CN110320100B - Double-camera Brinell hardness measuring device and measuring method based on machine vision - Google Patents
Double-camera Brinell hardness measuring device and measuring method based on machine vision Download PDFInfo
- Publication number
- CN110320100B CN110320100B CN201910596173.7A CN201910596173A CN110320100B CN 110320100 B CN110320100 B CN 110320100B CN 201910596173 A CN201910596173 A CN 201910596173A CN 110320100 B CN110320100 B CN 110320100B
- Authority
- CN
- China
- Prior art keywords
- image
- indentation
- camera
- brinell hardness
- view camera
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000007373 indentation Methods 0.000 claims abstract description 81
- 238000007542 hardness measurement Methods 0.000 claims abstract description 10
- 238000001514 detection method Methods 0.000 claims description 20
- 238000012937 correction Methods 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000000354 decomposition reaction Methods 0.000 claims description 8
- 239000000463 material Substances 0.000 claims description 8
- 230000009466 transformation Effects 0.000 claims description 8
- 229910000831 Steel Inorganic materials 0.000 claims description 7
- 239000010959 steel Substances 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 238000001228 spectrum Methods 0.000 claims description 4
- 239000000126 substance Substances 0.000 claims description 4
- 230000001360 synchronised effect Effects 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000006872 improvement Effects 0.000 claims description 2
- 238000005259 measurement Methods 0.000 abstract description 11
- 238000012545 processing Methods 0.000 description 10
- 238000004519 manufacturing process Methods 0.000 description 6
- 238000003708 edge detection Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000007546 Brinell hardness test Methods 0.000 description 1
- 238000007547 Knoop hardness test Methods 0.000 description 1
- 238000007550 Rockwell hardness test Methods 0.000 description 1
- 238000007551 Shore hardness test Methods 0.000 description 1
- 238000007545 Vickers hardness test Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000000227 grinding Methods 0.000 description 1
- 238000003702 image correction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N3/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N3/02—Details
- G01N3/06—Special adaptations of indicating or recording means
- G01N3/068—Special adaptations of indicating or recording means with optical indicating or recording means
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N3/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N3/40—Investigating hardness or rebound hardness
- G01N3/42—Investigating hardness or rebound hardness by performing impressions under a steady load by indentors, e.g. sphere, pyramid
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/0001—Type of application of the stress
- G01N2203/0003—Steady
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/0058—Kind of property studied
- G01N2203/0076—Hardness, compressibility or resistance to crushing
- G01N2203/0078—Hardness, compressibility or resistance to crushing using indentation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
- G01N2203/02—Details not specific for a particular testing method
- G01N2203/06—Indicating or recording means; Sensing means
- G01N2203/0641—Indicating or recording means; Sensing means using optical, X-ray, ultraviolet, infrared or similar detectors
- G01N2203/0647—Image analysis
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention belongs to the field of Brinell hardness measurement, and particularly relates to a double-camera Brinell hardness measurement device and method based on a machine vision method. The large-view-field video camera and the small-view-field video camera are mounted above the workbench and fixed on the vertical direction driving device. The invention greatly reduces the influence of manual intervention measurement, after the indentation of the standard hardness block is formed by the press, the newly added indentation is automatically identified and positioned by the large-vision camera, then the indentation circle fitting is carried out by the small-vision camera, and further the Brinell hardness is converted.
Description
Technical Field
The invention belongs to the field of Brinell hardness measurement, and particularly relates to a double-camera Brinell hardness measurement device and method based on a machine vision method.
Background
The mechanical properties of materials are of particular importance for the design and processing of various mechanical and constructional engineering, and certain properties are required for various mechanical parts or building components. The expression of hardness is not a definite physical quantity, no method for measuring hardness has a definite quantitative relation with a certain physical property at present, and hardness tests are widely applied to daily production and scientific research, particularly the field of mechanical manufacturing and material science research. The hardness reflects the mechanical property of the material to a certain extent, and plays a significant role in various mechanical designs and processing. At present, 6 hardness test methods which are standardized in China are adopted: brinell hardness test, vickers hardness test, richter hardness test, rockwell hardness test, shore hardness test, and knoop hardness test. The Brinell hardness is mainly proposed by a Swedish scholars, namely Bunaii (J.A. Brinell) in 1900 years, and the measuring method has the advantages of large indentation, small influence of material unevenness on the hardness value in the experimental process, high detection precision, relatively concentrated measuring results and capability of objectively and practically reflecting the mechanical property of the material, so that the method is one of the conventional methods. The main method comprises the following steps: the hardness is evaluated by the surface area of the spherical indentation, the surface area of the indentation is mainly calculated by the average diameter of the indentation obtained by measurement, and the numerical value obtained by dividing the test force under certain conditions by the surface area of the spherical indentation is taken as the Brinell hardness value.
At present, most of China also measures Brinell hardness manually by means of a traditional optical microscope, a sample is ground and polished by a handheld grinding machine, then a steel ball with a certain specification is pressed into the surface of the sample under pressure, and then an operator measures and reads the diameter of an indentation through the optical microscope.
In recent years, due to the rapid development of image processing technology and the absolute advantages of the image processing technology in the non-contact measurement field, the image processing technology has penetrated into various fields including high-precision measurement, defect detection, target identification and positioning and the like, the Brinell hardness measurement system based on machine vision not only enables the measurement hardness to be more visual, but also greatly improves the measurement precision after being combined with an effective image processing algorithm, reduces human factors, improves the stability and enhances the anti-interference capability.
Disclosure of Invention
In order to solve the problems, the invention provides a double-camera Brinell hardness measuring device and a measuring method based on a machine vision method.
The invention adopts the following technical scheme: a double-camera Brinell hardness measuring device based on a machine vision method comprises a base, wherein a workbench is arranged on the base, a horizontal X-direction driving device and a horizontal Y-direction driving device are installed on the workbench, a large-view camera and a small-view camera are installed above the workbench, and the large-view camera and the small-view camera are fixed on a vertical driving device.
Vertical direction drive arrangement is including installing the perpendicular stand on the base, install the lead screw of vertical setting on the perpendicular stand, the lead screw both ends are installed on the bearing frame, lead screw one end is passed through the hold-in range and is connected with motor I, install screw nut on the lead screw, screw nut is fixed with L type connecting plate, still be provided with the slide rail II of vertical setting on the perpendicular stand, L type connecting plate back is installed at slide rail II and can be followed slide rail II and slide, large field of vision camera and little field of vision camera pass through camera balance adjustment mechanism and install at L type connecting plate lower extreme.
The horizontal Y-direction driving device comprises a Y-direction moving platform, two first Y-direction moving slide rails are arranged at the bottom of the Y-direction moving platform, second Y-direction moving slide rails meshed with the two first Y-direction moving slide rails are respectively arranged on the outer sides of the two first Y-direction moving slide rails, and the second Y-direction moving slide rails are fixed on the base; horizontal X direction drive arrangement include two first X direction sliding rails that the workstation bottom set up side by side, the outside of two first X direction sliding rails is provided with respectively rather than the second X direction sliding rails of meshing, second X direction sliding rails installs on Y direction moving table, be provided with a set of drive structure on horizontal Y direction drive arrangement and the horizontal X direction drive arrangement respectively, drive structure includes the motor, feed bar and sliding bearing seat, the motor drive feed bar, the feed bar other end is fixed, set up sliding bearing seat on the feed bar, sliding bearing seat is fixed with the workstation.
A Y-direction moving groove is formed in the base, and the first Y-direction moving slide rail is arranged in the Y-direction moving groove; an X-direction moving groove is formed in the Y-direction moving table, and a first X-direction moving slide rail is arranged in the X-direction moving groove.
The camera balance adjusting mechanism comprises a first mounting plate fixed with an L-shaped connecting plate, a spherical groove is mounted on the first mounting plate, a spheroid capable of rotating randomly is placed in the spherical groove, a screw knob is further arranged on the spherical groove, a connecting rod is arranged on the spheroid and connected with a second mounting plate, and a large-view-field camera and a small-view-field camera are mounted on the second mounting plate.
A measuring method of a double-camera Brinell hardness measuring device based on a machine vision method is characterized in that: comprises the following steps of (a) carrying out,
s100, firstly, starting a motor to enable a lead screw to work through a synchronous belt, driving an L-shaped connecting plate to rise to a height suitable for a large-view camera to collect images, then placing a standard Brinell hardness block with indentations on a workbench, collecting the images of the Brinell hardness block by using the large-view camera, wherein the collected images are called as first images, after the collection is completed, selecting different loads and steel ball diameters according to different materials of the Brinell hardness block, re-performing indentation manufacturing on the Brinell hardness block, after the new indentation manufacturing is completed, placing the Brinell hardness block on the workbench again, and collecting the images of the Brinell hardness block again by using the large-view camera, wherein the collected images are called as second images.
S200, respectively carrying out image enhancement on the first image and the second image obtained by the large-field-of-view camera.
S300, image rotation correction is carried out on the second image obtained by the large-view camera, and the second image is converted into the first image position.
S400-frame difference positioning; carrying out difference on the second image obtained after the newly added indentation is subjected to rotation correction and the first image obtained before the indentation is added to obtain the specific position of the second image, wherein the formula is as follows;
in the above formulaAs a result of rotating the second image after taking the impression,to obtain the first image before indentation, the obtained target detection result image is。
S500, after the second image positioning identification of the newly added indentation is completed through the large-view camera, the view center of the small-view camera is moved to the center of the indentation circle through the movement of the workbench, the small-view camera starts to extract the accurate outline of the single indentation, and finally the small-view camera acquires the image of the single indentation circle.
S600, carrying out gray scale conversion and wavelet denoising pretreatment on the image obtained by the small-view camera.
S700, performing edge extraction on the image obtained by the small-view camera to obtain an edge profile of the indentation circle.
S800, performing circle fitting on the extracted indentation contour, and determining the diameter of the indentation circle.
S900, carrying out arc contour detection on the extremely large indentation circle, extracting coordinates of three characteristic points from the detected edge, and calculating to determine an equation of the circle so as to obtain the radius and the center position of the indentation circle.
S1000, converting the measured circle diameter pixel value into an actual length value through camera calibration, and further calculating to obtain a Brinell hardness value according to a calculation formula of Brinell hardness HB;
wherein F is a predetermined detection force, D is a predetermined detection steel ball diameter, and D is an indentation diameter left in the object after the predetermined force is applied and the object is held for a predetermined time.
The step S300 includes the steps of,
s301, respectively obtaining the first image and the second image, performing Fourier transform after image enhancement to obtain corresponding spectrograms, wherein the rotation transformation of the images in a space domain can be directly reflected in Fourier transform spectrums, and the rotation angles of the images are determined by analyzing the characteristics of the spectrograms.
S302, calculating gradient maps of the two images in the X direction and the Y direction, and summing edge images of the images.
S303, carrying out binarization on the two images to keep the most obvious linear characteristics in the frequency change of the images;
s304, positioning the straight lines in the frequency spectrogram by utilizing Hough transform, and respectively obtaining the angles of the straight lines in the two imagesAndthereby determining the angle of the image to be corrected as(ii) a After the binarization of the spectrogram is completed, straight line detection needs to be carried out on the characteristic information in the spectrogram, and angle information of the straight line is determined to determine an angle which needs to be corrected by the image.
S305, performing angle correction on the image through affine transformation.
Compared with the prior art, the invention has the beneficial effects that:
1. the invention greatly reduces the influence of manual intervention measurement, after the indentation of the standard hardness block is formed by the press, the newly added indentation is automatically identified and positioned by the large-vision camera, then the indentation circle fitting is carried out by the small-vision camera, and further the Brinell hardness is converted.
2. The invention has high measurement precision. And acquiring the image of the single indentation circle by using a small-field-of-view camera, and performing high-precision fitting to obtain a high-precision measurement result.
3. The invention can measure a plurality of indentation circles at one time. After the positioning is completed by the large-view camera, the small-view camera collects images of single indentation circles one by one, and finally Brinell hardness values of a plurality of groups of indentation circles are obtained.
Drawings
FIG. 1 is a schematic structural view of the present invention;
FIG. 2 is a side view of the structure of the present invention;
FIG. 3 is a view of a large and small field of view camera device;
FIG. 4 is a schematic view of a driving device;
FIG. 5 is a view of a camera balance adjustment mechanism;
FIG. 6 is a view of the horizontal X-direction drive;
FIG. 7 is a view of the first Y-direction movable slide rail and the second X-direction movable slide rail;
in the figure, 1-a motor, 2-a synchronous belt, 3-a lead screw, 4-a lead screw nut, 5-an L-shaped connecting plate, 6-a base, 7-a motor, 8-a light bar, 9-a sliding bearing seat, 10-an X-direction moving groove, 11-a second Y-direction moving slide rail, 12-a workbench, 14-a Brinell hardness block, 15-a large-field-of-view camera, 16-a small-field-of-view camera, 17-a camera balance adjusting mechanism, 18-a slide rail, 19-a vertical upright post, 20-a controller box, 21-a leveling base, 22-a Y-direction moving table, 23-an X-direction moving groove, 24-a first X-direction moving slide rail, 25-a second X-direction moving slide rail, 26-a first Y-direction moving slide rail, 17.1-a first mounting plate, 17.2-spherical recess, 17.3-screw knob, 17.4-spheroid, 17.5-second mounting plate.
Detailed Description
The following describes embodiments of the present invention in further detail with reference to the accompanying drawings.
A double-camera Brinell hardness measuring device based on a machine vision method comprises a base 6, wherein a workbench 12 is arranged on the base 6, a horizontal X-direction driving device and a horizontal Y-direction driving device are installed on the workbench 12, a large-view camera 15 and a small-view camera 16 are installed above the workbench 12, and the large-view camera 15 and the small-view camera 16 are fixed on the vertical-direction driving device.
Vertical direction drive arrangement is including installing perpendicular stand 19 on base 6, install the lead screw 3 of vertical setting on the perpendicular stand 19, install on the bearing frame at the lead screw 3 both ends, 3 one end of lead screw is passed through the hold-in range and is connected with motor I1, install lead screw nut 4 on the lead screw 3, lead screw nut 4 is fixed with L type connecting plate 5, still be provided with the slide rail II18 of vertical setting on the perpendicular stand 19, 5 backs of L type connecting plate are installed at slide rail II18 and can be followed slide rail II18 and slide, large-view camera 15 and small-view camera 16 pass through camera balance adjustment mechanism 17 and install at the lower extreme of L type connecting plate 5.
The horizontal Y-direction driving device comprises a Y-direction moving platform 22, two first Y-direction moving slide rails 26 are arranged at the bottom of the Y-direction moving platform 22, second Y-direction moving slide rails 11 meshed with the two first Y-direction moving slide rails 26 are respectively arranged on the outer sides of the two first Y-direction moving slide rails 26, and the second Y-direction moving slide rails 11 are fixed on the base 6; horizontal X direction drive arrangement include two first X direction movable slide rail 24 that workstation 12 bottom set up side by side, two first X direction movable slide rail 24's the outside is provided with respectively rather than the second X direction movable slide rail 25 of meshing, second X direction movable slide rail 25 is installed on Y direction mobile station 22, be provided with a set of drive structure on horizontal Y direction drive arrangement and the horizontal X direction drive arrangement respectively, drive structure includes motor 7, polished rod 8 and sliding bearing seat 9, motor 7 drive polished rod 8, the other end of polished rod 8 is fixed, set up sliding bearing seat 9 on the polished rod 8, sliding bearing seat 9 is fixed with workstation 12.
A Y-direction moving groove 10 is formed in the base 6, and a first Y-direction moving slide rail 26 is arranged in the Y-direction moving groove 10; the Y-direction moving table 22 is provided with an X-direction moving groove 23, and the first X-direction moving rail 24 is provided in the X-direction moving groove 23.
As shown in fig. 6. After the positioning of the large-field camera is completed, the workbench driving system sends out an instruction, and the platform X, Y moves towards the motor to receive a rotation instruction. Taking the X-direction movement of the worktable 12 as an example, when the motor 7 rotates, the light bar 8 is driven to rotate, and the worktable 12 is driven to realize the X-direction movement through the connection effect of the sliding bearing block 9. The inner side of the sliding bearing seat 9 is connected with the light bar 8, and the outer side is connected with the workbench 12. The X-direction moving rails 24 are a pair and engaged with each other, and the inner one is fixed to the table 12, the outer one is fixed to the Y-direction moving table 22, and the inner one is placed in the X-direction moving groove 23. The two sides of the working table 12 are respectively provided with a pair of slide rails which are arranged in the same space and are used for moving in the X direction.
When the worktable 12 moves in the X direction, two slide rails fixed with the worktable in the two groups of mutually meshed slide rails also move in the moving groove, thereby ensuring the accuracy of the X-direction movement of the worktable 12.
X, Y, the moving mechanism is the same, the Y-direction moving rail 11, the inner rail fixed to the Y-direction moving table 22, the outer rail fixed to the base 6, and the Y-direction moving groove 10 on the base 6.
The camera balance adjusting mechanism 17 comprises a first mounting plate 17.1 fixed with the L-shaped connecting plate 5, a spherical groove 17.2 is mounted on the first mounting plate 17.1, a spherical body 17.4 capable of rotating randomly is placed in the spherical groove 17.2, a screw knob 17.3 is further arranged on the spherical groove 17.2, a connecting rod is arranged on the spherical body 17.4 and connected with a second mounting plate 17.5, and a large-view camera 15 and a small-view camera 16 are mounted on the second mounting plate 17.5. The inner end of the screw knob 17.3 does not abut against the ball 17.4.
A measuring method of a double-camera Brinell hardness measuring device based on a machine vision method comprises the following steps,
s100, firstly, starting a motor 1 to enable a screw rod 3 to work through a synchronous belt 2, driving an L-shaped connecting plate 5 to rise to a height suitable for a large-view camera 15 to collect images, then placing a standard Brinell hardness block 14 with indentations on a workbench 12, using the large-view camera 15 to collect the images of the Brinell hardness block 14, wherein the collected images are called first images, selecting different loads and steel ball diameters according to different materials of the Brinell hardness block 14 after collection is finished, re-performing indentation manufacturing on the Brinell hardness block, placing the Brinell hardness block 14 on the workbench 12 again after new indentation manufacturing is finished, using the large-view camera 15 to perform image collection on the Brinell hardness block 14 again, and using the collected images to be called second images.
S200, respectively carrying out image enhancement on the first image and the second image obtained by the large-field-of-view camera 15; because the detected surface has a tiny indentation circle with a diameter less than 1mm, which may cause missing detection, the image needs to be enhanced, the visual effect of the image is improved, the profile characteristics of the indentation in the image are highlighted, and the profile of the tiny indentation circle is more obvious. The contrast of the indentation image is enhanced by selecting an image histogram equalization method, so that the details of the image are effectively improved, the algorithm is small in calculation amount, easy to implement, simple and effective in processing, and suitable for occasions needing online real-time detection, and the formula realized by the method can be expressed as follows:
wherein the first in the input imageThe frequency of occurrence of a gray level is expressed asAnd is andthus, therefore, it isThe obtained curve image is the gray level histogram of the input image. At the same time, the enhancement functionInterval of (1)Inner single-valued single-increment function, and satisfyTherefore, the gray value of each pixel point of the enhanced output image can be obtained as follows:
in the above-mentioned formula, the first and second,i.e. the values of the pixels of the output image, wherein。
S300, carrying out image rotation correction on the second image obtained by the large-view camera 15 to convert the second image into a first image position; an image rotation matching method based on Fourier transform is provided. According to the special property of Fourier transform, the angle correction of the image is realized by performing operations such as filtering, edge detection, binarization, feature extraction positioning, angle calculation, rotation correction and the like on the Fourier transform spectrogram, and the corrected image to be detected and the reference image are in the same direction, so that the image can be accurately positioned to the position of a newly added indentation when the image is subjected to differential processing subsequently. The detailed steps are as follows:
s301, respectively obtaining the first image and the second image, performing Fourier transform after image enhancement to obtain corresponding spectrograms, wherein the rotation transformation of the images in a space domain can be directly reflected in Fourier transform spectrums, and the rotation angles of the images are determined by analyzing the characteristics of the spectrograms.
S302, calculating gradient maps of the two images in the X direction and the Y direction, and summing edge images of the images: the brightness values in the spectrogram represent the intensity of the spectral transformation in the original image, so that image correction can be performed by determining the obvious straight-line features in the spectrogram. In order to make the image angle correction more accurate, before Fourier transformation, feature information of a frequency spectrum image needs to be extracted, firstly, gradient images in the X direction and the Y direction of the image are solved and summed to obtain all edge information of the image, then, the needed feature information is reserved through binarization operation, and the image correcting angle is determined according to the extraction of the feature information.
S303, carrying out binarization on the two images to keep the most obvious linear characteristics in the frequency change of the images;
s304, positioning the straight lines in the frequency spectrogram by utilizing Hough transform, and respectively obtaining the angles of the straight lines in the two imagesAndthereby determining the angle of the image to be corrected as(ii) a After the binaryzation of the spectrogram is completed, straight line detection needs to be carried out on the characteristic information in the spectrogram, and angle information of the straight line is determined to determine an angle which needs to be corrected of the image; the line detection of Hough transform is realized by statistical method, the local maximum value point obtained by voting statistics is found by traversing image space, and the point is regarded as the slope and intercept of the line in the detected image, and for any point, the line detection method can be used for detecting the local maximum value pointAll have a straight line through it:
whereinbIs the intercept of the straight line and is,kfor the slope of this line, the above formula can be expressed in polar form:
from the above formula, it can be seen that each straight line corresponds to a unique one in the spaceBy detecting the angle of two straight linesAndand determining the angle of the image to be rotated as follows:。
s305, performing angle correction on the image through affine transformation.
S400-frame difference positioning; carrying out difference on the second image obtained after the newly added indentation is subjected to rotation correction and the first image obtained before the indentation is added to obtain the specific position of the second image, wherein the formula is as follows;
in the above formulaAs a result of rotating the second image after taking the impression,to obtain the first image before indentation, the obtained target detection result image is。
S500, after the second image positioning identification of the newly added indentation is completed through the large-view camera 15, the view center of the small-view camera 16 is moved to the center of the indentation circle through the movement of the workbench 12, the small-view camera 16 starts to extract the accurate outline of the single indentation, and finally the small-view camera 16 acquires the image of the single indentation circle; the entire table 12 coordinates indicate the center of the field of view of the small field of view camera 16. As shown in the large and small field camera views of fig. 3, there is a relative position between the large field camera 15 and the small field camera 16. By reflecting this relative position in the image acquired by the large-field camera 15 by position calculation, the center of the field of view of the small-field camera 16 can be reflected in the field of view of the large-field camera 15. Namely, the view center of the small view camera 16 can be located in the image acquired by the large view camera 15, the coordinate of the workbench 12 shows the view center of the small view camera 16, the coordinate value of the workbench 12 at the center of the indentation circle is calculated by calculating the pixel distance between the center of the single indentation circle in the image acquired by the large view camera 15 and the view center of the small view camera 16, the pixel equivalent when the image of the Brinell hardness block 14 is acquired by the large view camera 15, and finally the view center of the small view camera 16 is moved to the center of the indentation circle by the movement of the workbench 12, so that the image acquisition of the single indentation circle by the small view camera 16 is realized.
S600, carrying out gray level conversion and wavelet denoising pretreatment on the image obtained by the small-view camera 16; the influence of the external environment on the measurement precision of the Brinell hardness indentation is reduced as much as possible, and a foundation is laid for the subsequent extraction of the indentation outline: according to the characteristics of Brinell hardness images and the requirements of projects in the system, the system selects the Haar wavelet for wavelet decomposition, the Haar wavelet has orthogonality, tight support and linear phase, noise information can be accurately and efficiently captured in the wavelet decomposition and reconstruction processes, the Haar wavelet decomposition calculation process is simple, the image processing efficiency can be improved, and the on-line real-time measurement requirements of the system are met.
On the basis, an improved adaptive wavelet threshold is provided, different thresholds are set according to different image decomposition layer numbers, and the improved wavelet threshold is as follows:
wherein the content of the first and second substances,the method is characterized in that the method is a method for improving the noise variance, N represents the length of a signal, j represents the number of decomposition layers, and the threshold T is relatively reduced along with the increase of j. In order to solve the problem that the hard threshold denoising is discontinuous at the threshold point and the deviation phenomenon of the soft threshold, an improved wavelet threshold denoising function is proposed, and the formula is as follows:
wherein the content of the first and second substances,for the new wavelet coefficients obtained after threshold denoising,Tin order to improve the threshold obtained by the method above,nin order to adjust the parameters of the device,the wavelet coefficients before de-noising. In the improved wavelet threshold function, when adjusting the parametersWhen it is obtainedI.e. the result of the input is the original wavelet coefficients, whenThe resulting threshold function approaches the soft threshold function before improvement. Thus, can be adjustednSuch that the modified threshold function varies between the conventional hard and soft thresholds, therebyAnd obtaining the optimal denoising effect.
S700, performing edge extraction on the image obtained by the small-view camera 16 to obtain an edge profile of an indentation circle; compared with other traditional edge detection operators, the Canny operator is adopted to extract the indentation circle, when the Canny operator is used for edge detection, the error rate is low, each edge point can be well positioned, and the false edge of the image can be removed through setting of high and low thresholds.
S800, performing circle fitting on the extracted indentation contour, and determining the diameter of an indentation circle; a least squares method is used which finds the best match function for a set of data by determining the sum of squares which minimizes the error. And realizing the fitting process to obtain an indentation circle, and solving the pixel value of the diameter of the indentation circle.
S900, carrying out arc contour detection on the extremely large indentation circle, extracting coordinates of three characteristic points from the detected edge, and calculating to determine an equation of the circle so as to obtain the radius and the center position of the indentation circle; due to the hardware limitation of the measuring system, when the diameter of the indentation circle is operated to be 3mm, only partial circular arcs of the indentation circle can be collected when the small-vision camera is used for shooting the indentation, aiming at the situation, the system adopts a three-point method to measure the radius of the circular arcs, firstly carries out image denoising processing on the collected indentation image, then carries out edge detection, extracts coordinates of three characteristic points from the detected edge to carry out calculation, determines the equation of the circle and obtains the radius and the circle center position of the indentation circle.
S1000, converting the measured circle diameter pixel value into an actual length value through camera calibration, and further calculating to obtain a Brinell hardness value according to a calculation formula of Brinell hardness HB;
wherein F is a predetermined detection force, D is a predetermined detection steel ball diameter, and D is an indentation diameter left in the object after the predetermined force is applied and the object is held for a predetermined time.
The above-mentioned details are provided for the purpose of describing the present invention in more detail, but the present invention is not limited to the details, and any modification, addition, or substitution made within the spirit of the present invention should be construed as being included in the scope of the present invention.
Claims (6)
1. A double-camera Brinell hardness measuring device based on a machine vision method is characterized in that: the device comprises a base (6), wherein a workbench (12) is arranged on the base (6), a horizontal X-direction driving device and a horizontal Y-direction driving device are installed on the workbench (12), a large-view camera (15) and a small-view camera (16) are installed above the workbench (12), and the large-view camera (15) and the small-view camera (16) are fixed on the vertical direction driving device;
the measuring method of the double-camera Brinell hardness measuring device based on the machine vision method comprises the following steps,
s100, firstly, starting a motor (1) to enable a lead screw (3) to work through a synchronous belt (2), driving an L-shaped connecting plate (5) to rise to a height suitable for a large-visual-field camera (15) to collect an image, then placing a standard Brinell hardness block (14) with an indentation on a workbench (12), collecting the image of the Brinell hardness block (14) through the large-visual-field camera (15), wherein the collected image is called a first image, after the collection is finished, selecting different loads and steel ball diameters according to different materials of the Brinell hardness block (14), re-indenting the Brinell hardness block, after the new indenting is finished, placing the Brinell hardness block (14) on the workbench (12), and performing image collection on the Brinell hardness block (14) again through the large-visual-field camera (15), wherein the collected image is called a second image;
s200, respectively carrying out image enhancement on a first image and a second image obtained by a large-field-of-view camera (15);
s300, carrying out image rotation correction on a second image obtained by the large-view camera (15) to convert the second image into a first image position;
s400-frame difference positioning; carrying out difference on the second image obtained after the newly added indentation is subjected to rotation correction and the first image obtained before the indentation is added to obtain the specific position of the second image, wherein the formula is as follows;
in the above formulaAs a result of rotating the second image after taking the impression,to obtain the first image before indentation, the obtained target detection result image is;
S500, after the second image positioning identification of the newly added indentation is completed through the large-view camera (15), the view center of the small-view camera (16) is moved to the center of an indentation circle through the movement of the workbench (12), the small-view camera (16) starts to extract the accurate outline of a single indentation, and finally the small-view camera (16) acquires the image of the single indentation circle;
s600, carrying out gray level conversion and wavelet denoising pretreatment on an image obtained by the small-view camera (16);
adopting Haar wavelet to carry out wavelet decomposition, and on the basis, providing an improved self-adaptive wavelet threshold, wherein different thresholds are set according to different image decomposition layer numbers, and the improved wavelet threshold is as follows:
wherein the content of the first and second substances,as noiseThe variance, N, represents the length of the signal, j is the number of decomposition layers, and the threshold T is relatively reduced along with the increase of j, in order to simultaneously solve the problem that the hard threshold denoising is discontinuous at the threshold point and the deviation phenomenon of the soft threshold, an improved wavelet threshold denoising function is provided, and the formula is as follows:
wherein the content of the first and second substances,for the new wavelet coefficients obtained after threshold denoising,Tin order to improve the threshold value obtained by the method,nin order to adjust the parameters of the device,the wavelet coefficients before denoising are obtained; in the improved wavelet threshold function, when adjusting the parametersWhen it is obtainedI.e. the result of the input is the original wavelet coefficients, whenThen, the obtained threshold function is close to the soft threshold function before improvement;
s700, performing edge extraction on an image obtained by a small-view camera (16) to obtain an edge profile of an indentation circle;
s800, performing circle fitting on the extracted indentation contour, and determining the diameter of an indentation circle;
s900, carrying out arc contour detection on the extremely large indentation circle, extracting coordinates of three characteristic points from the detected edge, and calculating to determine an equation of the circle so as to obtain the radius and the center position of the indentation circle;
s1000, converting the measured circle diameter pixel value into an actual length value through camera calibration, and further calculating to obtain a Brinell hardness value according to a calculation formula of Brinell hardness HB;
2. The dual-phase machine brinell hardness measurement device based on machine vision method of claim 1, wherein: vertical direction drive arrangement including installing perpendicular stand (19) on base (6), install lead screw (3) of vertical setting on perpendicular stand (19), install on lead screw (3) both ends on the bearing frame, lead screw (3) one end is passed through the hold-in range and is connected with motor I (1), install screw nut (4) on lead screw (3), screw nut (4) are fixed with L type connecting plate (5), still be provided with slide rail II (18) of vertical setting on perpendicular stand (19), L type connecting plate (5) back is installed at slide rail II (18) and can be followed slide rail II (18) and slide, install at L type connecting plate (5) lower extreme through camera balance adjustment mechanism (17) in big visual field camera (15) and little visual field camera (16).
3. The dual-phase machine brinell hardness measurement device based on machine vision method of claim 2, wherein: the horizontal Y-direction driving device comprises a Y-direction moving platform (22), two first Y-direction moving slide rails (26) are arranged at the bottom of the Y-direction moving platform (22), second Y-direction moving slide rails (11) meshed with the two first Y-direction moving slide rails (26) are respectively arranged on the outer sides of the two first Y-direction moving slide rails (26), and the second Y-direction moving slide rails (11) are fixed on the base (6); horizontal X direction drive arrangement include two first X direction movable slide rail (24) that workstation (12) bottom set up side by side, the outside of two first X direction movable slide rail (24) is provided with respectively rather than the second X direction movable slide rail (25) of meshing, second X direction movable slide rail (25) are installed on Y direction movable table (22), be provided with a set of drive structure on horizontal Y direction drive arrangement and the horizontal X direction drive arrangement respectively, drive structure includes motor (7), polished rod (8) and sliding bearing seat (9), motor (7) drive polished rod (8), polished rod (8) other end is fixed, set up sliding bearing seat (9) on polished rod (8), sliding bearing seat (9) are fixed with workstation (12).
4. The dual-phase machine brinell hardness measurement device based on machine vision method of claim 3, wherein: a Y-direction moving groove (10) is formed in the base (6), and a first Y-direction moving slide rail (26) is arranged in the Y-direction moving groove (10); an X-direction moving groove (23) is formed in the Y-direction moving table (22), and a first X-direction moving slide rail (24) is disposed in the X-direction moving groove (23).
5. The dual-phase machine brinell hardness measurement device based on machine vision method of claim 4, wherein: camera balance adjustment mechanism (17) include with first mounting panel (17.1) that L type connecting plate (5) are fixed, install spherical recess (17.2) on first mounting panel (17.1), placed spheroid (17.4) that can rotate at will in spherical recess (17.2), still be provided with spiral shell knob (17.3) on spherical recess (17.2), be provided with the connecting rod on spheroid (17.4), the connecting rod is connected with second mounting panel (17.5), install large-view camera (15) and small-view camera (16) on second mounting panel (17.5).
6. The dual-phase machine brinell hardness measurement device based on machine vision method of claim 5, wherein: the step S300 includes the steps of,
s301, respectively obtaining a first image and a second image, performing Fourier transform after image enhancement to obtain corresponding spectrograms, wherein the rotation transformation of the images in a space domain can be directly reflected in Fourier transform spectrums, and the rotation angles of the images are determined by analyzing the characteristics of the spectrograms;
s302, calculating gradient maps of the two images in the X direction and the Y direction, and summing edge images of the images;
s303, carrying out binarization on the two images to keep the most obvious linear characteristics in the frequency change of the images;
s304, positioning the straight lines in the frequency spectrogram by utilizing Hough transform, and respectively obtaining the angles of the straight lines in the two imagesAndthereby determining the angle of the image to be corrected as(ii) a After the binaryzation of the spectrogram is completed, straight line detection needs to be carried out on the characteristic information in the spectrogram, and angle information of the straight line is determined to determine an angle which needs to be corrected of the image;
s305, performing angle correction on the image through affine transformation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910596173.7A CN110320100B (en) | 2019-07-03 | 2019-07-03 | Double-camera Brinell hardness measuring device and measuring method based on machine vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910596173.7A CN110320100B (en) | 2019-07-03 | 2019-07-03 | Double-camera Brinell hardness measuring device and measuring method based on machine vision |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110320100A CN110320100A (en) | 2019-10-11 |
CN110320100B true CN110320100B (en) | 2022-04-01 |
Family
ID=68122447
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910596173.7A Active CN110320100B (en) | 2019-07-03 | 2019-07-03 | Double-camera Brinell hardness measuring device and measuring method based on machine vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110320100B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111239158A (en) * | 2020-03-13 | 2020-06-05 | 苏州鑫睿益荣信息技术有限公司 | Automobile instrument panel detection system and detection method based on machine vision |
CN112051160B (en) * | 2020-09-09 | 2022-04-19 | 中山大学 | Segment joint bending stiffness measuring method, system, equipment and storage medium |
CN112348031A (en) * | 2020-11-17 | 2021-02-09 | 安徽理工大学 | Improved wavelet threshold denoising method for removing fingerprint image mixed noise |
CN112706075A (en) * | 2020-12-17 | 2021-04-27 | 武昌船舶重工集团有限公司 | Leveling device for machining equiangular base plate |
CN112964307A (en) * | 2021-03-23 | 2021-06-15 | 深圳联钜自控科技有限公司 | Nut detection device |
CN113436214B (en) * | 2021-06-28 | 2022-08-23 | 山东大学 | Brinell hardness indentation circle measuring method and system and computer readable storage medium |
CN115468738B (en) * | 2022-10-31 | 2024-02-27 | 易思维(杭州)科技股份有限公司 | Measurement precision evaluation device and evaluation method of linear array camera measurement system |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104424641A (en) * | 2013-09-07 | 2015-03-18 | 无锡华御信息技术有限公司 | Detection method for image fuzzy tampering |
CN106408609A (en) * | 2016-09-13 | 2017-02-15 | 江苏大学 | Parallel mechanism end motion pose detection method based on binocular vision |
US10032256B1 (en) * | 2016-11-18 | 2018-07-24 | The Florida State University Research Foundation, Inc. | System and method for image processing using automatically estimated tuning parameters |
CN109261528A (en) * | 2018-09-03 | 2019-01-25 | 广州铁路职业技术学院(广州铁路机械学校) | Express delivery sorting method and device based on binocular vision |
Family Cites Families (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000009622A (en) * | 1998-06-18 | 2000-01-14 | Nittetsu Hokkaido Seigyo System Kk | Apparatus for measuring brinell hardness value |
CN101377457A (en) * | 2007-08-30 | 2009-03-04 | 上海电气集团上海电机厂有限公司 | Method for measuring brinell hardness |
CN104422628B (en) * | 2013-09-03 | 2017-05-24 | 北京时代之峰科技有限公司 | Indentation image identification method and system based on Vickers hardness |
CN106767443B (en) * | 2016-11-22 | 2019-08-30 | 中北大学 | A kind of fully automatic secondary element image detector and measurement method |
CN206875067U (en) * | 2017-05-27 | 2018-01-12 | 郑州工业应用技术学院 | Camera support camera angle governor motion |
CN108562487A (en) * | 2018-03-23 | 2018-09-21 | 西北工业大学 | Block of hardness impression diameter measurement method and device |
CN109460769A (en) * | 2018-11-16 | 2019-03-12 | 湖南大学 | A kind of mobile end system and method based on table character machining and identification |
-
2019
- 2019-07-03 CN CN201910596173.7A patent/CN110320100B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104424641A (en) * | 2013-09-07 | 2015-03-18 | 无锡华御信息技术有限公司 | Detection method for image fuzzy tampering |
CN106408609A (en) * | 2016-09-13 | 2017-02-15 | 江苏大学 | Parallel mechanism end motion pose detection method based on binocular vision |
US10032256B1 (en) * | 2016-11-18 | 2018-07-24 | The Florida State University Research Foundation, Inc. | System and method for image processing using automatically estimated tuning parameters |
CN109261528A (en) * | 2018-09-03 | 2019-01-25 | 广州铁路职业技术学院(广州铁路机械学校) | Express delivery sorting method and device based on binocular vision |
Non-Patent Citations (2)
Title |
---|
基于小波变换图像去噪及边缘检测研究;胡志峰;《中国优秀硕士学位论文全文数据库 信息科技辑》;20190115(第12期);全文 * |
射线检测图像的自适应多尺度积阈值降噪算法;党长营;《无损检测》;20180930(第9期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN110320100A (en) | 2019-10-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110320100B (en) | Double-camera Brinell hardness measuring device and measuring method based on machine vision | |
CN101561249B (en) | Method for automatically detecting fit dimension of surgical knife blade | |
CN102840829B (en) | Manual mark-basedsystem and manual mark-based method for measuring displacement field inside high temperature object area | |
CN109087274A (en) | Electronic device defect inspection method and device based on multidimensional fusion and semantic segmentation | |
CN102032875B (en) | Image-processing-based cable sheath thickness measuring method | |
CN111126174A (en) | Visual detection method for robot to grab parts | |
Dong et al. | Robust circular marker localization under non-uniform illuminations based on homomorphic filtering | |
CN107525768B (en) | Quality control method of DNA ploid analysis equipment | |
CN108288288A (en) | Accurate shaft size measurement method, the device and system of view-based access control model identification | |
DE112006002674T5 (en) | Methods and apparatus for practical 3D vision system | |
CN112465743B (en) | Periodic structure quality detection method | |
Zhang et al. | Research on tool wear detection based on machine vision in end milling process | |
CN109409290B (en) | Thermometer verification reading automatic identification system and method | |
CN112991287B (en) | Automatic indentation measurement method based on full convolution neural network | |
CN109489554A (en) | A kind of each layer parameter intelligent detecting method of full automatic cable and device | |
CN116091790B (en) | Online detection system for quality of driven disc of electrically-controlled silicone oil clutch | |
CN109507198B (en) | Mask detection system and method based on fast Fourier transform and linear Gaussian | |
CN109359604B (en) | Method for identifying instrument under shadow interference facing inspection robot | |
CN112489042A (en) | Metal product printing defect and surface damage detection method based on super-resolution reconstruction | |
CN108171742A (en) | A kind of method that the evaluation of carbon fibre composite surface topography three-dimensional is realized based on shade of gray feature | |
CN112907556A (en) | Automatic measuring method for abrasion loss of rotary cutter based on machine vision | |
CN115096206A (en) | Part size high-precision measurement method based on machine vision | |
CN102914479A (en) | Automatic Brinell hardness testing method | |
CN110544276A (en) | Least square method ellipse fitting piston skirt maximum point size measurement method | |
CN107220973A (en) | The hollow filter stick quick determination method of hexagon based on Python+OpenCV |
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 |