CN109900226A - A kind of metal surface quality detecting method of machine vision - Google Patents
A kind of metal surface quality detecting method of machine vision Download PDFInfo
- Publication number
- CN109900226A CN109900226A CN201711299633.7A CN201711299633A CN109900226A CN 109900226 A CN109900226 A CN 109900226A CN 201711299633 A CN201711299633 A CN 201711299633A CN 109900226 A CN109900226 A CN 109900226A
- Authority
- CN
- China
- Prior art keywords
- image
- dimensional
- machine vision
- quality detecting
- metal surface
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 17
- 239000002184 metal Substances 0.000 title claims abstract description 9
- 238000010586 diagram Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 230000003746 surface roughness Effects 0.000 description 4
- 230000009466 transformation Effects 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 239000000700 radioactive tracer Substances 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000005305 interferometry Methods 0.000 description 1
- 239000000155 melt Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000004439 roughness measurement Methods 0.000 description 1
- 238000000790 scattering method Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000033772 system development Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a kind of metal surface quality detecting methods of machine vision, this method comprises: S1, obtains micro- sequence image and pre-processed；S2 merges multilayer sequence image；S3 restores workpiece surface three-dimensional appearance and calculates three-dimensional parameter value.
Description
Technical field
The invention belongs to Machine Vision Recognition Technology fields, are related to a kind of metal surface quality detecting method of machine vision.
Background technique
Currently, tracer method and optical method are the most traditional in numerous detection methods, but since tracer method needle sizes have
The influence for the disadvantages of limit and syringe needle are easily damaged measured surface, use scope are limited；Optics rule mostly need to by means of special instrument,
Expensive, testing cost is excessively high, and also has certain limitation for roughness measurement, such as by the limit of optical wavelength
System.
In roughness assessment system, the extremely limited information of workpiece surface is only utilized in two-dimensional parameter, cannot effective earth's surface
Levy the performances such as supporting, wearability, the leakproofness of workpiece.
Three-dimensional evaluation is from workpiece surface region rather than profile traces obtain information, can recover the three-dimensional shaped of workpiece surface
Looks, can it is more acurrate, more fully workpiece surface appearance is evaluated, there is globality and of overall importance.
Summary of the invention
It is limited using the microscope depth of field it is an object of that present invention to provide a kind of metal surface quality detecting method of machine vision
Feature carries out multilayer sequence to workpiece surface and claps figure, then recovers the exact height of workpiece surface to image co-registration by software
Information, reconstructs the three-dimensional appearance on its surface, finally calculates the three-dimensional parameter of workpiece surface roughness, realizes to common work
Part three-dimensional surface shape is reasonably evaluated, and traditional contact pin type detection method damage measured surface, tradition are efficiently solved
Focusing, interferometry, the measuring speed of scattering method be slow, the inaccurate comprehensive problem of information.
In order to solve the above technical problems, the present invention adopts the following technical scheme that: a kind of metal surface matter of machine vision
Detecting method, which comprises S1 obtains micro- sequence image and pre-processed；S2 melts multilayer sequence image
It closes；S3 restores workpiece surface three-dimensional appearance and calculates three-dimensional parameter value.
The present invention have compared with prior art it is below the utility model has the advantages that
The present invention program feature limited using the microscope depth of field, carries out multilayer sequence to workpiece surface and claps figure, then pass through
Software recovers the exact height information of workpiece surface, reconstructs the three-dimensional appearance on its surface to image co-registration, finally calculates and goes to work
The three-dimensional parameter of part surface roughness is realized and is rationally and effectively evaluated common workpiece surface three-dimensional appearance.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of the metal surface quality detecting method of machine vision.
Fig. 2 is image preprocessing flow diagram.
Fig. 3 is that altitude information obtains schematic diagram.
Fig. 4 is Gauss interpolation schematic diagram.
Specific embodiment
With reference to the accompanying drawing and specific embodiment to the present invention carry out in further detail with complete explanation.It is understood that
It is that described herein the specific embodiments are only for explaining the present invention, rather than limitation of the invention.
Referring to Fig.1, the metal surface quality detecting method of a kind of machine vision of the invention, which comprises
S1 obtains micro- sequence image and is pre-processed；
S11 clap layer by layer to workpiece surface and be schemed using Z axis digital display microscope, and the full sequence image for taking bat all wraps
All information of the Z-direction in microscope containing workpiece；
S12 pre-processes image to reduce the influence of noise on image；
Because collected original image inevitably will receive noise, distortion, low contrast etc. influence, by picture into
Row pretreatment can reduce the adverse effect of these factors to achieve the purpose that improve discrimination.
