer, realize the self-calibration of image measurer.
The self-calibrating method of described a kind of image measurer, is characterized in that: it is as follows in the pixel displacement value process of X-axis and Y-axis to try to achieve two width adjacent images before and after measured workpiece translation: before setting translation, measured workpiece image is as reference picture f
1(x, y), measured workpiece image adjacent after translation is target image f
2(x, y), first at reference picture f
1choose in (x, y) with reference to subregion A
0for template, after measured workpiece translation, obtain target image f
2(x, y); Then with zero-mean normalized crosscorrelation criterion for criterion, computing reference image f
1(x, y) and target image f
2the similarity measure C of (x, y) each position, according to the value of similarity measure C at target image f
2in (x, y), search is oriented and reference subregion A
0corresponding target subregion A
1position; Last with reference picture f
1the reference subregion A chosen in (x, y)
0with target image f
2target subregion A in (x, y)
1for object, calculate the X of two width adjacent images before and after measured workpiece translation and Integer Pel shift value U and V of Y direction based on mathematical image related operation, then adopt the Displacement algorithm based on gradient, ask for sub-pixel displacement value Δ x and Δ y.
The self-calibrating method of described a kind of image measurer, is characterized in that: choose with reference to subregion A
0time, the measured surface selecting measured workpiece has the position of obvious characteristic for reference subregion A
0, usually choose defective locations; Or by having top article, top for imaging object with article, then forms a bright spot to defocused on image, in this, as described with reference to subregion A
0.
The self-calibrating method of described a kind of image measurer, is characterized in that: it is as follows in the actual physical size displacement value process of X-axis and Y-axis to try to achieve measured workpiece: according to reference to subregion A
0at reference picture f
1the position P of (x, y)
1grating scale reading, and target subregion A
1at target image f
2the position P of (x, y)
2grating scale reading, calculate the displacement value L of measured workpiece in X and Y direction actual physical size
xand L
y, and according to vectorial P
1p
2direction, determine with reference to subregion A
0corresponding target subregion A
1target image f will be positioned at
2which quadrant of (x, y).
The self-calibrating method of described a kind of image measurer, is characterized in that: the computing formula of the scale factor α of image measurer is:
Compared with the prior art, beneficial effect of the present invention is embodied in:
1, relative to traditional standard component method, the present invention, without the need to the high-precision calibrating plate by custom-made, can realize the self-calibration of image measurer.
2, the present invention has succinct, the efficient advantage of operation, and can demarcate image measurer at any time rapidly, cost is also lower.
Accompanying drawing explanation
Fig. 1 is principle of work schematic diagram of the present invention.
Embodiment
Below in conjunction with accompanying drawing, by embodiment, this patent image measurer self-calibrating method is described in further detail.
As shown in Figure 1.During measurement, workpiece is positioned over the glass table top of image measurer, and measuring instrument will realize display and capturing sample image by its micro imaging system.The present embodiment is based on mathematical image relative theory, with the two width adjacent images of measured workpiece in two dimensional surface before and after translation for object, calculate pixel displacement value, combine the reading value of grating scale when gathering image again, try to achieve actual physical size displacement value, the ratio of final both bases tries to achieve the scale factor α of image measurer, realizes the self-calibration of image measurer.
With the sample image f gathered before two-dimensional working stage translation in the present embodiment
1(x, y) is reference picture, the sample image f gathered after translation
2(x, y) is target image.Choosing with reference to subregion A
0time, it is area-of-interest that the present invention will select the position that workpiece measured surface has an obvious characteristic, usually chooses defect (cut, scratch and greasy dirt etc.) position; Or by class article such as drawing pin, iron nail, top for imaging object with them, then forms a bright spot to defocused on image, and in this, as with reference to subregion A
0, effectively can reduce target subregion A like this
1the interference of ambient background image log word image correlation computations, thus the erroneous judgement reducing target subregion, and improve computational accuracy.
When driving two-dimensional working stage translation, target subregion A be guaranteed
1intactly be positioned at target image f
2in (x, y), and control it as far as possible and be positioned at target image f
2the near central regions of (x, y), thus the impact that reduction lens distortion as much as possible causes result of calculation.
