CN116128873A - Bearing retainer detection method, device and medium based on image recognition - Google Patents
Bearing retainer detection method, device and medium based on image recognition Download PDFInfo
- Publication number
- CN116128873A CN116128873A CN202310348401.5A CN202310348401A CN116128873A CN 116128873 A CN116128873 A CN 116128873A CN 202310348401 A CN202310348401 A CN 202310348401A CN 116128873 A CN116128873 A CN 116128873A
- Authority
- CN
- China
- Prior art keywords
- image
- bearing retainer
- detected
- determining
- window
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 38
- 238000001914 filtration Methods 0.000 claims abstract description 65
- 230000003746 surface roughness Effects 0.000 claims abstract description 41
- 230000008859 change Effects 0.000 claims abstract description 37
- 230000007547 defect Effects 0.000 claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 23
- 230000006870 function Effects 0.000 claims description 22
- 238000012806 monitoring device Methods 0.000 claims description 20
- 238000003709 image segmentation Methods 0.000 claims description 16
- 238000011084 recovery Methods 0.000 claims description 8
- 239000000654 additive Substances 0.000 claims description 7
- 230000000996 additive effect Effects 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 4
- 230000003595 spectral effect Effects 0.000 claims description 3
- 230000001131 transforming effect Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 230000009466 transformation Effects 0.000 claims description 2
- 238000004590 computer program Methods 0.000 claims 2
- 238000007689 inspection Methods 0.000 claims 1
- 238000012545 processing Methods 0.000 abstract description 5
- 230000011218 segmentation Effects 0.000 abstract description 2
- 238000004519 manufacturing process Methods 0.000 description 8
- 238000005096 rolling process Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 6
- 238000009826 distribution Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000002372 labelling Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 208000025962 Crush injury Diseases 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- JEIPFZHSYJVQDO-UHFFFAOYSA-N iron(III) oxide Inorganic materials O=[Fe]O[Fe]=O JEIPFZHSYJVQDO-UHFFFAOYSA-N 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000003801 milling Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000004381 surface treatment Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration by the use of local operators
-
- G06T5/70—
-
- G06T5/77—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20056—Discrete and fast Fourier transform, [DFT, FFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
-
- 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
Abstract
The embodiment of the application discloses a bearing retainer detection method, equipment and medium based on image recognition. Belonging to the technical field of image data processing. Acquiring a first image corresponding to the bearing retainer to be detected, carrying out segmentation matching on the first image, and obtaining circumferential size information corresponding to the bearing retainer to be detected through exclusive or operation; acquiring a second image corresponding to the bearing retainer to be detected, and performing filtering restoration on the second image; determining window hole size information according to the change values of the sizes of the window holes of the bearing retainer to be detected in the filtered and restored second image; dividing the filtered and restored second image into a plurality of rectangular areas, and determining texture features of a plurality of adjacent areas based on gray values of pixel points in the rectangular areas to obtain surface roughness information; and obtaining defect information of the bearing retainer to be detected based on the circumferential size information, the window size information and the surface roughness information. The method reduces the detection cost.
Description
Technical Field
The present disclosure relates to the field of image data processing technologies, and in particular, to a method, an apparatus, and a medium for detecting a bearing holder based on image recognition.
Background
The main function of the bearing cage is to avoid direct contact between the rolling elements, to separate the rolling elements from each other and to guide the rolling of the rolling elements.
The manufacturing process of the bearing retainer mainly comprises the following steps: blanking, stretching, turning edges and bottoms, milling holes and pressing slope surfaces, expanding, surface treatment and the like. The bearing retainer is an important component of the rolling bearing, and once an abnormality occurs, the use of the rolling bearing is directly affected, so that the detection of the quality of the finished product of the bearing retainer is a necessary condition for ensuring the normal operation of the rolling bearing.
In the course of processing, the circularity of bearing retainer is difficult to guarantee, and the ellipse phenomenon easily appears, and simultaneously the position accuracy of bearing retainer fenestration and fenestration size etc. also need higher accuracy. In the prior art, the bearing retainer manufactured is usually detected in a manual mode, but the manual detection accuracy is low, and high labor cost is also easily caused.
