CN115953397B - Method and equipment for monitoring process preparation flow of conical bearing retainer - Google Patents
Method and equipment for monitoring process preparation flow of conical bearing retainer Download PDFInfo
- Publication number
- CN115953397B CN115953397B CN202310231627.7A CN202310231627A CN115953397B CN 115953397 B CN115953397 B CN 115953397B CN 202310231627 A CN202310231627 A CN 202310231627A CN 115953397 B CN115953397 B CN 115953397B
- Authority
- CN
- China
- Prior art keywords
- image
- conical bearing
- bearing retainer
- window
- workpiece
- 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 94
- 238000002360 preparation method Methods 0.000 title claims abstract description 35
- 238000012544 monitoring process Methods 0.000 title claims abstract description 18
- 230000002159 abnormal effect Effects 0.000 claims abstract description 81
- 239000002184 metal Substances 0.000 claims abstract description 60
- 230000002950 deficient Effects 0.000 claims abstract description 58
- 230000008569 process Effects 0.000 claims abstract description 52
- 238000012545 processing Methods 0.000 claims abstract description 25
- 230000007547 defect Effects 0.000 claims abstract description 18
- 238000001514 detection method Methods 0.000 claims abstract description 12
- 238000000605 extraction Methods 0.000 claims abstract description 8
- 238000003801 milling Methods 0.000 claims description 38
- 238000003708 edge detection Methods 0.000 claims description 23
- 238000005488 sandblasting Methods 0.000 claims description 20
- 230000006870 function Effects 0.000 claims description 16
- 238000004080 punching Methods 0.000 claims description 11
- 238000010330 laser marking Methods 0.000 claims description 9
- 238000003754 machining Methods 0.000 claims description 9
- 239000007769 metal material Substances 0.000 claims description 7
- 241001292396 Cirrhitidae Species 0.000 claims description 6
- 230000036544 posture Effects 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 238000012806 monitoring device Methods 0.000 claims description 5
- 230000004069 differentiation Effects 0.000 claims description 4
- 238000013145 classification model Methods 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 230000005764 inhibitory process Effects 0.000 claims description 3
- 230000001788 irregular Effects 0.000 claims description 3
- 239000000047 product Substances 0.000 description 21
- 238000012797 qualification Methods 0.000 description 9
- 238000004381 surface treatment Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000005096 rolling process Methods 0.000 description 3
- 230000002349 favourable effect Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000003825 pressing Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a monitoring method and equipment for a process preparation flow of a conical bearing retainer, belongs to the technical field of image processing, and is used for solving the technical problems that in the existing process preparation flow of the conical bearing retainer, defective products of each flow link are difficult to identify, the whole process preparation flow is easy to influence, and the yield of finished products is reduced. Comprising the following steps: performing anomaly detection on the edge thickness of the prefabricated metal plate, and then performing metal surface crack feature extraction on the conical bowl-shaped workpiece image to obtain a first residual image; performing image shape matching processing on the initial conical bearing retainer image to determine a second residual image; carrying out surface defect characteristic image recognition on the image of the finished conical bearing retainer, and determining a third residual image; and carrying out abnormal marking on the abnormal workpieces corresponding to the first, second and third residual images to obtain marked workpiece information.
Description
Technical Field
The application relates to the field of image processing and linkage control, in particular to a monitoring method and equipment for a process preparation flow of a conical bearing retainer.
Background
The main function of the conical bearing retainer is to avoid direct contact between the rolling elements, to separate the rolling elements from each other and to guide the rolling elements to roll. Existing conical bearing retainers are typically manufactured by machining or by integral stamping.
The existing conical bearing retainer is easy to cause certain negative effects in the aspects of precision, roundness, cost and the like when a single preparation method is adopted, meanwhile, in the process preparation flow of the conical bearing retainer, real-time quality monitoring is difficult to be carried out on each flow link, so that a problematic workpiece can always operate, the manufacturing cost of the retainer is increased, the finished product qualification rate of the conical bearing retainer is reduced, in the whole process flow, the defective products of each link are difficult to accurately identify due to the rapidness of the process flow, the subsequent process links can be possibly influenced, and unnecessary loss on some process flows is caused.
Disclosure of Invention
The embodiment of the application provides a monitoring method and equipment for a process preparation flow of a conical bearing retainer, which are used for solving the following technical problems: in the existing process preparation flow of the conical bearing retainer, defective products in each flow link are difficult to identify, the whole process preparation flow is easy to be influenced, and the finished product qualification rate of the conical bearing retainer is reduced.
The embodiment of the application adopts the following technical scheme:
in one aspect, an embodiment of the present application provides a method for monitoring a process preparation flow of a conical bearing retainer, including: the method comprises the steps of obtaining an image of the edge thickness of a prefabricated metal plate through a preset area array CCD camera; performing anomaly detection on the edge thickness image to obtain a plate edge detection result of the edge thickness image; wherein the prefabricated metal plate is a metal material to be processed for preparing the conical bearing retainer; based on the plate edge detection result, carrying out metal surface crack feature extraction on the conical bowl-shaped workpiece image which is acquired by a preset camera and is formed by punching, so as to obtain crack image features; classifying and identifying the types of the cracks of the crack image features according to a preset classifying and identifying model to obtain a first residual image; milling holes on the rest conical bowl-shaped workpiece to obtain an initial conical bearing retainer; according to a preset image shape matching template library, performing image shape matching processing on the pre-acquired initial conical bearing retainer image, and determining a second residual image; wherein the remaining conical bowl-shaped workpieces are conical bowl-shaped workpieces without defects; carrying out surface sand blasting on the rest initial conical bearing retainer to obtain a finished conical bearing retainer; carrying out image recognition on the surface defect characteristics of the pre-acquired image of the finished conical bearing retainer, and determining a third residual image; wherein the remaining initial conical bearing retainer is a stub-free initial conical bearing retainer; performing abnormal marking on the abnormal workpieces corresponding to the first defective image, the second defective image and the third defective image to obtain marked workpiece information; and classifying all the workpieces in the conveying belt according to the marked workpiece information to obtain a plurality of finished conical bearing retainer information meeting the standard.
According to the method and the device, the edge thickness of the metal material to be processed of the conical bearing retainer is detected, and the image detection of the defective workpiece is carried out on the unfinished workpiece in the process flows of stretching stamping, edge turning, bottom turning, hole milling, surface treatment and the like in the main process preparation flow, and the unfinished workpiece image obtained through each link in the conical bearing retainer is combined with image recognition and detection of the relevant size and the easily damaged part, so that the defective product of each flow link can be better recognized accurately in the process preparation flow, and meanwhile the recognized defective product is not subjected to the treatment of the subsequent link, so that the defective product always follows the movement of a conveying belt, the whole process flow is smoother, the whole process flow is not suspended halfway, and the influence of the whole process preparation flow is minimized. Meanwhile, the mark recognition of defective products is utilized, so that the finished conical bearing retainer can be better classified and recognized in the remembering way, and the finished product qualification rate of the conical bearing retainer is greatly improved.