In the scale factor α based on digital picture relative theory calculates, with reference to subregion A
0with target subregion A
1between only there is rigid body translation, and only have translation transformation, namely position changes, and direction and shape all remain unchanged, therefore in subregion, the displacement of each point is identical, therefore, can represent with zeroth order shape function:
When choosing correlation criterion, comparatively speaking, zero-mean normalized crosscorrelation (ZNCC) and zero-mean normalization squared difference have better noise immunity with (ZNSSD) correlation criterion, and to the insensitive characteristic of illumination variation, therefore the present embodiment chooses ZNCC correlation criterion, namely
And as criterion, at target image f
2searching target subregion A in (x, y)
1, thus calculate Integer Pel shift value U and V of two width adjacent images before and after workpiece translational motion.
In actual measurement, shift value general little be just Integer Pel, and for precision measurement, the displacement location precision of Pixel-level is far from being enough, therefore need adopt the displacement location of sub-pixel, thus improves the precision of locating.Displacement localization method mainly contains Surface Fitting, Newton-Raphson process of iteration and gradient method.Comparatively speaking, Surface Fitting precision is not as good as latter two, Newton-Raphson process of iteration is most widely used, but need to carry out interative computation, counting yield is not high, therefore the present embodiment chooses the Displacement algorithm based on gradient that can directly carry out solving, obtain sub-pixel displacement value Δ x and Δ y.
In the present embodiment, the computing formula of the scale factor α of image measurer is:
Concrete implementation step based on the image measurer self-calibration of digital picture relative theory is:
(1) perform the focus operation of image measurer, selected characteristic region is as reference subregion A
0, and moved to the near central regions of image, gather this position P
1sample image, be reference picture f
1(x, y), and preserve with reference to subregion A
0information, record P
1the grating scale reading of position.
(2) translation two-dimentional work bench among a small circle, on the one hand, guarantee with reference to subregion A
0do not shift out the visual field of microlens; On the other hand, make with reference to subregion A as far as possible
0still be positioned at the picture centre areas adjacent after translation, and gather this position P
2sample image, be target image f
2(x, y), and record P
2the grating scale reading of position.
(3) according to described P
1and P
2the grating scale reading of position, calculates the displacement value L of X and Y direction actual physical size
xand L
y, and according to vectorial P
1p
2direction, determine with reference to subregion A
0target image f will be positioned at
2which quadrant of (x, y), thus the scope reducing search calculating, reduce calculated amount, without the need to calculating entire image again.
(4) with two width adjacent image f
1(x, y) and f
2(x, y) is object, in conjunction with reference subregion A
0information, carry out digital picture related operation, namely with reference to subregion A
0for template, along target image f
2(x, y) movable platen, and the similarity measure C calculating each position, what the present embodiment was selected is ZNCC correlation criterion, therefore can search for according to the value of C and orient target subregion A
1position, thus try to achieve target subregion A
1relative to reference subregion A
0at Integer Pel shift value U and V of X and Y direction.
(5) again according to the Displacement algorithm based on gradient, the sub-pixel displacement value Δ x corresponding to Integer Pel shift value U and V and Δ y is calculated.
(6) final, by above-mentioned each value L
x, L
y, U, V, Δ x and Δ y bring the computing formula of scale factor α into, try to achieve α, realize the self-calibration of image measurer.