Disclosure of Invention
The embodiment of the application provides a bearing retainer detection method, device and medium based on image recognition, which are used for solving the following technical problems: in the prior art, the bearing retainer manufactured is usually detected in a manual mode, but the manual detection accuracy is low, and high labor cost is also easily caused.
The embodiment of the application adopts the following technical scheme:
the embodiment of the application provides a bearing retainer detection method based on image recognition. Acquiring a first image corresponding to a bearing retainer to be detected, performing image segmentation on the first image, performing image matching based on a plurality of edge points corresponding to the segmented image, and performing exclusive-or operation on the matched image to obtain circumferential size information corresponding to the bearing retainer to be detected; acquiring a second image corresponding to the bearing retainer to be detected, and performing filtering restoration on the second image based on a primary wiener filtering restoration result corresponding to the second image and the size of the second image; determining coordinates corresponding to a plurality of window holes of the bearing retainer to be detected in the filtered and restored second image, and determining a plurality of window hole size change values based on the coordinates corresponding to the plurality of window holes, so as to determine window hole size information corresponding to the bearing retainer to be detected according to the plurality of window hole size change values; dividing the filtered and restored second image into a plurality of rectangular areas, and determining a plurality of adjacent areas based on gray values of pixel points in the rectangular areas so as to determine surface roughness information corresponding to the bearing retainer to be detected based on texture features corresponding to the adjacent areas; and obtaining defect information of the bearing retainer to be detected based on the circumferential size information, the window size information and the surface roughness information.
According to the embodiment of the application, the first image is segmented, image matching is carried out according to the segmented first image, the circumference of the bearing retainer can be detected, the bearing retainer is guaranteed to be in a right circular shape, and accuracy of circumference detection is improved. Secondly, the embodiment of the application can solve the problem of unclear image caused by light or shake of the monitoring device and the like by filtering and recovering the second image. In addition, the embodiment of the application determines the size change value of the plurality of window holes according to the coordinates corresponding to the plurality of window holes respectively, so that shooting of each window hole is not required, whether the size of each window hole meets the requirement or not is not required to be measured independently, and the speed of detecting the size of the window hole is improved. According to the embodiment of the application, the texture features are determined through the pixel values of the plurality of neighborhoods in the second image, so that the surface roughness defect of the bearing retainer to be detected currently is determined through the texture features, the detection of the surface roughness defect is not needed, the accuracy of the detection of the surface defect is improved, and meanwhile the labor cost of the detection of the bearing retainer is reduced.
In one implementation manner of the present application, coordinates corresponding to a plurality of windows of a bearing retainer to be detected in a second image after filtering and restoring are determined, and a size change value of the plurality of windows is determined based on the coordinates corresponding to the plurality of windows, which specifically includes: inputting the filtered and restored second image into a preset window vertex detection model to mark a plurality of window vertices in the filtered and restored second image based on the preset window vertex detection model; determining coordinates corresponding to the plurality of window hole vertexes respectively based on the relative positions between the filtered and restored second image edge vertexes and the plurality of window hole vertexes; grouping coordinates corresponding to the window apexes, and determining size information corresponding to each window according to the coordinates corresponding to the window apexes in each group; and determining the size change values of the plurality of window holes based on the size information corresponding to the two adjacent window holes.
In one implementation manner of the present application, image segmentation is performed on the first image, image matching is performed based on a plurality of edge points corresponding to the segmented image, and exclusive or operation is performed on the matched image to obtain circumference size information corresponding to the bearing retainer to be detected, which specifically includes: preprocessing the first image to obtain a gray level image corresponding to the first image; image segmentation is carried out on the first image based on a gray value histogram corresponding to the gray image, so that a bearing retainer image is obtained; randomly selecting a plurality of edge points on the bearing retainer image, determining the diameters corresponding to the edge points, and determining the center point of the bearing retainer image based on the intersection points between the diameters; matching the bearing retainer image with the preset template image based on the center point of the bearing retainer image and the center point of the preset template image; the preset template image is a binary image corresponding to the bearing retainer with the standard size; performing exclusive-or operation on the two matched images to obtain an exclusive-or image; and determining a pixel point set corresponding to the connected domain in the exclusive or graph, determining the area of the connected domain based on the pixel point set, comparing the area of the connected domain with an area threshold value, and determining the circumferential size information corresponding to the bearing retainer to be detected based on the comparison result.