Referring to Fig. 2, pretreated basic step includes:
1) gray level image is converted by original image while eliminating noise, conversion formula are as follows:
F (x, y)=0.299R+0.578G+0.114B (1)
Wherein R, G, B are three color components, carry out burn into expansion and filtering to the image after gray processing and are eliminated with reaching
The purpose of noise, smoothed profile.
2) image enhancement, using piecewise linear transform between image degree of comparing stretching conversion to increase concave and convex plane
Contrast.The principle of this mode is to improve the grayscale dynamic range of image procossing.Its greyscale transformation function are as follows:
Wherein: (a1,a2) be original image f (x, y) tonal range；(b1,b2) be transformation after image f'(x, y) gray scale
Range；L-1 is the gray value upper limit.
3) border detection and slant correction, border detection are used using top cap transformation and canny operator, slant correction
Hough transform.Top cap shift theory: such as by morphological method: expansion, burn into opening operation, closed operation, from original image
Opening operation image is subtracted, the part left in image is exactly image border.Calculating formula:
H=F- (F ° of S) (3)
F is generally original image；S is morphology element.
S2 merges multilayer sequence image；
The key of image co-registration is how accurately to obtain its altitude information, and schematic diagram is as shown in Figure 3.It can from Fig. 3
Out, workpiece surface is rough, there is the part of focus in every sequence of layer image.
Microscope, which rises to clap from the minimum point of workpiece surface to highest point stepping, takes sequence image, each on workpiece surface
It is focus that a point, which always has the point in a corresponding sequence figure, and the present invention is able to reflect using variance Focus field emission array operator
The readability that image focuses, when a certain point focusing in image, Focus field emission array operator value is the largest.Therefore, sequence is found out
The maximum value of each pixel Focus field emission array operator in column image also just has found corresponding number of plies when its focusing, so as to
Roughly calculate every height value.Focus field emission array operator value focuses near top at it at Gaussian Profile, such as Fig. 4 institute
Show.
The variance Focus field emission array operator magnitude of any point (x, y) is { F on curved surfacei| i=1,2 ..., m },Indicate this point
Exact height value.According to Gaussian Profile, variance focus function is expressed as
In formulaThe average value of Gaussian Profile is represented, also the actual height value of this point；σFRepresent standard deviation.
Both sides take logarithm to obtain
By three measure value Fm、Fm-1、Fm+1And its corresponding dm、dm-1、dm+1Substituting into above formula with stepping Δ d can calculate
S3 restores workpiece surface three-dimensional appearance and calculates three-dimensional parameter value；
S31, the selection of three-dimensional parameter；
In new ISO4287 draft, it is proposed that 27 standardized range parameters, 3 spacing parameters and 12 synthesis
Parameter.In fact, the microscopic appearance of piece surface be it is complicated, each parameter can only characterize the one aspect of pattern.So being
Avoiding the parameter explosion phenomenon similar to two-dimensional parameter should lay particular emphasis on a small number of, basic, important when choosing three-dimensional parameter
Selection principle, this is also a direction of three-dimensional parameter evaluating system development.
The present invention combines actual needs and parameter selection principle, chooses the S in range parametera、Sq、Sz、SskAnd Sku, also
S in spatial parameterds.These parameter basic expressions three-dimensional surface roughness feature, and convenience of calculation.
Before solving three-dimensional parameter, need a datum level as evaluation benchmark, the three-dimensional evaluation that the present invention is utilized
Benchmark is promoted by two-dimensional silhouette middle line, that is, uses least square plane as the evaluation benchmark of three-dimensional parameter.If three-dimensional surface
For Z (x, y), least square datum plane Equation f (x, y)=a+bx+cy finds out the value of a, b, c according to the principle of least square, then
F (x, y)=a+bx+cy is substituted into, the least square plane equation for some measurement point can be found out.Surface roughness curved surface side
Journey l (x, y, z)=z (x, y)-f (x, y).
S32, three-dimensional parameter introduction；
1) surface arithmetic average deviation SaRefer on measured surface each point to reference datum distance average value:
S is orthographic projection of the sampling area on datum level in formula.
