CN106949848A - A kind of high-precision laser 3D profiles phone structural detection method - Google Patents
A kind of high-precision laser 3D profiles phone structural detection method Download PDFInfo
- Publication number
- CN106949848A CN106949848A CN201710153827.XA CN201710153827A CN106949848A CN 106949848 A CN106949848 A CN 106949848A CN 201710153827 A CN201710153827 A CN 201710153827A CN 106949848 A CN106949848 A CN 106949848A
- Authority
- CN
- China
- Prior art keywords
- data
- point
- laser
- model
- measurement
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
Abstract
The invention discloses a kind of high-precision laser 3D profiles phone structural detection method, including：Step 1) the previously-scanned sample once of laser measuring apparatus is used, the sweep parameter of sample, including laser power are set, image exposuring time, acquisition range gathers the laser scanning image sample of sample；Step 2) phone structural is placed to mobile work platform, fixed laser measurement head is in surface, and servomotor drives the movement of workbench supporting plate, detection is completed, system completed surface profile detection, sampling number St=pl/sstep in 2 seconds, pl is product measurement length, and sstep is sampling step length；Step 3) set up profile standard data model；Step 4) the preceding importing 3D MODEL C AD data of measurement, data are made up of thousands of to up to ten thousand tri patch, along slice map of path interception 3D model vertical with scanning direction etc..
Description
Technical field
The invention belongs to a kind of high-precision laser 3D profiles phone structural detection method.
Background technology
With the development of mobile terminal (mobile phone), the quality requirements more and more higher of phone structural.Mobile phone knot is detected in the past
The most of producers of component are used based on visually observation, and subjectivity is very big, and quality is difficult to stabilization.Small part enterprise uses machine in industry
The method detection phone structural of device vision, such as Huawei Tech Co., Ltd, Samsung, also deploys some test dresses
It is standby in handset production line, the main flow of these test equipments and detecting system still uses 2D vision techniques, uses different angles
Light source by profile shoot come, then analyzed and processed using image processing algorithm.The method data of this 2D visions are disturbed
Dynamic larger, often intensity of illumination, material change, the small change of light-source angle can all have a huge impact to data, detect
As a result it is inaccurate.Some AOI producers are filled using method and three-color LED multi-angle light source the manufacture image detection of statistical machine learning
It is standby, it can realize that the presence or absence of defect is detected to a certain extent, it is impossible to the defect in terms of accurate quantification size.
The method that 1 routine 2D machine vision is taken pictures using front or side, characteristics of image is easily influenceed by illumination, is produced
Result is surveyed by mistake；
2nd, the method for another processing Surface testing is to use intelligent image algorithm, for example, carried out with machine learning method
Characteristics of image is classified, and classification, but the not energetic size of this method are realized by great amount of samples training, while when needing longer
Between be trained and learn；
2nd, current laser profile degree instrument can be with output surface point cloud, but the narrower width of measurement, and precision and efficiency are paid no attention to
Think, the measurement simultaneously for mixed materials is unreliable.
3rd, detected using depth data, be related to 3D point cloud registration problems, current main flow 3D point cloud registration Algorithm
It is the iteration nearest neighbor algorithm (ICP) for adapting to free form surface registration, it is low for 3D vision-based detection efficiency, it is impossible to meet on-line testing
Demand.
The content of the invention
The technical problems to be solved by the invention are to provide a kind of high-precision laser 3D profiles phone structural detection method,
For solving the problem of prior art is present.
The technical scheme that present invention solution above-mentioned technical problem is taken is as follows：
A kind of high-precision laser 3D profiles phone structural detection method, is specifically included：
Step 1) the previously-scanned sample once of laser measuring apparatus is used, the sweep parameter of sample, including laser power are set,
Image exposuring time, acquisition range gathers the laser scanning image sample of sample；
Step 2) phone structural is placed to mobile work platform, fixed laser measurement head is in surface, servomotor band
Dynamic workbench supporting plate movement, completes detection, and system completed surface profile detection, sampling number St=pl/sstep, pl in 2 seconds
It is product measurement length, sstep is sampling step length；
Step 3) set up profile standard data model；
Step 4) import 3D MODEL C AD data before measurement, data are made up of thousands of to up to ten thousand tri patch, along with sweep
The slice map that the vertical path in direction intercepts 3D models is retouched, in each spatial triangle and slice plane shape of structural member 3D data
Into intersecting lens, calculate the intersection point of each edge of intersecting lens and spatial triangle, judge intersection point whether triangle end points composition
Line segment between, if record intersection point coordinate is used as Slice Sampling between 2 line ends；
After the completion of Slice Sampling point record, Slice Sampling point is ranked up, profiled outline data are formed, wherein, in wheel
Key point is detected in wide data acquisition system, tangent plane contour line is represented with the sequence of key point；
Also, each tangent plane data use (x1, pt1), (x2, pt2), (x3, pt3), (x4, pt4) ..., (xn, ptn)
Sequence key point is constituted；Ptn represents data vertex type, and flat data point and the big data point of curvature judge in detection process
Standard uses different weighted factors, including：
Represent that data vertex type is flat spot using ptn=1, ptn=2 represents that data point is flex point, works as curvature variation of data points
It is to be arranged to flex point more than some scope；
Numerous slice of datas is lined up according to reference axis, data altitude information is formed, altitude information is exactly to cut
Height value on piece curve, rectangular altitude data can be expressed as 2D gray scale pcolors, wherein, gray count formula is such as on image