In one implementation manner of the present application, a plurality of adjacent areas are determined based on gray values of each pixel point in the plurality of rectangular areas, so as to determine surface roughness information corresponding to a bearing retainer to be detected based on texture features corresponding to each adjacent area, and the method specifically includes: image segmentation is respectively carried out on the plurality of rectangular areas so as to obtain side images of the bearing retainer, which correspond to the plurality of rectangular areas respectively; acquiring pixel point coordinates corresponding to the side images of the bearing retainers respectively, and determining adjacent points corresponding to the pixel points respectively so as to determine a plurality of adjacent areas based on gray values of the adjacent points; determining texture features corresponding to the rectangular areas respectively based on the gray values and the coordinate information of the adjacent areas; and comparing the texture features corresponding to the rectangular areas with texture features of corresponding preset reference rectangular areas, and determining the surface roughness information corresponding to the bearing retainer to be detected based on the comparison result.
In one implementation manner of the present application, texture features corresponding to a plurality of rectangular areas are compared with texture features of corresponding preset reference rectangular areas, and based on a comparison result, surface roughness information corresponding to a bearing retainer to be detected is determined, and the method specifically includes: determining a corresponding texture type based on the texture features, and determining a preset confidence value corresponding to the texture type in a preset texture feature confidence table; the texture feature confidence list comprises a plurality of texture feature types and confidence values corresponding to the texture feature types respectively; determining the areas of adjacent areas corresponding to different texture types respectively, so as to determine the area occupation ratio of the texture types based on the areas of the adjacent areas, the total areas corresponding to the rectangular areas respectively and the confidence values; and under the condition that the area occupation ratio is larger than an area occupation ratio threshold value corresponding to texture features of a preset reference rectangular area, determining the surface roughness information corresponding to the bearing retainer to be detected based on the comparison result.
In one implementation manner of the present application, determining window size information corresponding to a bearing retainer to be detected according to a plurality of window size variation values specifically includes: acquiring window size variation values of second images corresponding to different orientations of the bearing retainer to be detected respectively; the second images corresponding to the different directions are respectively shot by the monitoring devices uniformly arranged at the different directions of the bearing retainer to be detected; and comparing the window size change values of the second image corresponding to different orientations with the preset reference window size change values corresponding to the corresponding orientations respectively to determine the window size information of the bearing retainer to be detected based on the comparison result.
In one implementation manner of the present application, based on the primary wiener filtering restoration result corresponding to the second image and the size of the second image, performing filtering restoration on the second image specifically includes: restoring the second image based on a preset wiener filter to obtain a first wiener filter restoration result corresponding to the second image; determining a second wiener filtering restoration image result corresponding to the second image according to the first wiener filtering restoration result, the preset point spread function and the preset additive noise function; determining noise corresponding to the second image based on the size corresponding to the original image of the second image and the restoration result of the first wiener filtering; determining a noise mean square value corresponding to the second image according to the first wiener filtering restoration result, the second wiener filtering restoration image result, the noise corresponding to the second image, a preset additive noise function and the size corresponding to the second image; and performing filtering restoration on the second image based on the noise mean square value corresponding to the second image and the first wiener filtering restoration result.