Claims (4)
1. the self-calibrating method of an image measurer, it is characterized in that: with measured workpiece in two dimensional surface before and after translation two width adjacent images for object, based on mathematical image relative theory, calculate the pixel displacement value in X-axis and Y-axis of two width adjacent images before and after measured workpiece translation according to zero-mean normalized crosscorrelation criterion; Combine the reading value of the grating scale of image measurer self configuration when gathering image before and after measured workpiece translation again, try to achieve the actual physical size displacement value of measured workpiece in X-axis and Y-axis; Finally according to the ratio of the pixel displacement value of two width adjacent images before and after measured workpiece actual physical size displacement value and measured workpiece translation, try to achieve the scale factor α of image measurer, realize the self-calibration of image measurer;
Try to achieve two width adjacent images before and after measured workpiece translation as follows in the pixel displacement value process of X-axis and Y-axis: before setting translation, measured workpiece image is as reference picture f
1(x, y), measured workpiece image adjacent after translation is target image f
2(x, y), first at reference picture f
1choose in (x, y) with reference to subregion A
0for template, after measured workpiece translation, obtain target image f
2(x, y); Then with zero-mean normalized crosscorrelation criterion for criterion, computing reference image f
1(x, y) and target image f
2the similarity measure C of (x, y) each position, according to the value of similarity measure C at target image f
2in (x, y), search is oriented and reference subregion A
0corresponding target subregion A
1position; Last with reference picture f
1the reference subregion A chosen in (x, y)
0with target image f
2target subregion A in (x, y)
1for object, calculate the X of two width adjacent images before and after measured workpiece translation and Integer Pel shift value U and V of Y direction based on mathematical image related operation, then adopt the Displacement algorithm based on gradient, ask for sub-pixel displacement value Δ x and Δ y.
2. the self-calibrating method of a kind of image measurer according to claim 1, is characterized in that: choose with reference to subregion A
0time, the measured surface selecting measured workpiece has the position of obvious characteristic for reference subregion A
0, usually choose defective locations; Or by having top article, top for imaging object with article, then forms a bright spot to defocused on image, in this, as described with reference to subregion A
0.
3. the self-calibrating method of a kind of image measurer according to claim 1, is characterized in that: it is as follows in the actual physical size displacement value process of X-axis and Y-axis to try to achieve measured workpiece: according to reference to subregion A
0at reference picture f
1the position P of (x, y)
1grating scale reading, and target subregion A
1at target image f
2the position P of (x, y)
2grating scale reading, calculate the displacement value L of measured workpiece in X and Y direction actual physical size
xand L
y, and according to vectorial P
1p
2direction, determine with reference to subregion A
0corresponding target subregion A
1target image f will be positioned at
2which quadrant of (x, y).
4. the self-calibrating method of a kind of image measurer according to claim 3, is characterized in that: the computing formula of the scale factor α of image measurer is:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310143160.7A CN103234454B (en) | 2013-04-23 | 2013-04-23 | A kind of self-calibrating method of image measurer |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310143160.7A CN103234454B (en) | 2013-04-23 | 2013-04-23 | A kind of self-calibrating method of image measurer |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103234454A CN103234454A (en) | 2013-08-07 |
CN103234454B true CN103234454B (en) | 2016-03-30 |
Family
ID=48882506
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310143160.7A Active CN103234454B (en) | 2013-04-23 | 2013-04-23 | A kind of self-calibrating method of image measurer |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103234454B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104034259B (en) * | 2014-05-21 | 2016-11-02 | 同济大学 | A kind of image measurer bearing calibration |
CN104867160B (en) * | 2015-06-17 | 2017-11-07 | 合肥工业大学 | A kind of directionality demarcation target demarcated for camera interior and exterior parameter |
JP6785092B2 (en) | 2016-08-19 | 2020-11-18 | 株式会社Screenホールディングス | Displacement detection device, displacement detection method and substrate processing device |