It is that follow-up matching detection more facilitates by the Z axis data conversion of model into gray value thus shown in formula (2)；
Pg (x, y)=255* (z (x, y)-z1)/(Z2-Z1) (2)
Step 5) 3D point cloud matching step is actual to structural member first before detection to measure 3D point cloud and input master pattern
Registration is carried out, actual measurement structure part is calculated relative to master pattern and displacement and the anglec of rotation, wherein, Model registration algorithm
It is iteration arest neighbors ICP algorithm, wherein, the judgment rule of ICP algorithm is：
After product measurement height map generation, height map is converted into gray scale pcolor；
Step 6) actual product height map and reference model height map generation after, both are subjected to registration process, it is registering
The method that method uses 2D locations of contours；
Step 7) after product registration operation, the range difference that model and actual product measure altitude information is calculated, reality
The data and model data of border phone structural measurement are compared, and least square method and Euclidean distance are used for comparative approach
Computational methods, according to different data vertex types use different matching degree computational methods；
Step 8) measurement process feedback control and data quality monitoring；
Wherein, including：Laser image collection and laser controlling are merged, calibration sample figure first before systematic survey
The statistic of picture, the brightness of image average of record standard sample, laser linewidth, after being measured every time in measurement process, inspection
Survey whether the line width of laser image, average meet requirement, export measurement result if meeting and requiring, otherwise notify user again
Measurement or multi output result optimize output result again after processing.Ensure the measurement essence of output result in this way
Degree.
Preferably, before above method of testing, also include：
Phone structural (4) is placed on test platform,【Relation and component names are accurately placed, accompanying drawing is refer to
1】And phone structural is irradiated using the blue wavelength 405nm laser line light source (1) that sends of semiconductor line structure light laser,
Wherein, in symmetrically arranged first CMOS camera modules (2) and the 2nd CMOS camera modules (3) collection laser line image, acquisition
Surface 3D point cloud data, wherein, camera optical axis and laser rays optical plane angle have selected 45 degree, wherein, imaging publicity is as follows：
Wherein,
Preferably, step 3) specifically include：
3D MODEL C AD data are imported before measurement, are then normalized；
Or, 3D point cloud data are obtained using master sample sampling, then standard three-dimensional point are generated after resampling, filtering
Cloud.
Preferably, step 3) specifically include：
Standard form is used as using after the 3D overall sizes data filtering processing of the actual measurement of standard sample.
Preferably, step 6) specifically include：
A) in reference model figure, manually by software interface choose conspicuousness position point feature, for example circular hole, square hole,
Right-angle side etc., calculates the geometry measure feature of location feature, including：Length, height, width, area, the segmentation of outline method vector sum
Value tag, the geometric properties of profile are preserved as template parameter；
B) binary segmentation is carried out to product height map using the partition value of reference model, spot inspection is carried out using Blob algorithms
Survey, blob features and reference model figure the positioning point feature detected is compared, and finds out most like spot as product
Position feature point, comparative product position feature point and reference model figure anchor point displacement relation, so as to realize actual product and ginseng
Examine Model registration.
Preferably, step 7) specifically include：
For ptn=1 flat surfaces number of types strong point, first to referring to point set fit Plane model, least square method is used
Calculate, then calculate actual samples point to plane or the distance of plane equation, calculating, which exceeds, judges different severity level errors
The number of the point of threshold distance, position, finally define decision tree classifier output final result using this patent；
For ptn=2 turning points measurement point, the circular arc camber model of flex point, Ran Houji are gone out using least square fitting
Calculate actual surface model and the distance between the radius of the surface model dis with reference to point set, using dis values as judgement according to
According to；
Wherein, filtered out in calculating parameter model process in order to avoid interference using iterative algorithm a certain proportion of
Hash point；Wherein, gating rate is from 10% to 60%.
The present invention is detected using laser profile technology measurement phone structural depth dimensions, surveys objective
The physical quantity that structural member authority embodies is measured, detection data precision is high, and wrong report and miss probability are low, can realize 0 rate of false alarm.Changing
Detect that without carrying out machine learning training when model, system flexibility degree is high, can be achieved quickly to remodel.Testing result and light
The position of spot is relevant, and the intensity of hot spot is unrelated, when LASER Light Source brightness changes, and laser rays center, which changes, to be far smaller than
The change of laser linewidth, therefore the uniformity of testing result is good.
Other features and advantages of the present invention will be illustrated in the following description, also, partly becomes from specification
Obtain it is clear that or being understood by implementing the present invention.The purpose of the present invention and other advantages can be by the explanations write
Specifically noted structure is realized and obtained in book, claims and accompanying drawing.
Brief description of the drawings
The present invention is described in detail below in conjunction with the accompanying drawings, to cause the above-mentioned advantage of the present invention definitely.Its
In,
Fig. 1 is the schematic diagram of high-precision laser 3D profiles phone structural detection method of the present invention；
Fig. 2 is the schematic flow sheet of high-precision laser 3D profiles phone structural detection method of the present invention；
Fig. 3 is the schematic diagram of high-precision laser 3D profiles phone structural detection method of the present invention.
Embodiment
Describe embodiments of the present invention in detail below with reference to drawings and Examples, how the present invention is applied whereby
Technological means solves technical problem, and reaches the implementation process of technique effect and can fully understand and implement according to this.Need explanation
As long as not constituting each embodiment in conflict, the present invention and each feature in each embodiment can be combined with each other,
The technical scheme formed is within protection scope of the present invention.
In addition, the step of the flow of accompanying drawing is illustrated can such as one group computer executable instructions department of computer science
Performed in system, and, although logical order is shown in flow charts, but in some cases, can be with different from herein
Order perform shown or described step.
Wherein, present invention survey phone structural procedure is as follows：
1st, phone structural is placed to test platform, uses blue wavelength 405nm semiconductor line structure light laser
Phone structural is irradiated, double CMOS cameras gather laser line image, obtain upper surface 3D point cloud data, in order to reach that highest is surveyed
Accuracy of measurement, camera optical axis and laser rays optical plane angle have selected 45 degree.According to formula (1), increase laser optical plane and
Angle between camera optical axis can improve measurement accuracy, but be limited to physical dimension, therefore for handset structure measurement side
45 degree of system architecture angle is used in method.Imaging arrangement is as shown in Figure 1.
Wherein,
Fig. 1 is described as follows:1st, laser line light source；2nd, CMOS camera modules A；3rd, CMOS camera modules B；4th, test system knot
Component；5th, test sample.
2nd, using the previously-scanned sample once of laser measuring apparatus, the sweep parameter of sample, including laser power, image are set
Time for exposure, acquisition range.Gather the laser scanning image sample of sample.
3rd, phone structural normally detects that running is as follows, places phone structural to mobile work platform, fixed to swash
Photo measure head is in surface, and servomotor drives the movement of workbench supporting plate, completes detection, and system completed surface profile in 2 seconds
Detection.Sampling number St=pl/sstep, pl are product measurement length, and sstep is sampling step length.System can be according to product not
With size and complex degree of structure, sampling step length is set.
4th, profile, which compares, needs to set up normal data model, and one of master pattern method for building up is importing 3D models before measurement
CAD data, is then normalized；Another mode is to obtain 3D point cloud data using master sample sampling, is then adopted again
Standard three-dimensional point cloud is generated after sample, filtering.The two of master pattern method for building up are the 3D profiles using the actual measurement of standard sample
Standard form is used as after sized data filtering process.
5th, the form stl forms that 3D MODEL C AD data use main flow 3D CAD design software supports, data are imported before measuring
It is made up of thousands of to up to ten thousand tri patch, in order to realize that profile is quickly compared and measured, patent of the present invention uses computational geometry
Method, the slice map of 3D models is intercepted along the path vertical with scanning direction, in each space three of structural member 3D data
Angular and slice plane formation intersecting lens, calculates the intersection point of each edge of intersecting lens and spatial triangle, judge intersection point whether
Between the line segment that the end points of triangle is constituted, if record intersection point coordinate is used as Slice Sampling point between 2 line ends.Section is adopted
After the completion of sampling point record, Slice Sampling point is ranked up, profiled outline data are formed.Directly use profiled outline DATA REASONING
More erroneous judgement can be produced, and data volume is also larger, therefore detect in outline data set key point, tangent plane contour line is used
The sequence of key point is represented.For the structural member of every kind of model, multigroup tangent plane data are created according to fixed step size, in the present invention
0.025mm, 0.05,0.075mm are used, 0.1mm, five kinds of step-lengths of 0.2mm create section model data.Each tangent plane data make
With (x1, pt1), (x2, pt2), (x3, pt3), (x4, pt4) ..., (xn, ptn) sequence key point composition.Ptn represents data
Vertex type, the standard that flat data point and the big data point of curvature judge in detection process uses different weighted factors, I
Represent that data vertex type is flat spot using ptn=1, ptn=2 represents that data point is flex point, when curvature variation of data points be more than certain
Individual scope is to be arranged to flex point.Numerous slice of datas is lined up according to reference axis, data altitude information, height is formed
Data are exactly height value on curve of cutting into slices, and rectangular altitude data can be expressed as 2D gray scale pcolors.Gray count on image
It is that follow-up matching detection more facilitates thus the Z axis data conversion of model into gray value shown in formula such as formula (2).
Pg (x, y)=255* (z (x, y)-z1)/(Z2-Z1) (2)
6th, 3D point cloud is matched, and measurement 3D point cloud actual to structural member first and input master pattern are matched somebody with somebody before detection
Standard, calculates actual measurement structure part relative to master pattern and displacement and the anglec of rotation.The Model registration algorithm commonly used at present
It is iteration arest neighbors ICP algorithm.The judgment rule of ICP algorithm is：
ICP algorithm is mainly used in free form surface three-dimensional splicing field, the precision and effect matched in regular texture detection field
Rate is not high.Actually phone structural point cloud can be represented in a datum plane with digital height, therefore we use height
Figure represents phone structural measurement result.Because test process slightly some may be tilted, therefore it is generated in height map
Preceding use ICP algorithm carries out angle of inclination correction, and angle searching scope is limited within +/- n degree, and n values use software design patterns, with
The reduction of hunting zone, amount of calculation is substantially reduced.