In one implementation manner of the present application, based on the noise mean square value corresponding to the second image and the first wiener filtering restoration result, the filtering restoration is performed on the second image, and specifically includes:
based on the function:
obtaining a noise mean square value corresponding to the second image and a ratio result between the power spectrum densities of the images corresponding to the second image; wherein, ratio results; />The mean square value of the noise corresponding to the second image; />Is the length of the second image; />Is the width of the second image; />Performing Fourier transform on the first wiener filtering recovery result to obtain data;
based on the function:
filtering and restoring the second image; wherein, the filtering recovery result is obtained; />Fourier transform of the preset point spread function; />Fourier transforming the degraded image; />For the ratio result, ++>And performing Fourier transform on the pixel points in the second image.
The embodiment of the application provides a bearing retainer detection device based on image recognition, 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, the instructions being executable by the at least one processor to enable the at least one processor to: acquiring a first image corresponding to a bearing retainer to be detected, performing image segmentation on the first image, performing image matching based on a plurality of edge points corresponding to the segmented image, and performing exclusive-or operation on the matched image to obtain circumference size information corresponding to the bearing retainer to be detected; acquiring a second image corresponding to the bearing retainer to be detected, and performing filtering restoration on the second image based on a primary wiener filtering restoration result corresponding to the second image and the size of the second image; determining coordinates corresponding to a plurality of window holes of the bearing retainer to be detected in the filtered and restored second image, and determining a plurality of window hole size change values based on the coordinates corresponding to the plurality of window holes, so as to determine window hole size information corresponding to the bearing retainer to be detected according to the plurality of window hole size change values; dividing the filtered and restored second image into a plurality of rectangular areas, and determining a plurality of adjacent areas based on gray values of pixel points in the rectangular areas so as to determine surface roughness information corresponding to the bearing retainer to be detected based on texture features corresponding to the adjacent areas; and obtaining defect information of the bearing retainer to be detected based on the circumferential size information, the window size information and the surface roughness information.
The embodiment of the application provides a nonvolatile computer storage medium, which stores computer executable instructions, wherein the computer executable instructions are configured to: acquiring a first image corresponding to a bearing retainer to be detected, performing image segmentation on the first image, performing image matching based on a plurality of edge points corresponding to the segmented image, and performing exclusive-or operation on the matched image to obtain circumference size information corresponding to the bearing retainer to be detected; acquiring a second image corresponding to the bearing retainer to be detected, and performing filtering restoration on the second image based on a primary wiener filtering restoration result corresponding to the second image and the size of the second image; determining coordinates corresponding to a plurality of window holes of the bearing retainer to be detected in the filtered and restored second image, and determining a plurality of window hole size change values based on the coordinates corresponding to the plurality of window holes, so as to determine window hole size information corresponding to the bearing retainer to be detected according to the plurality of window hole size change values; dividing the filtered and restored second image into a plurality of rectangular areas, and determining a plurality of adjacent areas based on gray values of pixel points in the rectangular areas so as to determine surface roughness information corresponding to the bearing retainer to be detected based on texture features corresponding to the adjacent areas; and obtaining defect information of the bearing retainer to be detected based on the circumferential size information, the window size information and the surface roughness information.
The above-mentioned at least one technical scheme that this application embodiment adopted can reach following beneficial effect: according to the embodiment of the application, the first image is segmented, image matching is carried out according to the segmented first image, the circumference of the bearing retainer can be detected, the bearing retainer is guaranteed to be in a right circular shape, and accuracy of circumference detection is improved. Secondly, the embodiment of the application can solve the problem of unclear image caused by light or shake of the monitoring device and the like by filtering and recovering the second image. In addition, according to the embodiment of the application, the size change value of the plurality of window holes is determined according to the coordinates corresponding to the plurality of window holes, so that shooting of each window hole is not required, whether the size of each window hole meets the requirement or not is not required to be measured independently, and the speed of detecting the size of the window hole is improved. According to the embodiment of the application, the texture features are determined through the pixel values of the plurality of neighborhoods in the second image, so that the surface roughness defect of the bearing retainer to be detected currently is determined through the texture features, the detection of the surface roughness defect is not needed, the accuracy of the detection of the surface defect is improved, and meanwhile the labor cost of the detection of the bearing retainer is reduced.