2) surface Root Mean Square deviation SqRefer to the root-mean-square value of each point deviation reference datum on measured surface, it reacts wheel
Exterior feature deviates the degree of datum level:
3) ten point height S of surfacez:
L in formulai(x,y,z)maxAnd lj(x,y,z)min(i=1,2 ..., 5) it is respectively 5 tops and 5 in domain for assessment
The amount of a lowest trough.
4) surface deflection degree SskIt is the shape for measuring profile bumps amplitude distribution curve, with two-dimensional silhouette degree of skewness
RskFunction is identical:
5) surface deflection degree SkuIt is the kurtosis of topographical height distribution and the measurement of steepness:
6) surface peak density SdsIndicate the number on the vertex for including in unit sampling area:
M and N refers to the number of the sampled point in the direction x and y in sampling area in formula；NsRefer to the number of unit sampling area inner vertex
Mesh；Δ x, Δ y refer specifically to sampled point spacing.
The above description is only a preferred embodiment of the present invention, is not intended to restrict the invention, for those skilled in the art
For, the invention can have various changes and changes.All any modifications made within the spirit and principles of the present invention are equal
Replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (1)
1. a kind of metal surface quality detecting method of machine vision, which is characterized in that the described method includes: S1, obtains micro- sequence
Image is simultaneously pre-processed；S2 merges multilayer sequence image；S3 restores workpiece surface three-dimensional appearance and calculates three-dimensional
Parameter value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711299633.7A CN109900226A (en) | 2017-12-10 | 2017-12-10 | A kind of metal surface quality detecting method of machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711299633.7A CN109900226A (en) | 2017-12-10 | 2017-12-10 | A kind of metal surface quality detecting method of machine vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109900226A true CN109900226A (en) | 2019-06-18 |
Family
ID=66940964
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711299633.7A Pending CN109900226A (en) | 2017-12-10 | 2017-12-10 | A kind of metal surface quality detecting method of machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109900226A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112747698A (en) * | 2019-10-29 | 2021-05-04 | 复盛应用科技股份有限公司 | Golf club head measurement method |
-
2017
- 2017-12-10 CN CN201711299633.7A patent/CN109900226A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112747698A (en) * | 2019-10-29 | 2021-05-04 | 复盛应用科技股份有限公司 | Golf club head measurement method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102032875B (en) | Cable sheath thickness measuring method based on image processing | |
| CN105931232B (en) | High-precision sub-pixel extraction method for the center of structured light strips | |
| CN108362469B (en) | Size and surface pressure measurement method and apparatus based on pressure sensitive paint and light-field camera | |
| CN110766748B (en) | A non-contact vehicle attitude measurement method | |
| CN104657587A (en) | Method for extracting center line of laser stripe | |
| CN104616325A (en) | Rapid and high-precision method for extracting light strip center on large surface | |
| CN104764407B (en) | A kind of fine measuring method of thickness of cable sheath | |
| CN112132886A (en) | Circular hole center rapid positioning and roundness detection method for aviation part | |
| CN102592284A (en) | Method for transforming part surface appearance three-dimensional high-density point cloud data into grayscale image | |
| CN107798698A (en) | Structured light strip center extracting method based on gray-level correction and adaptive threshold | |
| CN105387814A (en) | Automatic measurement system for surface parameters of prefabricated component of three-dimension braiding composite material | |
| CN110793462A (en) | Measuring method of index circle of nylon gear based on vision technology | |
| CN116433707B (en) | Accurate extraction method and system for optical center sub-pixels of line structure under complex background | |
| CN107358628A (en) | Linear array images processing method based on target | |
| CN110287887A (en) | A kind of scale recognition methods, device and computer storage medium | |
| CN112361989B (en) | Method for calibrating parameters of measurement system through point cloud uniformity consideration | |
| CN111415349A (en) | Method for detecting polyester filament yarn based on image processing technology | |
| CN112241964B (en) | Light strip center extraction method for line structured light non-contact measurement | |
| CN114580559A (en) | Speed measuring method based on monocular vision system | |
| CN114820812B (en) | Three-dimensional reconstruction method for high-reflection surface based on structured light | |
| CN111415378B (en) | Image registration method for automobile glass detection and automobile glass detection method | |
| CN110081832A (en) | Logic projection imaging measures the method and device of nontransparent object structures parameter | |
| CN112330667A (en) | Morphology-based laser stripe center line extraction method | |
| CN109900226A (en) | A kind of metal surface quality detecting method of machine vision | |
| CN111815580A (en) | Image edge identification method and small module gear module detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190618 |