After product measurement height map generation, height map is converted into the knot that light color in gray scale pcolor, image represents protrusion
Structure feature, dark color represents and falls in architectural feature.This patent is used according to the disposal methods gray scale pcolor of gray level image
2D image detecting methods are post-processed.
7th, after actual product height map and reference model height map are generated, both are subjected to registration process, registering method
Using the method for 2D locations of contours, quickly, a few tens of milliseconds can complete location Calculation to this method speed.Registration process is as follows：
A) in reference model figure, manually by software interface choose conspicuousness position point feature, for example circular hole, square hole,
Right-angle side etc., the geometry measure feature (length, height, width, area, outline method vector) and partition value for calculating location feature is special
Levy, the geometric properties of profile are preserved as template parameter；
B) binary segmentation is carried out to product height map using the partition value of reference model, spot inspection is carried out using Blob algorithms
Survey, blob features and reference model figure the positioning point feature detected is compared, and finds out most like spot as product
Position feature point, comparative product position feature point and reference model figure anchor point displacement relation, so as to realize actual product and ginseng
Examine Model registration.
8th, it is necessary to which the range difference for measuring altitude information to model and actual product is calculated, in fact after product registration operation
The data and model data of border phone structural measurement are compared, and least square method and Euclidean distance are used for comparative approach
Computational methods.Different matching degree computational methods are used according to different data vertex types.For ptn=1 flat surfaces classes
Type data point, first to referring to point set fit Plane model, is calculated using least square method, then calculates actual samples point to plane
Or the distance of plane equation, the number for exceeding the point for judging different severity level error threshold distances, position are calculated, is finally made
Decision tree classifier output final result is defined with this patent.For ptn=2 turning points measurement point, intended using least square method
The circular arc camber model of flex point is closed out, between the radius for the surface model for then calculating actual surface model and reference point set
Apart from dis, dis values are regard as the foundation judged.In order to avoid interference in calculating parameter model process, it is necessary to use iteration
Algorithm filters out a certain proportion of hash point, and the gating rate that this patent is used is from 10% to 60%, according to different product
Type system is adjusted.
9th, the feedback control and data quality monitoring of measurement process.The stability of light source is very in structured light measurement system
It is important.Laser image collection and laser controlling are merged by this patent, the system of calibration sample image first before systematic survey
Metering, the brightness of image average of record standard sample, laser linewidth after being measured every time in measurement process, detect laser
Whether the line width of image, average meet requirement, if meet require if export measurement result, otherwise notify user remeasure or
Person's multi output result optimizes output result again after processing.Ensure the measurement accuracy of output result in this way.Laser
The control of device is completed by the one block of laser controlling plate made, and control signal uses the voltage signal of 0-3v scopes.
Detected using laser profile technology measurement phone structural depth dimensions, objective ground measurement structure
The physical quantity that part authority embodies, detection data precision is high, and wrong report and miss probability are low, can realize 0 rate of false alarm.Changing detection type
Number when without carrying out machine learning training, system flexibility degree is high, can be achieved quickly to remodel.The position of testing result and hot spot
It is equipped with pass, and the intensity of hot spot is unrelated, when LASER Light Source brightness changes, laser rays center, which changes, is far smaller than laser rays
Wide change, therefore the uniformity of testing result is good.By the 3D point cloud data conversion of theoretical model into digital elevation data, and with
2D gray level images are represented.3D data can be so handled with 2D image algorithms.
Wherein, this case needs main points to be protected, is summarized as follows：
1st, novelty, which proposes laser optical plane and binocular cmos sensor Wide-angle imaging structure, is used for fine structure part
Detection, the angle of design is 45 degree, realizes that phone structural measurement accuracy optimizes in this way.
2nd, to adapt to the measurement of unlike material phone structural, laser intensity is monitored in measurement process, according to image
Line width and hot spot scattering degree adjust automatically laser energy are detected, high stable measurement process is realized.
3rd, 3D outline datas are changed into 2D gray-scale maps or pseudocolour picture, uses 2D vision Blob location algorithms and gray scale
Comparison algorithm realizes the rapid registering of outline data, quick detection.
4th, detection template STL data processing methods, by the normalized 2D images of template 3D point cloud data conversion.
5th, actual spot of measurement cloud data are represented with matrix digital altitude information, change into the image that fathoms, and call 2D
Contour detecting algorithm is positioned, and the data and model altitude information after positioning carry out additive operation, subtract each other rear residual image and set
Put a threshold value to be split, use the area of the effective defect of Blob detection statistics segmentation figures, length, width characteristics.
6th, reference model and actual product distance compare the different methods that employ, and flat surfaces use planar linear model
It is compared, turning point is compared using arc model.
7th, devising tree classificator is used for defect dipoles, and conspicuousness defect is distinguished and substantially good first by Weak Classifier
Product, then those suspected defects binding deficient feature distribution is further discriminated between, until false drop rate and loss meet visitor
Family is required.
It should be noted that for above method embodiment, in order to be briefly described, therefore it is all expressed as a series of
Combination of actions, but those skilled in the art should know, the application is not limited by described sequence of movement because
According to the application, some steps can be carried out sequentially or simultaneously using other.Secondly, those skilled in the art should also know
Know, embodiment described in this description belongs to preferred embodiment, involved action and module not necessarily the application
It is necessary.
It should be understood by those skilled in the art that, embodiments herein can be provided as method, system or computer program
Product.Therefore, the application can be using the reality in terms of complete hardware embodiment, complete software embodiment or combination software and hardware
Apply the form of example.
Moreover, the application can use available in one or more computers for wherein including computer usable program code
The computer program product that storage medium is implemented on (including but is not limited to magnetic disk storage, CD-ROM, optical memory etc.)
Form.
Finally it should be noted that：The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention,
Although the present invention is described in detail with reference to the foregoing embodiments, for a person skilled in the art, it still may be used
To be modified to the technical scheme described in foregoing embodiments, or to which part technical characteristic progress equivalent.
Within the spirit and principles of the invention, any modifications, equivalent substitutions and improvements made etc., should be included in the present invention's
Within protection domain.
Claims (6)
1. a kind of high-precision laser 3D profiles phone structural detection method, it is characterised in that including：
Step 1) the previously-scanned sample once of laser measuring apparatus is used, set the sweep parameter of sample, including laser power, image
Time for exposure, acquisition range gathers the laser scanning image sample of sample；
Step 2) phone structural is placed to mobile work platform, fixed laser measurement head is started building in surface, servomotor band
Make the movement of platform supporting plate, complete detection, system completed surface profile detection in 2 seconds, and sampling number St=pl/sstep, pl are productions
Product measure length, and sstep is sampling step length；
Step 3) set up profile standard data model；
Step 4) import 3D MODEL C AD data before measurement, data are made up of thousands of to up to ten thousand tri patch, along with scanning side
The slice map of 3D models is intercepted to vertical path, in each spatial triangle and slice plane the formation phase of structural member 3D data
Intersection, calculates the intersection point of each edge of intersecting lens and spatial triangle, judge intersection point whether triangle end points composition line
Between section, if record intersection point coordinate is used as Slice Sampling between 2 line ends；
After the completion of Slice Sampling point record, Slice Sampling point is ranked up, profiled outline data are formed, wherein, in number of contours
According to key point is detected in set, tangent plane contour line is represented with the sequence of key point；
Also, each tangent plane data use (x1, pt1), (x2, pt2), (x3, pt3), (x4, pt4) ..., (xn, ptn) sequence
Key point is constituted；Ptn represents data vertex type, the standard that flat data point and the big data point of curvature judge in detection process
Using different weighted factors, including：
Represent that data vertex type is flat spot using ptn=1, ptn=2 represents that data point is flex point, when curvature variation of data points is more than
Some scope is to be arranged to flex point；
Numerous slice of datas is lined up according to reference axis, data altitude information is formed, altitude information is exactly that section is bent
Height value on line, rectangular altitude data can be expressed as 2D gray scale pcolors, wherein, gray count formula such as formula on image
(2) it is that follow-up matching detection more facilitates by the Z axis data conversion of model into gray value thus shown in；
Pg (x, y)=255* (z (x, y)-z1)/(Z2-Z1) (2)
Step 5) 3D point cloud matching step, measurement 3D point cloud actual to structural member first and input master pattern are carried out before detection
Registration, calculates actual measurement structure part relative to master pattern and displacement and the anglec of rotation, wherein, Model registration algorithm is to change
For arest neighbors ICP algorithm, wherein, the judgment rule of ICP algorithm is：
After product measurement height map generation, height map is converted into gray scale pcolor；
Step 6) actual product height map and reference model height map generation after, both are subjected to registration process, registering method
Use the method for 2D locations of contours；
Step 7) after product registration operation, the range difference that model and actual product measure altitude information is calculated, actual hand
The data and model data of machine structural member measurement are compared, for comparative approach using least square method and the meter of Euclidean distance
Calculation method, different matching degree computational methods are used according to different data vertex types；
Step 8) measurement process feedback control and data quality monitoring；
Wherein, including：Laser image collection and laser controlling are merged, calibration sample image first before systematic survey
Statistic, the brightness of image average of record standard sample, laser linewidth, after being measured every time in measurement process, detection swashs
Whether the line width of light image, average meet requirement, export measurement result if meeting and requiring, otherwise notify that user remeasures
Or multi output result optimizes output result again after processing.Ensure the measurement accuracy of output result in this way.
2. high-precision laser 3D profiles phone structural detection method according to claim 1, it is characterised in that in the above
Before method of testing, also include：
Phone structural (4) is placed on test platform,【Relation and component names are accurately placed, accompanying drawing 1 is refer to】And
The laser line light source (1) sent using blue wavelength 405nm semiconductor line structure light laser irradiates phone structural, its
In, symmetrically arranged first CMOS camera modules (2) and the 2nd CMOS camera modules (3) collection laser line image obtain upper table
Face 3D point cloud data, wherein, camera optical axis and laser rays optical plane angle have selected 45 degree, wherein, imaging publicity is as follows：
Wherein,
3. high-precision laser 3D profiles phone structural detection method according to claim 1, it is characterised in that step 3)
Specifically include：
3D MODEL C AD data are imported before measurement, are then normalized；
Or, 3D point cloud data are obtained using master sample sampling, then standard three-dimensional point cloud are generated after resampling, filtering.
4. high-precision laser 3D profiles phone structural detection method according to claim 1, it is characterised in that step 3)
Specifically include：
Standard form is used as using after the 3D overall sizes data filtering processing of the actual measurement of standard sample.
5. high-precision laser 3D profiles phone structural detection method according to claim 1, it is characterised in that step 6)
Specifically include：
A) in reference model figure, conspicuousness is manually chosen by software interface and positions point feature, such as circular hole, square hole, right angle
Side etc., calculates the geometry measure feature of location feature, including：Length, height, width, area, outline method vector sum partition value are special
Levy, the geometric properties of profile are preserved as template parameter；
B) binary segmentation is carried out to product height map using the partition value of reference model, spot detection is carried out using Blob algorithms,
Blob features and reference model figure the positioning point feature detected is compared, and finds out position of the most like spot as product
Characteristic point, comparative product position feature point and reference model figure anchor point displacement relation, so as to realize actual product and refer to mould
Type registration.
6. high-precision laser 3D profiles phone structural detection method according to claim 1, it is characterised in that step 7)
Specifically include：
For ptn=1 flat surfaces number of types strong point, first to referring to point set fit Plane model, calculated using least square method,
Then calculate actual samples point arrive plane or the distance of plane equation, calculate exceed judge different severity level error thresholds away from
From the number of point, position, finally define decision tree classifier output final result using this patent；
For ptn=2 turning points measurement point, the circular arc camber model of flex point is gone out using least square fitting, is then calculated
The distance between the radius of surface model of actual surface model and reference point set dis, dis values is used as the foundation judged；
Wherein, filtered out in calculating parameter model process in order to avoid interference using iterative algorithm a certain proportion of mixed and disorderly
Data point；Wherein, gating rate is from 10% to 60%.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710153827.XA CN106949848B (en) | 2017-03-15 | 2017-03-15 | A kind of high-precision laser 3D profile phone structural detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710153827.XA CN106949848B (en) | 2017-03-15 | 2017-03-15 | A kind of high-precision laser 3D profile phone structural detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106949848A true CN106949848A (en) | 2017-07-14 |
| CN106949848B CN106949848B (en) | 2019-11-29 |
Family
ID=59471964
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710153827.XA Active CN106949848B (en) | 2017-03-15 | 2017-03-15 | A kind of high-precision laser 3D profile phone structural detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106949848B (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107621235A (en) * | 2017-08-26 | 2018-01-23 | 东莞市三姆森光电科技有限公司 | Mobile phone curved surface housing profile measuring method and its measuring apparatus based on Spectral Confocal technology |
| CN108195289A (en) * | 2017-12-29 | 2018-06-22 | 上海华太信息技术有限公司 | A kind of method measured using line laser to workpiece specific position parameter |
| CN108592847A (en) * | 2018-07-27 | 2018-09-28 | Oppo（重庆）智能科技有限公司 | The method that the groove depth of electronic device is measured using measuring device |
| CN108682012A (en) * | 2018-05-15 | 2018-10-19 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of 3D bend glass profile pattern defect inspection methods for sweeping laser based on line |
| CN109579765A (en) * | 2017-09-29 | 2019-04-05 | 上海时元互联网科技有限公司 | Three-dimensional point cloud data matching method and system based on servo motor and controller |
| CN109596058A (en) * | 2019-02-01 | 2019-04-09 | 东莞中科蓝海智能视觉科技有限公司 | The size detection recognition methods of plastic workpiece |
| CN109739243A (en) * | 2019-01-30 | 2019-05-10 | 东软睿驰汽车技术（沈阳）有限公司 | A kind of vehicle positioning method, automatic Pilot control method and related system |
| CN110049243A (en) * | 2019-04-19 | 2019-07-23 | 博众精工科技股份有限公司 | Image-pickup method, device, equipment and medium |
| CN110276371A (en) * | 2019-05-05 | 2019-09-24 | 杭州电子科技大学 | A kind of container angle recognition methods based on deep learning |
| CN110288644A (en) * | 2018-03-14 | 2019-09-27 | 浙江大学山东工业技术研究院 | The measurement method at the refractory brick surface inclination angle based on fit Plane normal vector |
| CN111193855A (en) * | 2020-03-24 | 2020-05-22 | 无锡创驰电气有限公司 | Binocular 3D camera |
| CN111336947A (en) * | 2020-03-02 | 2020-06-26 | 南昌航空大学 | Mirror surface object line laser scanning method based on binocular point cloud fusion |
| CN111412859A (en) * | 2020-03-24 | 2020-07-14 | 无锡创驰电气有限公司 | Monocular 3D camera |
| CN111504223A (en) * | 2020-04-22 | 2020-08-07 | 荆亮 | Blade profile measuring method, device and system based on line laser sensor |
| CN112102223A (en) * | 2019-06-18 | 2020-12-18 | 通用电气精准医疗有限责任公司 | Method and system for automatically setting scanning range |
| CN113267122A (en) * | 2021-05-12 | 2021-08-17 | 温州大学瓯江学院 | Industrial part size measurement method based on 3D vision sensor |
| CN114399507A (en) * | 2022-03-25 | 2022-04-26 | 季华实验室 | Mobile phone appearance quality detection method and device, electronic equipment and storage medium |
| CN116051542A (en) * | 2023-03-06 | 2023-05-02 | 深圳市深视智能科技有限公司 | Defect detection method and defect detection device |
| CN117053718A (en) * | 2023-10-11 | 2023-11-14 | 贵州黔程弘景工程咨询有限责任公司 | Beam bottom linear model generation method based on beam bottom linear measurement |
| CN117299583A (en) * | 2023-11-25 | 2023-12-29 | 东莞市欧宇精密科技有限公司 | Control method of optical screening machine |
| CN117299583B (en) * | 2023-11-25 | 2024-05-03 | 东莞市欧宇精密科技有限公司 | Control method of optical screening machine |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5546189A (en) * | 1994-05-19 | 1996-08-13 | View Engineering, Inc. | Triangulation-based 3D imaging and processing method and system |
| US6915004B2 (en) * | 2002-02-28 | 2005-07-05 | Cti Pet Systems, Inc. | Continuous tomography bed motion data processing apparatus and method |
| CN101639452A (en) * | 2009-09-11 | 2010-02-03 | 北京科技大学 | 3D detection method for rail surface defects |
| CN101907439A (en) * | 2010-03-17 | 2010-12-08 | 中国二十二冶集团有限公司 | Stimulated measurement and detection method in architectural steel structure fabrication |
| CN102592284A (en) * | 2012-02-27 | 2012-07-18 | 上海交通大学 | Method for transforming part surface appearance three-dimensional high-density point cloud data into grayscale image |
| CN105526882A (en) * | 2015-12-28 | 2016-04-27 | 西南交通大学 | Turnout wear detection system and detection method based on structured light measurement |
-
2017
- 2017-03-15 CN CN201710153827.XA patent/CN106949848B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5546189A (en) * | 1994-05-19 | 1996-08-13 | View Engineering, Inc. | Triangulation-based 3D imaging and processing method and system |
| US6915004B2 (en) * | 2002-02-28 | 2005-07-05 | Cti Pet Systems, Inc. | Continuous tomography bed motion data processing apparatus and method |
| CN101639452A (en) * | 2009-09-11 | 2010-02-03 | 北京科技大学 | 3D detection method for rail surface defects |
| CN101907439A (en) * | 2010-03-17 | 2010-12-08 | 中国二十二冶集团有限公司 | Stimulated measurement and detection method in architectural steel structure fabrication |
| CN102592284A (en) * | 2012-02-27 | 2012-07-18 | 上海交通大学 | Method for transforming part surface appearance three-dimensional high-density point cloud data into grayscale image |
| CN105526882A (en) * | 2015-12-28 | 2016-04-27 | 西南交通大学 | Turnout wear detection system and detection method based on structured light measurement |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107621235A (en) * | 2017-08-26 | 2018-01-23 | 东莞市三姆森光电科技有限公司 | Mobile phone curved surface housing profile measuring method and its measuring apparatus based on Spectral Confocal technology |
| CN107621235B (en) * | 2017-08-26 | 2024-02-27 | 东莞市三姆森光电科技有限公司 | Method and equipment for measuring contour of curved surface shell of mobile phone based on spectral confocal technology |
| CN109579765B (en) * | 2017-09-29 | 2020-09-18 | 上海时元互联网科技有限公司 | Three-dimensional point cloud data matching method and system based on servo motor and controller |
| CN109579765A (en) * | 2017-09-29 | 2019-04-05 | 上海时元互联网科技有限公司 | Three-dimensional point cloud data matching method and system based on servo motor and controller |
| CN108195289A (en) * | 2017-12-29 | 2018-06-22 | 上海华太信息技术有限公司 | A kind of method measured using line laser to workpiece specific position parameter |
| CN110288644A (en) * | 2018-03-14 | 2019-09-27 | 浙江大学山东工业技术研究院 | The measurement method at the refractory brick surface inclination angle based on fit Plane normal vector |
| CN108682012A (en) * | 2018-05-15 | 2018-10-19 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of 3D bend glass profile pattern defect inspection methods for sweeping laser based on line |
| CN108592847A (en) * | 2018-07-27 | 2018-09-28 | Oppo（重庆）智能科技有限公司 | The method that the groove depth of electronic device is measured using measuring device |
| CN109739243A (en) * | 2019-01-30 | 2019-05-10 | 东软睿驰汽车技术（沈阳）有限公司 | A kind of vehicle positioning method, automatic Pilot control method and related system |
| CN109596058A (en) * | 2019-02-01 | 2019-04-09 | 东莞中科蓝海智能视觉科技有限公司 | The size detection recognition methods of plastic workpiece |
| CN110049243A (en) * | 2019-04-19 | 2019-07-23 | 博众精工科技股份有限公司 | Image-pickup method, device, equipment and medium |
| CN110276371A (en) * | 2019-05-05 | 2019-09-24 | 杭州电子科技大学 | A kind of container angle recognition methods based on deep learning |
| CN110276371B (en) * | 2019-05-05 | 2021-05-07 | 杭州电子科技大学 | Container corner fitting identification method based on deep learning |
| CN112102223A (en) * | 2019-06-18 | 2020-12-18 | 通用电气精准医疗有限责任公司 | Method and system for automatically setting scanning range |
| CN111336947A (en) * | 2020-03-02 | 2020-06-26 | 南昌航空大学 | Mirror surface object line laser scanning method based on binocular point cloud fusion |
| CN111412859A (en) * | 2020-03-24 | 2020-07-14 | 无锡创驰电气有限公司 | Monocular 3D camera |
| CN111193855A (en) * | 2020-03-24 | 2020-05-22 | 无锡创驰电气有限公司 | Binocular 3D camera |
| CN111193855B (en) * | 2020-03-24 | 2021-09-03 | 无锡创驰电气有限公司 | Binocular 3D camera |
| CN111504223A (en) * | 2020-04-22 | 2020-08-07 | 荆亮 | Blade profile measuring method, device and system based on line laser sensor |
| CN111504223B (en) * | 2020-04-22 | 2022-05-31 | 荆亮 | Blade profile measuring method, device and system based on line laser sensor |
| CN113267122A (en) * | 2021-05-12 | 2021-08-17 | 温州大学瓯江学院 | Industrial part size measurement method based on 3D vision sensor |
| CN114399507A (en) * | 2022-03-25 | 2022-04-26 | 季华实验室 | Mobile phone appearance quality detection method and device, electronic equipment and storage medium |
| CN114399507B (en) * | 2022-03-25 | 2022-06-17 | 季华实验室 | Mobile phone appearance quality detection method and device, electronic equipment and storage medium |
| CN116051542A (en) * | 2023-03-06 | 2023-05-02 | 深圳市深视智能科技有限公司 | Defect detection method and defect detection device |
| CN117053718A (en) * | 2023-10-11 | 2023-11-14 | 贵州黔程弘景工程咨询有限责任公司 | Beam bottom linear model generation method based on beam bottom linear measurement |
| CN117053718B (en) * | 2023-10-11 | 2023-12-12 | 贵州黔程弘景工程咨询有限责任公司 | Beam bottom linear model generation method based on beam bottom linear measurement |
| CN117299583A (en) * | 2023-11-25 | 2023-12-29 | 东莞市欧宇精密科技有限公司 | Control method of optical screening machine |
| CN117299583B (en) * | 2023-11-25 | 2024-05-03 | 东莞市欧宇精密科技有限公司 | Control method of optical screening machine |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106949848B (en) | 2019-11-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106949848B (en) | A kind of high-precision laser 3D profile phone structural detection method | |
| Xu et al. | Line structured light calibration method and centerline extraction: A review | |
| CN107064170A (en) | One kind detection phone housing profile tolerance defect method | |
| CN113252700B (en) | Structural crack detection method, equipment and system | |
| CN104913737A (en) | Component quality checking device based on line laser three-dimensional measurement and detection method of device | |
| CN102800096B (en) | Robustness estimation algorithm of camera parameter | |
| CN104359403A (en) | Plane part size measurement method based on sub-pixel edge algorithm | |
| CN108090896B (en) | Wood board flatness detection and machine learning method and device and electronic equipment | |
| CN104024793A (en) | Shape inspection method and device | |
| CN106814083A (en) | Filter plate defect detecting system and its detection method | |
| CN103411533A (en) | Structured light self-adapting repeated multi-exposure method | |
| CN107345789A (en) | A kind of pcb board hole location detecting device and method | |
| CN113920081A (en) | Cutter wear degree detection method | |
| CN109141273B (en) | DMD-based high-speed moving target deformation measurement system and method | |
| CN110110758A (en) | A kind of surface roughness classification method based on convolutional neural networks | |
| US8830464B2 (en) | Film thickness, refractive index, and extinction coefficient determination for film curve creation and defect sizing in real time | |
| CN113983929A (en) | Method for measuring and evaluating position degree and profile degree of blade-shaped hole of runner blade-shaped plate | |
| JP5226510B2 (en) | System and method for measuring curvature of optical surfaces | |
| Yu et al. | Visual texture-based 3-D roughness measurement for additive manufacturing surfaces | |
| CN107727013A (en) | Surface roughness measuring method based on support vector regression | |
| Li et al. | Triangulation-based edge measurement using polyview optics | |
| CN115656182A (en) | Sheet material point cloud defect detection method based on tensor voting principal component analysis | |
| CN110017998A (en) | Vehicle checking method, device and equipment | |
| Frangione et al. | Multi-step approach for automated scaling of photogrammetric micro-measurements | |
| JP2015503110A (en) | Sensor for measuring surface non-uniformity |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP02 | Change in the address of a patent holder |
Address after: 518000 room 501-504, building 2, Chongwen Park, Nanshan wisdom Park, 3370 Liuxian Avenue, Fuguang community, Taoyuan Street, Nanshan District, Shenzhen City, Guangdong Province Patentee after: SHENZHEN SINCEVISION TECHNOLOGY Co.,Ltd. Address before: 518000 room 519, block a, Shenzhen Zhihui innovation center, Taoyuanju south of Qianjin 2nd Road, Bao'an District, Shenzhen City, Guangdong Province Patentee before: SHENZHEN SINCEVISION TECHNOLOGY Co.,Ltd. |
|
| CP02 | Change in the address of a patent holder |