In one possible implementation, the image acquisition of the edge thickness of the prefabricated metal plate is performed by presetting an area array CCD camera; and performing anomaly detection on the edge thickness image to obtain a plate edge detection result of the edge thickness image, wherein the plate edge detection result specifically comprises: shooting the side edge area of the prefabricated metal plate at multiple angles through a plurality of area array CCD cameras to obtain the edge thickness image; the edge thickness images are a plurality of edge port images of the prefabricated metal plate; acquiring a plane pixel image related to the thickness of the prefabricated metal plate in the edge port image; carrying out gray processing on the planar pixel image to obtain an initial gray contrast image, and carrying out point location identification on demarcation points with different pixel gray values in the initial gray contrast image through a preset linear fitting function to obtain a demarcation point coordinate set in the initial gray contrast image; according to the coordinate set of the demarcation points, performing point location curve fitting on a plurality of demarcation points to obtain a gray value demarcation curve; judging the curvature of the gray value demarcation curve; if the curvature value of the gray value demarcation curve is a false fraction, the side edge area of the prefabricated metal plate is a head warping area; if the curvature value of the gray value demarcation curve is a true fraction, the side edge area of the prefabricated metal plate is a buckling tail area; determining the head raising area or the tail buckling area as an edge abnormal area, and determining the edge areas of the rest plates as edge normal areas; wherein, the panel edge detection result includes: the edge abnormal region and the edge normal region.
According to the method and the device, the edge thickness of the prefabricated metal plate in the earlier stage preparation process of the conical bearing retainer is identified, the head warping and tail buckling conditions of the metal plate can be accurately detected and identified in the blanking process, subsequent abnormal processing of the edge abnormal areas is avoided, and the yield of the conical bearing retainer is improved.
In a possible implementation manner, based on the plate edge detection result, the metal surface crack feature extraction is performed on the punch-formed conical bowl-shaped workpiece image acquired by the preset camera, so as to obtain a crack image feature, which specifically includes: drawing a metal plate corresponding to an edge normal region in the plate edge detection result by a preset metal hydraulic drawing machine, and punching the drawn metal plate in a conical bowl shape by a preset metal punching machine to obtain a conical bowl-shaped workpiece; photographing the surface of the conical bowl-shaped workpiece through a preset camera to obtain an image of the conical bowl-shaped workpiece; acquiring a bowl wall image of a bowl body area in the conical bowl-shaped workpiece image; wherein, the bowl wall image includes: an inner bowl wall image and an outer bowl wall image; carrying out pixel gray scale processing on the bowl wall image, and carrying out metal surface crack feature extraction of pixels in all directions on the bowl wall image subjected to gray scale processing through Gabor image transformation to obtain texture structure features and abnormal curve features of the bowl wall image; wherein the directions include: transverse, longitudinal and diagonal directions; according to a preset Grouplet kurtosis algorithm, identifying irregular texture features of the bowl wall image to obtain abnormal texture features; according to the coefficient transformation of the association domain in the Grouplet kurtosis, carrying out weighted association on the abnormal texture characteristics to obtain coefficients based on the association domain of the abnormal texture characteristics; and correspondingly associating the coefficients with each pixel point in the abnormal curve characteristic, and determining the crack image characteristic in the bowl wall image.
In a possible implementation manner, according to a preset classification recognition model, classification recognition of the fracture type is performed on the fracture image features to obtain a first residual image, which specifically includes: and carrying out classification, identification and judgment on the crack types in the crack image features through a pretrained KNN classification model: judging the pixel color depth of a gray pixel area corresponding to the crack image features; if a plurality of light areas exist in the gray pixel area, the crack image features are crack image features, and a crack image corresponding to the crack image features is determined to be a first residual image; based on the position of a milling hole window preset in the conical bowl-shaped workpiece, judging the position of a pixel dense region according to the length of a crack pixel and the width of the crack pixel corresponding to the crack image characteristic; if the position of the pixel dense region and the position of the milling hole window are overlapped, and the lengths of the fracture pixels and the widths of the fracture pixels are smaller than or equal to the size of the milling hole window, determining the fracture image corresponding to the fracture image characteristics as a first normal image; wherein, milling hole window size includes: window height and window width; and if the position of the pixel dense region and the position of the milling hole window are not overlapped, or the length of the crack pixel and the width of the crack pixel are larger than the size of the milling hole window, determining the crack image as a first residual image.
According to the method and the device for detecting the metal cracks of the conical bowl-shaped workpiece, through the image recognition of the metal cracks of the conical bowl-shaped workpiece, the qualification condition of the unfinished workpiece can be better monitored, and meanwhile, through treatment on some tiny cracks or tiny cracks which do not affect the subsequent hole milling, edge turning and bottom turning conditions is achieved, the qualification condition of the product is guaranteed to the greatest extent, meanwhile, cost saving is achieved, and the qualification rate of the final product is improved.
In one possible embodiment, the remaining conical bowl-shaped workpiece is subjected to hole milling to obtain an initial conical bearing retainer; and according to a preset image shape matching template library, carrying out image shape matching processing on the pre-acquired initial conical bearing retainer image to determine a second defective image, wherein the method specifically comprises the following steps of: determining a plurality of conical bowl-shaped workpieces corresponding to the first residual images as first residual workpieces, and determining the rest conical bowl-shaped workpieces as normal conical bowl-shaped workpieces according to the identified first residual workpieces; milling holes on the normal conical bowl-shaped workpiece through a preset machining lathe to obtain the initial conical bearing retainer; wherein the first defective work piece is not subjected to hole milling treatment; acquiring an initial conical bearing retainer image and an initial conical bearing retainer image size by a camera preinstalled above the conveyor belt; wherein the initial conical bearing retainer image size comprises: image length and image width; according to a preset rectangular frame, uniformly intercepting a window, a window beam and an end ring in the initial conical bearing retainer image to obtain a first clipping image; judging the gray variance of the window for the first clipping image; the gray variance of the window is the gray pixel range inside the window; if the gray variance is smaller than or equal to a first preset threshold value, the first clipping image is a normal window image; otherwise, the first clipping image is an abnormal window image; obtaining the number of pixels of the window beam in the first clipping image to obtain a window beam pixel value; performing circumferential function fitting on end ring pixel points in the first clipping image to obtain a circumferential function curve based on the end ring pixel points, and determining the end ring circumferential curvature of the function curve; based on different models of the initial conical bearing retainer, acquiring a normal window image set, a window beam pixel value set and an end ring circumferential curvature set which are in one-to-one correspondence with various postures according to the various postures of the initial conical bearing retainer under the standard size, and establishing an image shape matching template library based on the initial conical bearing retainer under the standard size; matching the normal window image, the window beam pixel value and the end ring circumference ratio in the initial conical bearing retainer image with the corresponding standard size in the image shape matching template library to obtain a matching result; and if the matching is consistent, determining the initial conical bearing retainer image as a second normal image, and if the matching is inconsistent or the first clipping image is an abnormal window image, determining the initial conical bearing retainer image as a second defective image.