CN110858403B (en) * | 2018-08-22 | 2022-09-27 | 杭州萤石软件有限公司 | Method for determining scale factor in monocular vision reconstruction and mobile robot |
CN109015646B (en) * | 2018-08-22 | 2021-07-23 | 中科新松有限公司 | Position information self-calibration method, device, equipment and storage medium |
CN110285770B (en) * | 2019-07-31 | 2020-08-07 | 中山大学 | Bridge deflection change measuring method, device and equipment |
CN110264490B (en) * | 2019-08-15 | 2019-12-10 | 成都新西旺自动化科技有限公司 | sub-pixel precision edge extraction method applied to machine vision system |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102006055758A1 (en) * | 2006-11-25 | 2008-05-29 | Deutsches Zentrum für Luft- und Raumfahrt e.V. | Camera calibrating method for optical measuring system, involves marking one of cross-over points in checker pattern by position information, which is readable by image processing of images of pattern and has coding of orientation |
CN101382418A (en) * | 2008-10-16 | 2009-03-11 | 西安工业大学 | Absolute displacement measuring set and method based on digital picture |
CN101813693A (en) * | 2010-05-06 | 2010-08-25 | 北京大学 | Cell in-situ active deformation measurement method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4189925B2 (en) * | 2003-09-05 | 2008-12-03 | 福岡県 | Surface displacement measuring method and surface displacement measuring apparatus |
-
2013
- 2013-04-23 CN CN201310143160.7A patent/CN103234454B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102006055758A1 (en) * | 2006-11-25 | 2008-05-29 | Deutsches Zentrum für Luft- und Raumfahrt e.V. | Camera calibrating method for optical measuring system, involves marking one of cross-over points in checker pattern by position information, which is readable by image processing of images of pattern and has coding of orientation |
CN101382418A (en) * | 2008-10-16 | 2009-03-11 | 西安工业大学 | Absolute displacement measuring set and method based on digital picture |
CN101813693A (en) * | 2010-05-06 | 2010-08-25 | 北京大学 | Cell in-situ active deformation measurement method |
Non-Patent Citations (2)
Title |
---|
三维影像仪中影像测量关键技术研究;付琰;《中国优秀硕士学位论文全文数据库 信息科技辑》;20130315(第3期);正文第18-19页第3.3节及图3-4 * |
数字图像相关中的亚像素位移定位算法进展;潘兵等;《力学进展》;20050825;第35卷(第3期);第345-346页第2部分第1-2段,第349-350页第3部分第3.5节及图2 * |
Also Published As
Publication number | Publication date |
---|---|
CN103234454A (en) | 2013-08-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103234454B (en) | A kind of self-calibrating method of image measurer | |
CN103292701A (en) | Machine-vision-based online dimensional measurement method of precise instrument | |
US20100310128A1 (en) | System and Method for Remote Measurement of Displacement and Strain Fields | |
CN103913293B (en) | Coded structured light system optics TRANSFER MODEL measuring system | |
Liu et al. | An improved online dimensional measurement method of large hot cylindrical forging | |
CN104359403A (en) | Plane part size measurement method based on sub-pixel edge algorithm | |
CN104657711B (en) | A kind of readings of pointer type meters automatic identifying method of robust | |
CN102376089A (en) | Target correction method and system | |
Xia et al. | An accurate and robust method for the measurement of circular holes based on binocular vision | |
CN105180806A (en) | Trans-scale geometrical parameter measurement method based on microscopic visual sense | |
CN107589069B (en) | Non-contact type measuring method for object collision recovery coefficient | |
JP2012058076A (en) | Three-dimensional measurement device and three-dimensional measurement method | |
CN101515331B (en) | Neighborhood correlated sub-pixel positioning method | |
CN109751964B (en) | High-precision non-contact pipe diameter measuring method and device | |
CN102853786A (en) | Apparatus and method for detecting flatness | |
CN104680526A (en) | Method for measuring uniform particle size based on auto-correlation algorithm | |
CN110223355B (en) | Feature mark point matching method based on dual epipolar constraint | |
CN108564621B (en) | Structured light strip center extraction method and device for track detection | |
CN104807407A (en) | Mobile phone length measuring method based on bank card | |
Zhao et al. | Displacement monitoring for slope stability evaluation based on binocular vision systems | |
Liang et al. | Displacement measurement of specimen surfaces with damaged areas by digital image correlation | |
Antoš et al. | Real-time optical measurement of displacements using subpixel image registration | |
CN104754323A (en) | Calibration method of camera optical axis detection apparatus | |
CN106441162A (en) | Cylindrical object verticality noncontact detection device and method | |
CN106407987B (en) | Electro-optic crystal optical axis exposure point extraction method based on image cross-correlation matching |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |