CN113432543A - Method for measuring diameter size of edge part - Google Patents

Method for measuring diameter size of edge part Download PDF

Info

Publication number
CN113432543A
CN113432543A CN202110675472.7A CN202110675472A CN113432543A CN 113432543 A CN113432543 A CN 113432543A CN 202110675472 A CN202110675472 A CN 202110675472A CN 113432543 A CN113432543 A CN 113432543A
Authority
CN
China
Prior art keywords
edge
pixel
linear array
value
point
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
Application number
CN202110675472.7A
Other languages
Chinese (zh)
Inventor
梅武军
袁国堂
蒋婷婷
龚哲
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Research Institute of Zhejiang University Taizhou
Original Assignee
Research Institute of Zhejiang University Taizhou
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Research Institute of Zhejiang University Taizhou filed Critical Research Institute of Zhejiang University Taizhou
Priority to CN202110675472.7A priority Critical patent/CN113432543A/en
Publication of CN113432543A publication Critical patent/CN113432543A/en
Pending legal-status Critical Current

Links

Images

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
    • G01B11/08—Measuring arrangements characterised by the use of optical techniques for measuring diameters

Abstract

The invention discloses a method for measuring the diameter size of an edge part, which specifically comprises the following steps: 1) linear array CCD data processing step, 2) data fitting positioning edge point step, 3) edge coarse positioning step, 4) size initial measurement step, 5) size measurement step; the invention provides a method for measuring the diameter size of an edge part, which is reasonable in design and high in precision.

Description

Method for measuring diameter size of edge part
Technical Field
The invention relates to the technical field of size measurement, in particular to a method for measuring the diameter size of an edge part.
Background
In the prior art, when measuring the diameter of a tiny cylindrical workpiece such as a metal nozzle, a measurer fits a bolt with a certain size on the outer side of the workpiece to measure. In the case of automatic measurement, the workpiece is placed on a V-table or the like, a stylus of a contact measuring instrument is brought into contact with an outer diameter portion of the workpiece to rotate the workpiece, and an inner diameter dimension of the workpiece is determined from a displacement amount of the stylus at that time. Alternatively, the end face of a workpiece placed on a V-table or the like is imaged by a CCD camera, and the inner diameter of the workpiece is obtained by image processing based on the obtained image data.
However, when the bolt gauge is used for measurement, a measurer needs to perform measurement by manual work piece by piece, and thus, the bolt gauge has a disadvantage of requiring much labor. In addition, the method has the disadvantage of lacking accuracy due to manual operation. In the measurement using the linear array CCD measuring instrument, the influence of many factors can be caused, such as the fluctuation of a background image, which is related to the background light intensity; there are many noises (stray points) in the background, and especially in the measurement under the dynamic complex environment, the occurrence of the noises is random; again, the edges of the target are curvilinear, not sharp, and asymmetric. It is difficult to accurately derive the diameter size of the measurement part.
Disclosure of Invention
The invention overcomes the defects of the prior art and provides the method for measuring the diameter size of the edge part, which is reasonable in design and high in precision.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a method for measuring the diameter size of an edge part specifically comprises the following steps:
1) linear array CCD data processing: carrying out data noise elimination and noise filtering elimination processing on the size measurement chart obtained by the linear array CCD;
2) and (3) data fitting and positioning edge points: obtaining the maximum slope of a certain position by fitting the data processed in the step 1) through a least square method, and taking the maximum slope as an edge point of a laser measuring point;
3) coarse positioning of edges: according to the edge point distribution data obtained in the step 2), carrying out pixel positioning through a Canny operator, and determining an approximate area of an actual edge;
4) primary size determination: deriving the fitted linear function to obtain the optimal edge coordinates of the left side and the right side between the intervals [ a, b ], [ c and d ], so as to obtain the most accurate coordinate value;
5) and (3) size determination: measuring the diameter of the part, and adopting the following basic model:
Figure BDA0003120782000000021
wherein d is the part diameter (mm); h is the single linear array pixel width (mm); delta P is the number of pixels of the linear array shielded by the part; beta is the magnification of the optical system.
Then Δ p ═ pR-pLIn which P isRRepresents a point at which the slope of the right edge is maximum; pLRepresents a point at which the slope of the left edge is maximum;
the part diameter is then given by the following formula:
Figure BDA0003120782000000022
wherein d is the part diameter (mm); h is the single linear array pixel width (mm); j is the pixel position of the right linear array shielded by the part; i is the left linear array pixel position shielded by the part; beta is the magnification of the optical system.
Further, the data noise elimination and noise filtering elimination processing in the step 1) is as follows:
taking the main image as a base line, giving two thresholds d and N, and letting H (i) represent the gray value of the ith pixel, assuming no noise at the 1 st pixel, i.e. H (1) is a normal gray value, assuming the number of pixels is M, from 1 to M-1 for the pixel i;
if there is | Hi+1-HiIf | is less than d, the gray value is proved to be continuously changed, and H (i +1) does not belong to the noise point to be eliminated;
the front-end processing of the linear array CCD data is realized by a first-order RC low-pass filter, and if the output impedance of the circuit is very large and the input impedance is very small, the following formula can be obtained:
Figure BDA0003120782000000031
wherein the impedance of the capacitor is:
Figure BDA0003120782000000032
and ω 2 π f with a cutoff frequency of
Figure BDA0003120782000000033
The signal at this frequency, through this circuit, the output voltage and the input voltage have the relationship:
Figure BDA0003120782000000034
or an expression in the time domain:
Figure BDA0003120782000000035
after the above formula is discretized, the final formula is obtained as follows:
Figure BDA0003120782000000036
wherein, V0(k) To output the result, K is the coefficient, Vi(k) For the input data, R is the resistance value of the RC filter, C is the capacitance value of the RC filter, and T is the filter frequency.
Further, the least square method for fitting the straight line in the step 2) is as follows:
the fitting straight line is set as: y is kx + a and error is di＝yi- (kx + a); the least square method is as follows:
Figure BDA0003120782000000041
the parameters k, a of this equation are derived to obtain the following two equations:
Figure BDA0003120782000000042
Figure BDA0003120782000000043
the two equations form a system of equations, solve for a, b, and introduce the means as follows:
Figure BDA0003120782000000044
Figure BDA0003120782000000045
Figure BDA0003120782000000046
Figure BDA0003120782000000047
to obtain
Figure BDA0003120782000000048
Figure BDA0003120782000000049
Wherein k is the slope value of the fitted straight line, a is the compensation coefficient of the fitted straight line,
Figure BDA00031207820000000410
is the average value of the gray value accumulations of the pixel points,
Figure BDA00031207820000000411
and obtaining a function of a least square method fitting straight line for the position value of the pixel point.
Further, the specific steps of the Canny operator in the step 3) are as follows:
firstly, convolution is carried out on an image gray function by using a Gaussian function, the image gray function is set to be f (x), and the one-dimensional zero-mean Gaussian function is shown as the following formula:
Figure BDA0003120782000000051
σ is the standard deviation of normal distribution, e is a natural constant, and x is an input value;
the result of the gaussian function convolving the image gray function is as follows:
h(x)＝f(x)*g(x)
wherein "+" denotes convolution;
the first derivation is followed to obtain a gradient g (x) as follows:
Figure BDA0003120782000000052
finally, the gradient extreme value judgment is carried out as follows:
G(xi)＝max(G(x))
wherein xiFor the coarse positioning of pixel point at the edge, pixel point x is positioned at this momentiGray value f (x) ofi) And point xiAnd respectively storing the adjacent points and the gray values for subsequent processing.
The invention has the beneficial effects that:
the size of the measured object breaks through the limitation that the existing linear array CCD-based size measurement technology is mostly used for measuring small-size objects, and a sub-pixel precision edge detection algorithm is provided; and (3) providing an average distance method to distinguish the upper edge and the lower edge of the part, and finally determining the size of the part by using an imaging principle and the intersection point of the straight line and the surface of the part.
Drawings
FIG. 1 is an overall process flow diagram of the present invention;
FIG. 2 is a schematic diagram of a linear array CCD pixel of the present invention;
FIG. 3 is a graph of data obtained from a filtered drawing of the present invention;
FIG. 4 is a graph of a step edge image gradient analysis according to the present invention;
FIG. 5 is a schematic diagram of a linear array CCD pixel after M times of subdivision interpolation according to the present invention;
FIG. 6 is a flow chart of the sub-pixel high precision edge location data of the present invention;
FIG. 7 is a data diagram of an exemplary CCD of the present invention;
fig. 8 is a structural view of the present invention.
The labels in the figure are: the device comprises a left part 1, a bump 11, a strip-shaped groove 12, a left right trapezoid part 13, a left quadrilateral part 14, a heat dissipation hole 15, a left linear array CCD image acquisition device 16, a guide plate 17, a right part 2, a right trapezoid part 21, a right quadrilateral part 22, a laser sheet 23, a transmitter circuit board 24, a laser head fixing seat 25, a bottom 3 and a guide groove 31.
Detailed Description
The invention is further described with reference to the following figures and specific examples. It should be noted that the examples are only for specifically illustrating the present invention, and the purpose thereof is to make the technical solution of the present invention better understood by those skilled in the art, and should not be construed as limiting the present invention.
Example 1:
as shown in fig. 1 to 8, a method for measuring the diameter and size of an edge part, the size measuring instrument includes a left linear array CCD image collecting device 16, a right light reflecting structure, and a part with a size to be detected placed in the middle, the part is shot by a linear array CCD camera of the left linear array CCD image collecting device 16, and then data processing is performed by a processor in the left linear array CCD image collecting device 16.
The size measuring instrument is integrally in a bilateral symmetry structure and comprises a left part 1, a right part 2 and a bottom 3, wherein the left part 1 and the right part 2 are identical in shape, and the left part 1 and the right part 2 are respectively arranged at two ends of the bottom 3. Two sides of one side of the bottom part 3, which is far away from the left part 1, are provided with guide grooves 31, so that the size measuring instrument can be conveniently arranged on other equipment.
The left portion 1 includes a left rectangular trapezoid portion 13 and a left quadrangular portion 14, and the left rectangular trapezoid portion 13 and the left quadrangular portion 14 are communicated. One side of the left right-angle trapezoid part 13 close to the left quadrilateral part 14 is provided with a left linear array CCD image acquisition device 16, the left linear array CCD image acquisition device 16 is fixed by arranging a bump 11 at the bottom 3 of the left right-angle trapezoid part 13, and the bump 11 is provided with a strip-shaped groove 12 for placing the left linear array CCD image acquisition device 16. A splayed guide plate 17 is arranged between the left linear array CCD image acquisition device 16 and one side of the left right-angled trapezoid part 13 far away from the left quadrilateral part 14. A heat radiation hole 15 is provided on one side of the left quadrangle portion 14 near the dimension measuring instrument bottom 3. The left linear array CCD image acquisition device 16 mainly performs laser reception and information processing.
The right part 2 comprises a right rectangular trapezoid part 21 and a right quadrilateral part 22, and the right rectangular trapezoid part 21 is communicated with the right quadrilateral part 22. The right trapezoid part 21 is provided with a laser sheet 23 and a transmitter circuit board 24; the right quadrangle part 22 is provided with a laser head fixing seat 25 for fixing a laser head. The transmitter circuit board 24 is disposed on one side of the right quadrangle portion 22.
The method for measuring the diameter size of the part specifically comprises the following steps:
1) linear array CCD data processing: and carrying out data noise elimination and noise filtering elimination on the dimension measurement chart acquired by the linear array CCD. Because the image obtained by the linear array CCD has three obvious characteristics: (1) the background image is fluctuant and is related to the background light intensity; (2) there are many noises (stray points) in the background, and especially in the measurement under the dynamic complex environment, the occurrence of the noises is random; (3) the edges of the target are curvilinear, not sharp, and asymmetric.
The noise is mainly generated by power supply noise, a front-end acquisition circuit of the linear array CCD, a signal amplification circuit and an AD conversion circuit, the amplitude and the appearance position of the noise are generally random and are basically isolated points. In order to eliminate the stray points, the following comparison method is adopted for elimination:
taking the main image as a base line, giving two thresholds d and N, and letting H (i) represent the gray value of the ith pixel, assuming no noise at the 1 st pixel, i.e. H (1) is a normal gray value, assuming the number of pixels is M, from 1 to M-1 for the pixel i;
if there is | Hi+1-HiIf | is less than d, the gray value is proved to be continuously changed, and H (i +1) does not belong to the noise point which should be eliminated. If | Hi+1-HiIf | > d, it means that the gray value of the i +1 th pixel has too large variation, and there is a possibility of being a stray point, but there is a possibility of being a target signal. Subsequent processing may be retained.
The noise filtering elimination is realized by adopting a first-order RC low-pass filter to process the front end of the linear array CCD data, and the following formula can be obtained on the assumption that the output impedance of the circuit is very large and the input impedance is very small:
Figure BDA0003120782000000081
where j is a constant parameter, the impedance of the capacitor is:
Figure BDA0003120782000000082
and ω 2 π f with a cutoff frequency of
Figure BDA0003120782000000083
The signal at this frequency, through this circuit, the output voltage and the input voltage have the relationship:
Figure BDA0003120782000000084
or an expression in the time domain:
Figure BDA0003120782000000085
after the above formula is discretized, the final formula is obtained as follows:
Figure BDA0003120782000000091
wherein, V0(k) To output the result, K is the coefficient, Vi(k) For the input data, R is the resistance value of the RC filter, C is the capacitance value of the RC filter, and T is the filter frequency.
Depending on the actual design requirements of the product, and taking into account the effect of the low-pass filter circuit on the overall measurement, it is generally recommended to use a frequency of 200Hz, i.e. fcutThe RC value was calculated at 200, and the filtered drawing data is shown in fig. 3.
2) And (3) data fitting and positioning edge points: obtaining the maximum slope at a certain position by fitting the data processed in the step 1) through a least square method, and taking the maximum slope as an edge point of a laser measuring point.
The method for fitting the straight line by the least square method comprises the following steps:
the fitting straight line is set as: y is kx + a and error is di＝yi- (kx + a); the least square method is as follows:
Figure BDA0003120782000000092
the parameters k, a of this equation are derived to obtain the following two equations:
Figure BDA0003120782000000093
Figure BDA0003120782000000094
the two equations form a system of equations, solve for a, b, and introduce the means as follows:
Figure BDA0003120782000000095
Figure BDA0003120782000000096
Figure BDA0003120782000000097
Figure BDA0003120782000000101
to obtain
Figure BDA0003120782000000102
Figure BDA0003120782000000103
Wherein k is the slope value of the fitted straight line, a is the compensation coefficient of the fitted straight line,
Figure BDA0003120782000000104
is the average value of the gray value accumulations of the pixel points,
Figure BDA0003120782000000105
and obtaining a function of a least square method fitting straight line for the position value of the pixel point.
Specifically, taking fig. 7 as an example, after the MCU receives a group of data at the edge of the CCD, the detailed 16-bit data is: FD C9 CB FF C3A 2 AB A381718D 786057535E 3E 313022282D is 22 data in total, X is a natural number from 1 to 22, and y is 22 data values in the first 16-ary system, which can be derived from the above formula as follows:
Figure BDA0003120782000000111
Figure BDA0003120782000000112
Figure BDA0003120782000000113
Figure BDA0003120782000000114
Figure BDA0003120782000000115
Figure BDA0003120782000000116
the axis formula of the part tool calculated by fitting is as follows:
y＝-9.67154x+233.495
according to the comparison of the gray value of the edge on the picture and the fitted data, the data is basically well processed.
3) Coarse positioning of edges: and (3) carrying out pixel positioning through a Canny operator according to the edge point distribution data obtained in the step 2), and determining an approximate area of the actual edge.
Before the sub-pixel edge is positioned on the whole, the position of an edge point is determined by adopting a pixel level edge detection method, and then the pixel positioning is carried out according to the gray distribution near the edge point; selecting a Canny operator to extract the pixel-level edge, firstly, performing Gaussian smoothing on the image, and realizing the Gaussian smoothing through Gaussian convolution; a simple one-dimensional first order differential operation is then performed on the smoothed image.
The core idea of the implementation of the Canny operator is that the Canny operator is adopted to carry out pixel-level edge rough positioning to determine the approximate area of the actual edge; and then only for coarsely located pixel regions.
The basic idea of detecting a step edge is that a step edge in an image corresponds to a local extreme point of the image gradient, so that an image edge can be detected by finding the extreme point of the image gradient. However, the actual image may not be very sharp due to the low pass filtering effect of the optical system and the circuitry, and the noise in the scene. That is to say, the approximation of the image gradient faces two requirements of noise suppression effect and accurate positioning, and the two requirements are mutually influenced and cannot be simultaneously met, and the details are sacrificed if the noise is reduced; if the details are kept, the noise cannot be suppressed effectively. The scheme adopts the first derivative of the Gaussian function as a linear operator, smoothes the image to inhibit noise, can select an optimal compromise scheme between noise interference resistance and accurate positioning, and well solves the contradiction. This smoothing is performed over the entire image plane. The nature of the canny operator is the first derivative of the gaussian function, which is the optimal approximation operator for the product of signal-to-noise ratio and localization.
The Canny operator comprises the following specific steps:
firstly, convolution is carried out on an image gray function by using a Gaussian function, the image gray function is set to be f (x), and the Gaussian function of one-dimensional zero mean is shown as the following formula:
Figure BDA0003120782000000121
σ is the standard deviation of normal distribution, e is a natural constant, and x is an input value;
then, the results are as follows:
h(x)＝f(x)*g(x)
wherein "+" denotes convolution;
the first derivation is followed to obtain a gradient g (x) as follows:
Figure BDA0003120782000000131
finally, the gradient extreme value judgment is carried out as follows:
G(xi)＝max(G(x))
wherein xiFor the coarse positioning of pixel point at the edge, pixel point x is positioned at this momentiGray value f (x) ofi) And point xiAnd respectively storing the adjacent points and the gray values for subsequent processing.
4) Primary size determination: acquiring a part surface intersection point by adopting a pixel mean value interpolation mode to determine the size of the part; the pixel mean interpolation method is specifically as follows:
finding the position of an edge central section in data acquired by the linear array CCD, and interpolating in an M-time subdivision interpolation mode of an average value; finding suitable intervals a in the edge area1、a2、a3、a4Four edge region pixels are found to be matched with q1、q2、q3、q4Gray value, inserting a proper number of points a by meann，qnPoints, i.e. currently the firstSub-pixel coordinate of degree N is anThe specific insertion method is as follows:
Figure BDA0003120782000000132
finding out the corresponding gray value q by the least square method fitting function in the step 2)nThereby obtaining corresponding complete interpolated data;
5) and (3) size determination: and (4) carrying out secondary data processing by repeating the step 3) and the step 4), and accurately positioning the edge data to obtain the diameter size data of the part.
Example 2:
this embodiment is substantially the same as embodiment 1, and is different from this embodiment in that step 4) obtains the optimal left and right edge coordinates between the intervals [ a, b ], [ c, d ] by deriving the fitted straight-line function, thereby obtaining a most accurate coordinate value.
Measuring the diameter size of the part in the step 5), wherein the adopted basic model is as follows:
Figure BDA0003120782000000141
wherein d is the part diameter (mm); h is the single linear array pixel width (mm); delta P is the number of pixels of the linear array shielded by the part; beta is the magnification of the optical system.
Then Δ p ═ pR-pLIn which P isRRepresents a point at which the slope of the right edge is maximum; pLRepresents a point at which the slope of the left edge is maximum;
the part diameter is then given by the following formula:
Figure BDA0003120782000000142
wherein d is the part diameter (mm); h is the single linear array pixel width (mm); j is the pixel position of the right linear array shielded by the part; i is the left linear array pixel position shielded by the part; beta is the magnification of the optical system.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these modifications and improvements should also be considered as the protection scope of the present invention.

Claims (4)

1. A method for measuring the diameter size of an edge part is characterized by comprising the following steps:
1) linear array CCD data processing: carrying out data noise elimination and noise filtering elimination processing on the size measurement chart obtained by the linear array CCD;
2) and (3) data fitting and positioning edge points: obtaining the maximum slope of a certain position by fitting the data processed in the step 1) through a least square method, and taking the maximum slope as an edge point of a laser measuring point;
3) coarse positioning of edges: according to the edge point distribution data obtained in the step 2), carrying out pixel positioning through a Canny operator, and determining an approximate area of an actual edge;
4) primary size determination: deriving the fitted linear function to obtain the optimal edge coordinates of the left side and the right side between the intervals [ a, b ], [ c and d ], so as to obtain the most accurate coordinate value;
5) and (3) size determination: measuring the diameter of the part, and adopting the following basic model:
Figure FDA0003120781990000011
wherein d is the part diameter (mm); h is the single linear array pixel width (mm); delta P is the number of pixels of the linear array shielded by the part; beta is the magnification of the optical system.
Then Δ p ═ pR-pLIn which P isRRepresents a point at which the slope of the right edge is maximum; pLRepresents a point at which the slope of the left edge is maximum;
the part diameter is then given by the following formula:
Figure FDA0003120781990000012
wherein d is the part diameter (mm); h is the single linear array pixel width (mm); j is the pixel position of the right linear array shielded by the part; i is the left linear array pixel position shielded by the part; beta is the magnification of the optical system.
2. The method for measuring the diameter size of the edge part according to claim 1, wherein the data noise elimination and noise filtering elimination in the step 1) is as follows:
taking the main image as a base line, giving two thresholds d and N, and letting H (i) represent the gray value of the ith pixel, assuming no noise at the 1 st pixel, i.e. H (1) is a normal gray value, assuming the number of pixels is M, from 1 to M-1 for the pixel i;
if there is | Hi+1-HiIf | is less than d, the gray value is proved to be continuously changed, and H (i +1) does not belong to the noise point to be eliminated;
the front-end processing of the linear array CCD data is realized by a first-order RC low-pass filter, and if the output impedance of the circuit is very large and the input impedance is very small, the following formula can be obtained:
Figure FDA0003120781990000021
wherein the impedance of the capacitor is:
Figure FDA0003120781990000022
and ω 2 π f with a cutoff frequency of
Figure FDA0003120781990000023
The signal at this frequency, through this circuit, the output voltage and the input voltage have the relationship:
Figure FDA0003120781990000024
or an expression in the time domain:
Figure FDA0003120781990000025
after the above formula is discretized, the final formula is obtained as follows:
Figure FDA0003120781990000026
wherein, V0(k) To output the result, K is the coefficient, Vi(k) For the input data, R is the resistance value of the RC filter, C is the capacitance value of the RC filter, and T is the filter frequency.
3. The method for measuring the diameter size of the edge part according to claim 1, wherein the least square method for fitting the straight line in the step 2) is as follows:
the fitting straight line is set as: y-kx + a error di＝yi- (kx + a) least squares:
Figure FDA0003120781990000031
the parameters k, a of this equation are derived to obtain the following two equations:
Figure FDA0003120781990000032
Figure FDA0003120781990000033
the two equations form a system of equations, solve for a, b, and introduce the means as follows:
Figure FDA0003120781990000034
Figure FDA0003120781990000035
Figure FDA0003120781990000036
Figure FDA0003120781990000037
to obtain
Figure FDA0003120781990000038
Figure FDA0003120781990000039
Wherein k is the slope value of the fitted straight line, a is the compensation coefficient of the fitted straight line,
Figure FDA00031207819900000310
is the average value of the gray value accumulations of the pixel points,
Figure FDA00031207819900000311
and obtaining a function of a least square method fitting straight line for the position value of the pixel point.
4. The method for measuring the diameter dimension of the edge part according to claim 1, wherein the Canny operator in the step 3) comprises the following specific steps:
firstly, convolution is carried out on an image gray function by using a Gaussian function, the image gray function is set to be f (x), and the one-dimensional zero-mean Gaussian function is shown as the following formula:
Figure FDA0003120781990000041
σ is the standard deviation of normal distribution, e is a natural constant, and x is an input value;
the result of the gaussian function convolving the image gray function is as follows:
h(x)＝f(x)*g(x)
wherein "+" denotes convolution;
the first derivation is followed to obtain a gradient g (x) as follows:
Figure FDA0003120781990000042
finally, the gradient extreme value judgment is carried out as follows:
G(xi)＝max(G(x))
wherein xiFor the coarse positioning of pixel point at the edge, pixel point x is positioned at this momentiGray value f (x) ofi) And point xiAnd respectively storing the adjacent points and the gray values for subsequent processing.
CN202110675472.7A 2021-06-18 2021-06-18 Method for measuring diameter size of edge part Pending CN113432543A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110675472.7A CN113432543A (en) 2021-06-18 2021-06-18 Method for measuring diameter size of edge part

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110675472.7A CN113432543A (en) 2021-06-18 2021-06-18 Method for measuring diameter size of edge part

Publications (1)

Publication Number Publication Date
CN113432543A true CN113432543A (en) 2021-09-24

Family

ID=77756516

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110675472.7A Pending CN113432543A (en) 2021-06-18 2021-06-18 Method for measuring diameter size of edge part

Country Status (1)

Country Link
CN (1) CN113432543A (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DD241470A1 (en) * 1985-10-02 1986-12-10 Smab Forsch Entw Rat DEVICE FOR TOUCH-FREE MEASUREMENT OF EXTERNAL DIAMETERS
CN104359403A (en) * 2014-11-21 2015-02-18 天津工业大学 Plane part size measurement method based on sub-pixel edge algorithm
CN104732536A (en) * 2015-03-18 2015-06-24 广东顺德西安交通大学研究院 Sub-pixel edge detection method based on improved morphology
CN104981105A (en) * 2015-07-09 2015-10-14 广东工业大学 Detecting and error-correcting method capable of rapidly and accurately obtaining element center and deflection angle
WO2015177268A1 (en) * 2014-05-23 2015-11-26 Ventana Medical Systems, Inc. Systems and methods for detection of biological structures and/or patterns in images
CN106247969A (en) * 2016-09-21 2016-12-21 哈尔滨工业大学 A kind of deformation detecting method of industrial magnetic core element based on machine vision
CN106568384A (en) * 2017-01-17 2017-04-19 浙江大学台州研究院 Size measurement method for large-chamfer quartz wafer
CN109612390A (en) * 2018-12-17 2019-04-12 江南大学 Large-size workpiece automatic measuring system based on machine vision

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DD241470A1 (en) * 1985-10-02 1986-12-10 Smab Forsch Entw Rat DEVICE FOR TOUCH-FREE MEASUREMENT OF EXTERNAL DIAMETERS
WO2015177268A1 (en) * 2014-05-23 2015-11-26 Ventana Medical Systems, Inc. Systems and methods for detection of biological structures and/or patterns in images
CN104359403A (en) * 2014-11-21 2015-02-18 天津工业大学 Plane part size measurement method based on sub-pixel edge algorithm
CN104732536A (en) * 2015-03-18 2015-06-24 广东顺德西安交通大学研究院 Sub-pixel edge detection method based on improved morphology
CN104981105A (en) * 2015-07-09 2015-10-14 广东工业大学 Detecting and error-correcting method capable of rapidly and accurately obtaining element center and deflection angle
CN106247969A (en) * 2016-09-21 2016-12-21 哈尔滨工业大学 A kind of deformation detecting method of industrial magnetic core element based on machine vision
CN106568384A (en) * 2017-01-17 2017-04-19 浙江大学台州研究院 Size measurement method for large-chamfer quartz wafer
CN109612390A (en) * 2018-12-17 2019-04-12 江南大学 Large-size workpiece automatic measuring system based on machine vision

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
QUN HAO: "CCD probe: application of linear array CCD in large dimension measurement", 《VIDEOMETRICS V》, pages 330 - 334 *
刘国阳: "基于机器视觉的微小零件尺寸测量技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑》, no. 03, pages 138 - 2012 *
刘霞: "工业零件形状尺寸的机器视觉检测系统的研究", 《中国优秀硕士学位论文全文数据库信息科技辑》, no. 03, pages 138 - 827 *

Similar Documents

Publication Publication Date Title
CN103292701B (en) The online dimension measurement method of accurate device based on machine vision
CN110163918B (en) Line structure cursor positioning method based on projective geometry
CN106197612B (en) A kind of transparent bottled liquid-level detecting method based on machine vision
US4924506A (en) Method for directly measuring area and volume using binocular stereo vision
CN105627939A (en) Device, method and system for detecting micro gap based on industrial equipment
CN109489566A (en) Lithium battery diaphragm material cuts checking method for width, detection system and device
CN108088381B (en) Non-contact type micro gap width measuring method based on image processing
CN102809494A (en) Knife-edge measuring method of modulation transfer function of digital X-ray imaging system
CN108876860A (en) A kind of image calibration method for pipe bubble offset measurement
CN113436156A (en) Linear array CCD-based sub-pixel edge part diameter size measurement method
CN112504240B (en) Laser demarcation device calibration system and calibration method
CN105758312B (en) A kind of absolute grating scale and its striped boundary alignment method
CN108827197A (en) A kind of linear array industry CT homogeneous material dimension measurement method for reducing edge degradation and influencing
KR20000034922A (en) Removal of noise from a signal obtained with an imaging system
CN110017797A (en) A kind of dimension measurement result Evaluation of Uncertainty method based on image contour surface split plot design
CN215725699U (en) Dimension measuring device
CN113432543A (en) Method for measuring diameter size of edge part
CN113607058B (en) Straight blade size detection method and system based on machine vision
CN116880353A (en) Machine tool setting method based on two-point gap
CN116448053A (en) Method for positioning off-state light spots based on laser triangulation ranging system
CN107192348A (en) A kind of high-precision 3D vision measuring methods
CN113554688B (en) O-shaped sealing ring size measurement method based on monocular vision
Kee et al. A simple approach to fine wire diameter measurement using a high-resolution flatbed scanner
Pham et al. A Mobile Vision-based System for Gap and Flush Measuring between Planar Surfaces using ArUco Markers
CN110070528B (en) Two-dimensional detection method applied to online positioning of die insert workpiece machining

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