According to the method and the device for detecting the defective products in the process stage, through monitoring the window, the window beam and the roundness of the initial conical bearing retainer after milling the holes and based on the standard image shape matching template, defective workpieces can be detected rapidly and accurately, and due to the fact that the conveying belt moves rapidly and accurately through the shape matching template, the defective workpieces can be identified simply, conveniently and accurately, then the defective workpieces are not processed in the follow-up process, cost consumption is reduced, and production efficiency is improved.
In a possible implementation manner, matching the normal window image, the window beam pixel value and the end ring circumference ratio in the initial conical bearing retainer image with the corresponding standard sizes in the image shape matching template library to obtain a matching result, which specifically includes: extracting a normal window image set, a window beam pixel value set and an end ring circumference curvature set in the image shape matching template library based on a HALCON algorithm in the image shape matching template library, and determining a plurality of original shape matching templates according to template operators in the HALCON algorithm; generating templates from the normal window image, the window beam pixel values and the end ring circumference ratio in the initial conical bearing retainer image to obtain a target shape matching template; and performing data-level coverage matching on the target shape matching template and the plurality of original shape matching templates, and determining the matching result according to the similarity of data coverage.
In a possible embodiment, the remaining initial conical bearing holder is surface blasted, resulting in a finished conical bearing holder; and performing image recognition of the surface defect characteristics on the pre-acquired image of the finished conical bearing retainer to determine a third residual image, comprising the following specific steps: determining a plurality of initial conical bearing retainers corresponding to the second defective image as second defective workpieces, and determining the rest initial conical bearing retainers as normal initial conical bearing retainers according to the identified second defective workpieces; performing sand blasting on the normal initial conical bearing retainer through a sand blasting machine with preset multiple angles to obtain the finished conical bearing retainer; wherein the second residual workpiece and the first residual workpiece are not subjected to sand blasting; according to multi-angle irradiation of the adjustable light source system, acquiring images of the finish machining conical bearing retainer under multiple angles through a preset camera; gray scale processing is carried out on the image of the finished conical bearing retainer, so that a gray scale image of the finished conical bearing retainer is obtained; carrying out edge information identification on the gray level image through an image edge detection differential operator to obtain edge characteristics; performing non-maximum value inhibition interpolation processing of the gray level map according to the convolution differentiation of the central edge operator of the edge feature and the edge in the gray level map, and obtaining a gray level histogram based on the gray level map based on the gradient distribution characteristic of the Gaussian function; dividing the region of the upper and lower limit threshold values of the gray level histogram according to the preset pixel gray level, and determining the average gray level value of the gray level map; comparing the average gray value with a first preset threshold value to determine the third residual image; wherein the third defective image is a finished conical bearing holder image having surface defect features.
According to the method and the device for identifying the multi-angle surface defect characteristics of the workpiece after sand blasting, the identification of the workpiece with the surface defects, namely the identification of defective workpieces generated in the sand blasting process, can further guarantee the qualification rate of subsequent finished workpieces, meanwhile, hidden defect blind spots can be monitored well based on the identification of the multi-angle sand blasting and the surface defect characteristics, and the identification and judgment can be carried out more accurately and rapidly based on the threshold division of the gray level histogram.
In a possible implementation manner, the method for performing anomaly marking on the abnormal workpiece corresponding to the first defective image, the second defective image and the third defective image to obtain marked workpiece information specifically includes: generating a binary workpiece serial number by using a plurality of abnormal conical bowl-shaped workpieces corresponding to the first residual images, a plurality of abnormal initial conical bearing retainers corresponding to the second residual images and a plurality of abnormal finish conical bearing retainers corresponding to the third residual images to obtain a workpiece marking serial number; generating even number in decimal form corresponding to the workpiece mark number; and marking the abnormal workpiece with the abnormal sequence number according to the even sequence number, and determining the marked workpiece information corresponding to the abnormal workpiece.
According to the method and the device, the problem of logic operation can be better solved by marking the workpiece serial numbers in the binary form, the method and the device are further expanded into digital operation, the method and the device can be better adapted to complex environments of workshops, have stronger interference and reliability, are favorable for primarily marking abnormal workpieces, are then converted into decimal even serial numbers favorable for workers to view, namely, the abnormal serial numbers are formed intuitively and easily understood, and finally, the marked workpiece information of the abnormal workpieces is regenerated.
In a possible implementation manner, according to the marked workpiece information, sorting the finished workpieces of all the workpieces in the conveyor belt to obtain a plurality of finished conical bearing retainer information meeting the standard, specifically including: tracking a plurality of marked workpiece information in real time according to a preset 3D object tracking algorithm, and determining the real-time position of each abnormal workpiece; carrying out laser marking on the real-time position of each abnormal workpiece through a laser guidance system to obtain laser marking workpiece information; and based on a preset time interval, carrying out sorting and removing treatment on the laser marked workpiece information in the conveying belt so as to screen out the corresponding finished workpiece and obtain a plurality of finished conical bearing retainer information meeting the standard.
On the other hand, the embodiment of the application also provides a monitoring device for the process preparation flow of the conical bearing retainer, which comprises the following components: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform a method of monitoring a process preparation flow of a conical bearing holder as described in any one of the embodiments above.
According to the method and the device, the edge thickness of the metal material to be processed of the conical bearing retainer is detected, and the image detection of the defective workpiece is carried out on the unfinished workpiece in the process flows of stretching stamping, edge turning, bottom turning, hole milling, surface treatment and the like in the main process preparation flow, and the unfinished workpiece image obtained through each link in the conical bearing retainer is combined with image recognition and detection of the relevant size and the easily damaged part, so that the defective product of each flow link can be better recognized accurately in the process preparation flow, and meanwhile the recognized defective product is not subjected to the treatment of the subsequent link, so that the defective product always follows the movement of a conveying belt, the whole process flow is smoother, the whole process flow is not suspended halfway, and the influence of the whole process preparation flow is minimized. Meanwhile, the mark recognition of defective products is utilized, so that the finished conical bearing retainer can be better classified and recognized in the remembering way, and the finished product qualification rate of the conical bearing retainer is greatly improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. In the drawings:
fig. 1 is a flowchart of a monitoring method of a process preparation flow of a conical bearing retainer according to an embodiment of the present application;
fig. 2 is a schematic structural view of a monitoring device for a process manufacturing flow of a conical bearing retainer according to an embodiment of the present application.
Detailed Description
In order to better understand the technical solutions in the present application, the following description will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, shall fall within the scope of the present application.
The embodiment of the application provides a method for monitoring a process preparation flow of a conical bearing retainer, as shown in fig. 1, specifically comprising steps S101-S106:
s101, acquiring an image of the edge thickness of the prefabricated metal plate through a preset area array CCD camera. And performing anomaly detection on the edge thickness image to obtain a plate edge detection result of the edge thickness image. Wherein the prefabricated metal plate is a metal material to be processed for preparing the conical bearing retainer.
Specifically, a plurality of area array CCD cameras are used for shooting the side edge area of the prefabricated metal plate at multiple angles to obtain an edge thickness image. The edge thickness images are a plurality of edge port images of the prefabricated metal plate. And obtaining a plane pixel image related to the thickness of the prefabricated metal plate in the edge port image. Carrying out gray processing on the planar pixel image to obtain an initial gray contrast image, and carrying out point location identification on a plurality of demarcation points with different pixel gray values in the initial gray contrast image by presetting a linear fitting function to obtain a demarcation point coordinate set in the initial gray contrast image.
Further, according to the coordinate set of the demarcation points, fitting the point location curve of the demarcation points to obtain a gray value demarcation curve. And judging the curvature of the gray value demarcation curve. And if the curvature value of the gray value demarcation curve is a false fraction, the side edge area of the prefabricated metal plate is a warp area. And if the curvature value of the gray value demarcation curve is a true fraction, the side edge area of the prefabricated metal plate is a buckling tail area. And determining the head raising area or the tail buckling area as an edge abnormal area, and determining the edge areas of the rest plates as edge normal areas.
Wherein, panel edge detection result includes: an edge abnormal region and an edge normal region.
In one embodiment, a plurality of area array CCD cameras in a laser blanking machine are used for detecting curves of the edge thickness of a metal plate, point location identification is carried out on a plurality of demarcation points with different pixel gray values in an image according to an obtained planar pixel image related to the thickness of the metal plate, coordinate sets of the different pixel demarcation points are obtained, a gray value demarcation curve is constructed, whether the edge of the metal plate has a head warping or tail buckling condition or not is obtained according to the curvature value of the curve, and an edge image area corresponding to the abnormal curvature values is determined to be an edge abnormal area.
S102, based on a plate edge detection result, extracting the metal surface crack characteristics of the conical bowl-shaped workpiece image which is acquired by a preset camera and formed by punching, and obtaining crack image characteristics. And carrying out classification recognition of the fracture types on the fracture image features according to a preset classification recognition model to obtain a first residual image.
Specifically, a metal plate corresponding to an edge normal region in a plate edge detection result is stretched through a preset metal hydraulic stretching machine, and a conical bowl-shaped workpiece is obtained by punching the stretched metal plate through a preset metal punching machine. And photographing the surface of the conical bowl-shaped workpiece through a preset camera to obtain a conical bowl-shaped workpiece image. And acquiring a bowl wall image of a bowl body area in the conical bowl-shaped workpiece image. Wherein, bowl wall image includes: an inner bowl wall image and an outer bowl wall image. And carrying out pixel gray scale processing on the bowl wall image, and carrying out metal surface crack characteristic extraction of pixels in all directions on the bowl wall image subjected to gray scale processing through Gabor image transformation to obtain texture structure characteristics and abnormal curve characteristics of the bowl wall image. Wherein, each direction includes: transverse, longitudinal, and diagonal directions.
Further, according to a preset Grouplet kurtosis algorithm, irregular texture features are identified for the texture features of the bowl wall image, and abnormal texture features are obtained. And carrying out weighted association on the abnormal texture features according to the coefficient transformation of the association domain in the Grouplet kurtosis to obtain the coefficients based on the association domain of the abnormal texture features. And correspondingly associating the coefficients with each pixel point in the abnormal curve characteristics to determine the crack image characteristics in the bowl wall image.
Further, classifying, identifying and judging the types of the cracks in the crack image features through a pre-trained KNN classifying model: and judging the pixel color depth of the gray pixel area corresponding to the crack image characteristics. If a plurality of light areas exist in the gray pixel area, the crack image features are crack image features, and the crack image corresponding to the crack image features is determined to be a first residual image.
Based on the position of a milling hole window preset in the conical bowl-shaped workpiece, judging the position of a pixel dense region by using the length of a crack pixel and the width of the crack pixel corresponding to the crack image characteristics:
if the position of the pixel dense region and the position of the milling hole window are in an overlapping state, and the length of the crack pixels and the width of the crack pixels are smaller than or equal to the size of the milling hole window, determining a crack image corresponding to the crack image characteristics as a first normal image. Wherein, milling hole window size includes: window height and window width.
If the position of the pixel dense region and the position of the milling hole window are not overlapped, or the length of the crack pixel and the width of the crack pixel are larger than the size of the milling hole window, determining the crack image as a first residual image.
In one embodiment, a conical bowl-shaped workpiece is obtained by punching a stretched metal plate through a metal hydraulic stretcher, a conical bowl-shaped workpiece image is obtained by photographing the conical bowl-shaped workpiece through a preset camera, then the type of metal surface crack characteristics of the obtained metal surface images corresponding to the inner bowl wall image and the outer bowl wall image are identified to judge different crack image characteristics, then the gray pixel area corresponding to the crack image characteristics is subjected to pixel color depth judgment based on a KNN classification model, whether a crack or a split image exists in the crack image is determined, the images are determined to be a first residual image, namely a defective workpiece, then the pixel length of the crack and the position of a milling hole window are judged, so that whether the metal crack belongs to an irrelevant crack is judged, if the process of a subsequent milling hole is not affected, the crack image is judged to be a normal image, the subsequent process is continued, if the subsequent hole milling process is affected, the crack image is judged to be a metal crack image which cannot be continuously added, and the first residual image is determined.
S103, milling holes on the rest conical bowl-shaped workpiece to obtain the initial conical bearing retainer. And performing image shape matching processing on the pre-acquired initial conical bearing retainer image according to a preset image shape matching template library, and determining a second residual image. Wherein the remaining conical bowl-shaped workpieces are conical bowl-shaped workpieces without defects.
Specifically, a plurality of conical bowl-shaped workpieces corresponding to the first residual images are determined to be first residual workpieces, and the remaining conical bowl-shaped workpieces are determined to be normal conical bowl-shaped workpieces according to the identified first residual workpieces. And (3) carrying out hole milling treatment on the normal conical bowl-shaped workpiece through a preset machining lathe to obtain the initial conical bearing retainer. Wherein, the first defective work piece is not subjected to hole milling treatment. An initial cone bearing retainer image and an initial cone bearing retainer image size are acquired by a camera pre-mounted above the conveyor belt. Wherein the initial conical bearing retainer image size comprises: image length and image width.
Further, according to a preset rectangular frame, unified image interception is carried out on the window, the window beam and the end ring in the initial conical bearing retainer image, and a first clipping image is obtained. And judging the gray variance of the window for the first clipping image. Wherein the gray variance of the window is the gray pixel range inside the window. And if the gray variance is smaller than or equal to a first preset threshold value, the first clipping image is a normal window image. Otherwise, the first clipping image is an abnormal window image. And acquiring the pixel number of the window beam in the first clipping image to obtain a window beam pixel value. Performing circumferential function fitting on the end ring pixel points in the first clipping image to obtain a circumferential function curve based on the end ring pixel points, and determining the end ring circumferential curvature of the function curve.
Further, based on different models of the initial conical bearing retainer, according to various postures of the initial conical bearing retainer under the standard size, a normal window image set, a window beam pixel value set and an end ring circumferential curvature set which are in one-to-one correspondence with the various postures are obtained, and an image shape matching template library based on the initial conical bearing retainer under the standard size is established.
And further, matching the normal window image, the window beam pixel value and the end ring circumference ratio in the initial conical bearing retainer image with the corresponding standard size in the image shape matching template library to obtain a matching result. If the matching is consistent, the initial conical bearing retainer image is determined to be a second normal image, and if the matching is inconsistent or the first clipping image is an abnormal window image, the initial conical bearing retainer image is determined to be a second defective image.
The method comprises the steps of extracting a normal window image set, a window beam pixel value set and an end ring circumference curvature set from an image shape matching template library based on a HALCON algorithm in the image shape matching template library, and determining a plurality of original shape matching templates according to template operators in the HALCON algorithm. And generating templates by using the normal window image, the window beam pixel value and the end ring circumference ratio in the initial conical bearing retainer image to obtain a target shape matching template. And performing coverage matching on the data layer by using the target shape matching template and a plurality of original shape matching templates, and determining a matching result according to the similarity of data coverage.
In one embodiment, the first cropped image comprises: step 1: inputting an initial conical bearing retainer image I0 with the size of
. Step 2: with a size of +.>
The initial conical bearing holder image I0 is cut out, and the rectangular frame length direction includes a window a, a window beam b, and an end ring c in the initial conical bearing holder image. Step 3: the upper left corner of the initial conical bearing retainer image is taken as an original point, the width of the image is taken as an abscissa, and the length is taken as an ordinate. Setting the initial ordinate of the top end of the rectangular frame as S, wherein the S generally meets the following conditions: s= (W0-W1)/2, and the gray variance of the gray pixel range inside the window at the ordinate is calculated. Step 4: judging whether the image belongs to normal clipping (namely whether the gray pixel value in the window meets a preset threshold value) according to the gray variance, performing the next operation, and if the gray variance is smaller than or equal to a first preset threshold value, the first clipping image is a normal window image. Otherwise, the first clipping image is an abnormal window image.
In one embodiment, the machining lathe is used for carrying out the operations of turning edges, turning bottoms and milling holes on the normal conical bowl-shaped workpiece identified in the previous working procedure, the conical bowl-shaped workpiece corresponding to the first residual image is not subjected to any operation, the machining lathe continues to move along with the conveying belt, the machining lathe only carries out subsequent processing on the normal conical bowl-shaped workpiece, then obtains an initial conical bearing retainer, photographs the initial conical bearing retainer according to a preset monitoring camera, acquires an initial conical bearing retainer image, then carries out image interception on the initial conical bearing retainer image according to a preset rectangular frame, then carries out gray variance judgment of a window, judges whether the window size in the initial conical bearing retainer image meets preset standards, then acquires the window Liang Xiangsu value and the end ring circumferential curvature, carries out data-layer coverage matching on the normal window image, the window beam pixel value and the end ring circumferential rate in the initial conical bearing retainer image and a plurality of original shape matching templates based on a template, determines the matching degree according to the similarity of data which can be mutually covered, then cuts the initial conical bearing retainer image or determines that the first conical bearing retainer image is inconsistent or the first bearing retainer image is the abnormal, and determines that the first bearing retainer image is the corresponding to the first bearing image is the second image.
In one embodiment, according to conical bearing retainers of different types, slope pressing treatment is further performed on the window beam at the initial conical bearing retainers, then an initial conical bearing retainer image after the slope pressing treatment is acquired again, a pixel coordinate system corresponding to slope inclination in the image is identified and judged, the initial conical bearing retainer image which does not meet the slope inclination requirement is also judged to be a second residual image, the initial conical bearing retainers corresponding to all the qualified initial conical bearing retainer images continue to perform subsequent surface treatment procedures, and the initial conical bearing retainers corresponding to the second residual image do not participate in the surface sand blasting treatment procedure, move along the conveying belt until abnormal workpieces are judged to be classified.
And S104, carrying out surface sand blasting on the rest initial conical bearing retainers to obtain the finished conical bearing retainers. And performing image recognition on the pre-acquired image of the finished conical bearing retainer on the surface defect characteristics to determine a third residual image. Wherein the remaining initial conical bearing holder is an initial conical bearing holder without residues.
Specifically, a plurality of initial cone bearing retainers corresponding to the second defective image are determined as second defective work pieces, and the remaining initial cone bearing retainers are determined as normal initial cone bearing retainers according to the identified second defective work pieces. And (3) carrying out sand blasting on the normal initial conical bearing retainer by a sand blasting machine with preset multiple angles to obtain the finished conical bearing retainer. Wherein the second residual workpiece and the first residual workpiece are not subjected to sand blasting.
Further, according to the multi-angle irradiation of the adjustable light source system, the image of the finished conical bearing retainer under the multi-angle is acquired through a preset camera. And carrying out gray scale treatment on the image of the finished conical bearing retainer to obtain a gray scale image of the finished conical bearing retainer. And identifying the edge information of the gray level image by using an image edge detection differential operator to obtain edge characteristics.
Further, according to the convolution differentiation of the central edge operator of the edge characteristic and the edge in the gray level diagram, non-maximum value inhibition interpolation processing of the gray level diagram is carried out, and a gray level histogram based on the gray level diagram is obtained based on the gradient distribution characteristic of the Gaussian function.
Further, according to the preset pixel gray level, the region division of the upper and lower threshold values is carried out on the gray level histogram, and the average gray value of the gray level map is determined. And comparing the average gray value with a first preset threshold value to determine a third residual image. Wherein the third defective image is a finished conical bearing retainer image having surface defect features.
In one embodiment, before the initial conical bearing retainer is sandblasted, the window hole is optionally expanded, then sandblasted, then an image edge detection differential operator is performed on the finished conical bearing retainer image, the edge characteristics of the gray level image are identified, the region division of the upper and lower limit threshold values is performed on the gray level histogram according to the preset pixel gray level, the average gray level value of the gray level image is determined, when the average gray level value is greater than the first preset threshold value, the existence of the surface defect characteristic of the finished conical bearing retainer image can be judged, and the finished conical bearing retainer image is determined to be a third residual image, otherwise, the normal image of the workpiece is determined. And (3) selecting secondary sand blasting treatment for the finished conical bearing retainer corresponding to the third residual image, then continuing to identify the image of the related surface defect characteristic according to the convolution differentiation of the central edge operator of the edge characteristic and the edge in the gray scale image, and automatically ending the sand blasting treatment if the image is still the third residual image, and carrying out subsequent procedures.
S105, carrying out abnormal marking on the abnormal workpiece corresponding to the first residual image, the second residual image and the third residual image to obtain marked workpiece information.
Specifically, generating a binary workpiece serial number by using a plurality of abnormal conical bowl-shaped workpieces corresponding to the first residual images, a plurality of abnormal initial conical bearing retainers corresponding to the second residual images and a plurality of abnormal finish conical bearing retainers corresponding to the third residual images to obtain workpiece marking serial numbers. An even number in decimal form corresponding to the workpiece marking number is generated.
Further, according to the even number, the abnormal workpiece is marked with the abnormal number, and marked workpiece information corresponding to the abnormal workpiece is determined.
In one embodiment, after the sand blasting treatment, a plurality of abnormal conical bowl-shaped workpieces corresponding to the first defective image, a plurality of abnormal initial conical bearing holders corresponding to the second defective image and a plurality of abnormal finish conical bearing holders corresponding to the third defective image are reserved in the conveyor belt, and collectively referred to as abnormal workpieces, and then the abnormal workpieces are subjected to binary form workpiece serial number generation treatment, and then converted into binary form workpiece serial number generation treatment which is easy for a worker to check, and finally marked workpiece information corresponding to the abnormal workpieces is determined.
S106, classifying all the workpieces in the conveying belt according to the marked workpiece information to obtain a plurality of finished conical bearing retainer information meeting the standard.
Specifically, according to a preset 3D object tracking algorithm, the information of a plurality of marked workpieces is tracked in real time, and the real-time position of each abnormal workpiece is determined. And carrying out laser marking on the real-time position of each abnormal workpiece through a laser guidance system to obtain laser marking workpiece information.
Further, based on a preset time interval, sorting and eliminating the laser marked workpiece information in the conveying belt to screen out corresponding finished workpieces and obtain a plurality of finished conical bearing retainer information meeting the standard.
In one embodiment, the laser guidance system performs laser marking on the abnormal workpieces and generates corresponding laser marking workpiece information, then a worker can screen each abnormal workpiece according to the laser marking workpiece information, and only a plurality of finished conical bearing retainer information meeting the standard is reserved in the final conveying belt and the corresponding system.
In addition, the embodiment of the application further provides a monitoring device for a process preparation flow of the conical bearing retainer, as shown in fig. 2, the monitoring device 200 for the process preparation flow of the conical bearing retainer specifically includes:
At least one processor 201. And a memory 202 communicatively coupled to the at least one processor 201; wherein the memory 202 stores instructions executable by the at least one processor 201 to enable the at least one processor 201 to perform:
the method comprises the steps of obtaining an image of the edge thickness of a prefabricated metal plate through a preset area array CCD camera; performing anomaly detection on the edge thickness image to obtain a plate edge detection result of the edge thickness image; wherein the prefabricated metal plate is a metal material to be processed for preparing the conical bearing retainer;
based on a plate edge detection result, carrying out metal surface crack feature extraction on the conical bowl-shaped workpiece image which is acquired by a preset camera and is formed by punching, so as to obtain crack image features; classifying and identifying the types of the cracks of the crack image features according to a preset classifying and identifying model to obtain a first residual image;
milling holes on the rest conical bowl-shaped workpiece to obtain an initial conical bearing retainer; according to a preset image shape matching template library, performing image shape matching processing on the pre-acquired initial conical bearing retainer image, and determining a second residual image; wherein the rest conical bowl-shaped workpieces are conical bowl-shaped workpieces without defectiveness;
Carrying out surface sand blasting on the rest initial conical bearing retainer to obtain a finished conical bearing retainer; carrying out image recognition on the surface defect characteristics of the pre-acquired image of the finished conical bearing retainer, and determining a third residual image; wherein the remaining initial conical bearing retainer is an initial conical bearing retainer without residue;
carrying out abnormal marking on the abnormal workpieces corresponding to the first residual image, the second residual image and the third residual image to obtain marked workpiece information;
and classifying all the workpieces in the conveying belt according to the marked workpiece information to obtain a plurality of finished conical bearing retainer information meeting the standard.
According to the method and the device, the edge thickness of the metal material to be processed of the conical bearing retainer is detected, and the image detection of the defective workpiece is carried out on the unfinished workpiece in the process flows of stretching stamping, edge turning, bottom turning, hole milling, surface treatment and the like in the main process preparation flow, and the unfinished workpiece image obtained through each link in the conical bearing retainer is combined with image recognition and detection of the relevant size and the easily damaged part, so that the defective product of each flow link can be better recognized accurately in the process preparation flow, and meanwhile the recognized defective product is not subjected to the treatment of the subsequent link, so that the defective product always follows the movement of a conveying belt, the whole process flow is smoother, the whole process flow is not suspended halfway, and the influence of the whole process preparation flow is minimized. Meanwhile, the mark recognition of defective products is utilized, so that the finished conical bearing retainer can be better classified and recognized in the remembering way, and the finished product qualification rate of the conical bearing retainer is greatly improved.
All embodiments in the application are described in a progressive manner, and identical and similar parts of all embodiments are mutually referred, so that each embodiment mainly describes differences from other embodiments. In particular, for the apparatus embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference is made to the description of the method embodiments in part.
The foregoing describes specific embodiments of the present application. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the embodiments of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of the present application should be included in the scope of the claims of the present application.
Claims (6)
1. A method of monitoring a process preparation flow of a conical bearing retainer, the method comprising:
the method comprises the steps of obtaining an image of the edge thickness of a prefabricated metal plate through a preset area array CCD camera; and performing anomaly detection on the edge thickness image to obtain a plate edge detection result of the edge thickness image, wherein the plate edge detection result specifically comprises: shooting the side edge area of the prefabricated metal plate at multiple angles through a plurality of area array CCD cameras to obtain the edge thickness image; the edge thickness images are a plurality of edge port images of the prefabricated metal plate;
acquiring a plane pixel image related to the thickness of the prefabricated metal plate in the edge port image;
carrying out gray processing on the planar pixel image to obtain an initial gray contrast image, and carrying out point location identification on demarcation points with different pixel gray values in the initial gray contrast image through a preset linear fitting function to obtain a demarcation point coordinate set in the initial gray contrast image;
according to the coordinate set of the demarcation points, performing point location curve fitting on a plurality of demarcation points to obtain a gray value demarcation curve; judging the curvature of the gray value demarcation curve;
If the curvature value of the gray value demarcation curve is a false fraction, the side edge area of the prefabricated metal plate is a head warping area; if the curvature value of the gray value demarcation curve is a true fraction, the side edge area of the prefabricated metal plate is a buckling tail area;
determining the head raising area or the tail buckling area as an edge abnormal area, and determining the edge areas of the rest plates as edge normal areas;
wherein, the panel edge detection result includes: the edge abnormal region and the edge normal region; the prefabricated metal plate is a metal material to be processed for preparing the conical bearing retainer;
based on the plate edge detection result, carrying out metal surface crack feature extraction on the conical bowl-shaped workpiece image which is acquired by a preset camera and is formed by punching, so as to obtain crack image features; and according to a preset classification and identification model, classifying and identifying the types of the cracks of the crack image features to obtain a first residual image, which specifically comprises the following steps: classifying, identifying and judging the types of the cracks in the crack image features through a pre-trained KNN classification model;
judging the pixel color depth of a gray pixel area corresponding to the crack image features; if a plurality of light areas exist in the gray pixel area, the crack image features are crack image features, and a crack image corresponding to the crack image features is determined to be a first residual image;
Based on the position of a milling hole window preset in the conical bowl-shaped workpiece, judging the position of a pixel dense region according to the length of a crack pixel and the width of the crack pixel corresponding to the crack image characteristic;
if the position of the pixel dense region and the position of the milling hole window are overlapped, and the lengths of the fracture pixels and the widths of the fracture pixels are smaller than or equal to the size of the milling hole window, determining the fracture image corresponding to the fracture image characteristics as a first normal image; wherein, milling hole window size includes: window height and window width;
if the position of the pixel dense region and the position of the milling window are not overlapped, or the length of the crack pixel and the width of the crack pixel are larger than the size of the milling window, determining the crack image as a first residual image;
milling holes on the rest conical bowl-shaped workpiece to obtain an initial conical bearing retainer; and according to a preset image shape matching template library, carrying out image shape matching processing on the pre-acquired initial conical bearing retainer image to determine a second defective image, wherein the method specifically comprises the following steps of: determining a plurality of conical bowl-shaped workpieces corresponding to the first residual images as first residual workpieces, and determining the rest conical bowl-shaped workpieces as normal conical bowl-shaped workpieces according to the identified first residual workpieces;
Milling holes on the normal conical bowl-shaped workpiece through a preset machining lathe to obtain the initial conical bearing retainer; wherein the first defective work piece is not subjected to hole milling treatment;
acquiring an initial conical bearing retainer image and an initial conical bearing retainer image size by a camera preinstalled above the conveyor belt; wherein the initial conical bearing retainer image size comprises: image length and image width;
according to a preset rectangular frame, uniformly intercepting a window, a window beam and an end ring in the initial conical bearing retainer image to obtain a first clipping image; judging the gray variance of the window for the first clipping image; the gray variance of the window is the gray pixel range inside the window;
if the gray variance is smaller than or equal to a first preset threshold value, the first clipping image is a normal window image; otherwise, the first clipping image is an abnormal window image;
obtaining the number of pixels of the window beam in the first clipping image to obtain a window beam pixel value;
performing circumferential function fitting on end ring pixel points in the first clipping image to obtain a circumferential function curve based on the end ring pixel points, and determining the end ring circumferential curvature of the function curve;
Based on different models of the initial conical bearing retainer, acquiring a normal window image set, a window beam pixel value set and an end ring circumferential curvature set which are in one-to-one correspondence with various postures according to the various postures of the initial conical bearing retainer under the standard size, and establishing an image shape matching template library based on the initial conical bearing retainer under the standard size;
matching the normal window image, the window beam pixel value and the end ring circumference ratio in the initial conical bearing retainer image with the corresponding standard size in the image shape matching template library to obtain a matching result; if the matching is consistent, determining the initial conical bearing retainer image as a second normal image, and if the matching is inconsistent or the first clipping image is an abnormal window image, determining the initial conical bearing retainer image as a second residual image; wherein the remaining conical bowl-shaped workpieces are conical bowl-shaped workpieces without defects;
carrying out surface sand blasting on the rest initial conical bearing retainer to obtain a finished conical bearing retainer; and performing image recognition of the surface defect characteristics on the pre-acquired image of the finished conical bearing retainer to determine a third residual image, comprising the following specific steps: determining a plurality of initial conical bearing retainers corresponding to the second defective image as second defective workpieces, and determining the rest initial conical bearing retainers as normal initial conical bearing retainers according to the identified second defective workpieces;
Performing sand blasting on the normal initial conical bearing retainer through a sand blasting machine with preset multiple angles to obtain the finished conical bearing retainer; wherein the second residual workpiece and the first residual workpiece are not subjected to sand blasting;
according to multi-angle irradiation of the adjustable light source system, acquiring images of the finish machining conical bearing retainer under multiple angles through a preset camera; gray scale processing is carried out on the image of the finished conical bearing retainer, so that a gray scale image of the finished conical bearing retainer is obtained;
carrying out edge information identification on the gray level image through an image edge detection differential operator to obtain edge characteristics;
performing non-maximum value inhibition interpolation processing of the gray level map according to the convolution differentiation of the central edge operator of the edge feature and the edge in the gray level map, and obtaining a gray level histogram based on the gray level map based on the gradient distribution characteristic of the Gaussian function;
dividing the region of the upper and lower limit threshold values of the gray level histogram according to the preset pixel gray level, and determining the average gray level value of the gray level map;
comparing the average gray value with a first preset threshold value to determine the third residual image; wherein the third defective image is a finished conical bearing holder image having surface defect features; the remaining initial conical bearing retainer is an initial conical bearing retainer without residues;
Performing abnormal marking on the abnormal workpieces corresponding to the first defective image, the second defective image and the third defective image to obtain marked workpiece information;
and classifying all the workpieces in the conveying belt according to the marked workpiece information to obtain a plurality of finished conical bearing retainer information meeting the standard.
2. The method for monitoring the process preparation flow of the conical bearing retainer according to claim 1, wherein the step of extracting the metal surface crack characteristics from the stamped conical bowl-shaped workpiece image acquired by a preset camera based on the plate edge detection result, and obtaining crack image characteristics specifically comprises the following steps:
drawing a metal plate corresponding to an edge normal region in the plate edge detection result by a preset metal hydraulic drawing machine, and punching the drawn metal plate in a conical bowl shape by a preset metal punching machine to obtain a conical bowl-shaped workpiece; photographing the surface of the conical bowl-shaped workpiece through a preset camera to obtain an image of the conical bowl-shaped workpiece;
acquiring a bowl wall image of a bowl body area in the conical bowl-shaped workpiece image; wherein, the bowl wall image includes: an inner bowl wall image and an outer bowl wall image;
Carrying out pixel gray scale processing on the bowl wall image, and carrying out metal surface crack feature extraction of pixels in all directions on the bowl wall image subjected to gray scale processing through Gabor image transformation to obtain texture structure features and abnormal curve features of the bowl wall image; wherein the directions include: transverse, longitudinal and diagonal directions;
according to a preset Grouplet kurtosis algorithm, identifying irregular texture features of the bowl wall image to obtain abnormal texture features; according to the coefficient transformation of the association domain in the Grouplet kurtosis, carrying out weighted association on the abnormal texture characteristics to obtain coefficients based on the association domain of the abnormal texture characteristics;
and correspondingly associating the coefficients with each pixel point in the abnormal curve characteristic, and determining the crack image characteristic in the bowl wall image.
3. The method for monitoring the process preparation flow of the conical bearing retainer according to claim 1, wherein the matching of the shape templates is performed on the normal window image, the window beam pixel value and the end ring circumference ratio in the initial conical bearing retainer image and the corresponding standard size in the image shape matching template library to obtain a matching result, specifically comprising the following steps:
Extracting a normal window image set, a window beam pixel value set and an end ring circumference curvature set in the image shape matching template library based on a HALCON algorithm in the image shape matching template library, and determining a plurality of original shape matching templates according to template operators in the HALCON algorithm;
generating templates from the normal window image, the window beam pixel values and the end ring circumference ratio in the initial conical bearing retainer image to obtain a target shape matching template;
and performing data-level coverage matching on the target shape matching template and the plurality of original shape matching templates, and determining the matching result according to the similarity of data coverage.
4. The method for monitoring the process preparation flow of the conical bearing retainer according to claim 1, wherein the abnormal workpieces corresponding to the first defective image, the second defective image and the third defective image are marked abnormally to obtain marked workpiece information, and the method specifically comprises the steps of:
generating a binary workpiece serial number by using a plurality of abnormal conical bowl-shaped workpieces corresponding to the first residual images, a plurality of abnormal initial conical bearing retainers corresponding to the second residual images and a plurality of abnormal finish conical bearing retainers corresponding to the third residual images to obtain a workpiece marking serial number;
Generating even number in decimal form corresponding to the workpiece mark number;
and marking the abnormal workpiece with the abnormal sequence number according to the even sequence number, and determining the marked workpiece information corresponding to the abnormal workpiece.
5. The method for monitoring the process preparation flow of the conical bearing retainer according to claim 1, wherein the classifying of finished products is performed on all the workpieces in the conveyor belt according to the marked workpiece information to obtain a plurality of finished conical bearing retainer information meeting the standard, and the method specifically comprises the following steps:
tracking a plurality of marked workpiece information in real time according to a preset 3D object tracking algorithm, and determining the real-time position of each abnormal workpiece;
carrying out laser marking on the real-time position of each abnormal workpiece through a laser guidance system to obtain laser marking workpiece information;
and based on a preset time interval, carrying out sorting and removing treatment on the laser marked workpiece information in the conveying belt so as to screen out the corresponding finished workpiece and obtain a plurality of finished conical bearing retainer information meeting the standard.
6. A monitoring device for a process preparation flow of a conical bearing holder, characterized in that the device comprises:
At least one processor; the method comprises the steps of,
a memory communicatively coupled to the at least one processor; wherein,,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform a method of monitoring a process preparation flow of a conical bearing holder according to any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310231627.7A CN115953397B (en) | 2023-03-13 | 2023-03-13 | Method and equipment for monitoring process preparation flow of conical bearing retainer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310231627.7A CN115953397B (en) | 2023-03-13 | 2023-03-13 | Method and equipment for monitoring process preparation flow of conical bearing retainer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115953397A CN115953397A (en) | 2023-04-11 |
| CN115953397B true CN115953397B (en) | 2023-06-02 |
Family
ID=85896310
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310231627.7A Active CN115953397B (en) | 2023-03-13 | 2023-03-13 | Method and equipment for monitoring process preparation flow of conical bearing retainer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115953397B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116109633B (en) * | 2023-04-12 | 2023-06-23 | 山东金帝精密机械科技股份有限公司 | Window detection method and device for bearing retainer |
| CN116586925B (en) * | 2023-07-19 | 2023-09-19 | 山东金帝精密机械科技股份有限公司 | Large-scale bearing retainer production method, equipment and medium based on images |
| CN116665138B (en) * | 2023-08-01 | 2023-11-07 | 临朐弘泰汽车配件有限公司 | Visual detection method and system for stamping processing of automobile parts |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115359053A (en) * | 2022-10-19 | 2022-11-18 | 江苏万森绿建装配式建筑有限公司 | Intelligent detection method and system for defects of metal plate |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111724126B (en) * | 2020-06-12 | 2024-03-12 | 北京科技大学顺德研究生院 | Accurate tracing method and system for quality abnormality of process flow |
| KR102579783B1 (en) * | 2020-09-28 | 2023-09-18 | (주)미래융합정보기술 | Vision inspection system by using remote learning of product defects image |
| CN112837302B (en) * | 2021-02-09 | 2024-02-13 | 广东拓斯达科技股份有限公司 | Method and device for monitoring state of die, industrial personal computer, storage medium and system |
| CN113591790A (en) * | 2021-08-16 | 2021-11-02 | 上海铂端科技有限公司 | System, method and device for realizing production line assembly flow behavior monitoring based on computer vision, processor and storage medium thereof |
| CN115393363B (en) * | 2022-10-31 | 2023-03-17 | 山东金帝精密机械科技股份有限公司 | Production early warning method, equipment and medium for bearing retainer |
| CN115493843B (en) * | 2022-11-18 | 2023-03-10 | 聊城市义和轴承配件有限公司 | Quality monitoring method and equipment based on bearing retainer |
-
2023
- 2023-03-13 CN CN202310231627.7A patent/CN115953397B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115359053A (en) * | 2022-10-19 | 2022-11-18 | 江苏万森绿建装配式建筑有限公司 | Intelligent detection method and system for defects of metal plate |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115953397A (en) | 2023-04-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115953397B (en) | Method and equipment for monitoring process preparation flow of conical bearing retainer | |
| CN110163853B (en) | Edge defect detection method | |
| CN109685760B (en) | MATLAB-based SLM powder bed powder laying image convex hull depression defect detection method | |
| CN103649989B (en) | The analysis of the digital picture of outer surface of tire and the process of pseudo-measurement point | |
| CN115880248B (en) | Surface scratch defect identification method and visual detection equipment | |
| CN106780526A (en) | A kind of ferrite wafer alligatoring recognition methods | |
| CN109767445B (en) | High-precision PCB defect intelligent detection method | |
| CN108647706B (en) | Article identification classification and flaw detection method based on machine vision | |
| CN111127402A (en) | Visual detection method for welding quality of robot | |
| CN106355590B (en) | Mold residue visual detection method and device based on image difference making | |
| CN111539927B (en) | Detection method of automobile plastic assembly fastening buckle missing detection device | |
| CN111402238A (en) | Defect identification system realized through machine vision | |
| CN115018846B (en) | AI intelligent camera-based multi-target crack defect detection method and device | |
| CN113077437B (en) | Workpiece quality detection method and system | |
| CN108844961A (en) | A kind of temperature controller case vision detection system and method | |
| CN112102278A (en) | Metal workpiece machining surface defect detection method based on computer vision | |
| CN116128873A (en) | Bearing retainer detection method, device and medium based on image recognition | |
| CN115239728A (en) | Fire-fighting equipment identification method | |
| CN108986160A (en) | A kind of image laser center line extraction method containing specular light interference | |
| CN117808799A (en) | Chamfering equipment processing quality detection method based on artificial intelligence | |
| CN114418935A (en) | Detection and identification method for mobile phone TP frame screw hole and screw locking device | |
| CN113139943A (en) | Method and system for detecting appearance defects of open circular ring workpiece and computer storage medium | |
| CN113758439A (en) | Method and device for measuring geometric parameters on line in hot ring rolling forming process | |
| CN116681664B (en) | Detection method and device for operation of stamping equipment | |
| CN107121063A (en) | The method for detecting workpiece |
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 |