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 that other drawings may be obtained according to these drawings without inventive effort to a person skilled in the art. In the drawings:
FIG. 1 is a flowchart of a method for detecting a bearing retainer based on image recognition according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a bearing retainer detection device based on image recognition according to an embodiment of the present application;
reference numerals:
200 image recognition based bearing retainer detection apparatus, 201 processor, 202 memory.
Detailed Description
The embodiment of the application provides a bearing retainer detection method, equipment and medium based on image recognition.
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 following describes in detail the technical solution proposed in the embodiments of the present application through the accompanying drawings.
Fig. 1 is a flowchart of a method for detecting a bearing retainer based on image recognition according to an embodiment of the present application. As shown in fig. 1, the bearing-holder detection method based on image recognition includes the steps of:
step 101, acquiring a first image corresponding to a bearing retainer to be detected, performing image segmentation on the first image, performing image matching based on a plurality of edge points corresponding to the segmented image, and performing exclusive-or operation on the matched image to obtain circumference size information corresponding to the bearing retainer to be detected.
In one embodiment of the present application, the first image is preprocessed to obtain a gray scale image corresponding to the first image. And carrying out image segmentation on the first image based on the gray value histogram corresponding to the gray image to obtain the bearing retainer image. And randomly selecting a plurality of edge points on the bearing retainer image, determining the diameters corresponding to the edge points, and determining the center point of the bearing retainer image based on the intersection points between the diameters. Matching the bearing retainer image with the preset template image based on the center point of the bearing retainer image and the center point of the preset template image; the preset template image is a binary image corresponding to the bearing retainer with the standard size. And performing exclusive-or operation on the two matched images to obtain an exclusive-or image. And determining a pixel point set corresponding to the connected domain in the exclusive or graph, determining the area of the connected domain based on the pixel point set, comparing the area of the connected domain with an area threshold value, and determining the circumferential size information corresponding to the bearing retainer to be detected based on the comparison result.
Specifically, after the bearing retainer to be detected is processed, shooting the bearing retainer to be detected at different angles is carried out. The bearing retainer to be detected is placed at a preset shooting place and fixed, and a plurality of monitoring devices are uniformly arranged around the shooting place, for example, 4 monitoring devices can be uniformly arranged around the shooting place, so that the side face of the bearing retainer to be detected can be completely shot in four directions of front, back, left and right. Meanwhile, 1 monitoring device is arranged above the bearing retainer to be detected, and the top view of the bearing retainer to be detected is shot through the monitoring device above.
Further, shooting a first image by a monitoring device above the bearing retainer to be detected, wherein the first image is a top view corresponding to the bearing retainer to be detected. The first image may be preprocessed, for example, by denoising and graying, so as to obtain a gray scale image corresponding to the first image. And determining gray values corresponding to each pixel point in the gray map, establishing a gray histogram according to the gray values of each pixel point, and dividing the first image through the gray histogram to divide the background in the first image and the bearing retainer to obtain a divided bearing retainer image.
Further, the roundness of the bearing retainer to be detected is detected according to the bearing retainer image obtained after the segmentation. Specifically, a plurality of edge points are randomly selected on the circumferential edge of the bearing retainer, the diameters corresponding to the plurality of edge points at present are determined, the obtained diameters are connected, and the intersection points corresponding to the diameters can be obtained and can be used as the center point of the bearing retainer to be detected. And matching the center point of the bearing retainer to be detected with the center point of the preset template image, and comparing the image size of the bearing retainer to be detected with that of the preset template image. The preset template image is a top view corresponding to a standard bearing retainer meeting the current production requirement. And the shooting angle position of the preset template image is the same as the shooting angle position corresponding to the first image of the bearing retainer to be detected currently.
Further, after the first image of the bearing retainer to be detected is matched with the preset template image, if the roundness of the bearing retainer to be detected does not meet the requirement, a difference is generated between the circumferential edge position and the preset template image, so that the first image and the preset template image are subjected to exclusive OR operation, and different areas of the first image and the preset template image can be obtained. And establishing a coordinate system based on the matched images to determine the coordinates corresponding to the pixel points of the different areas, thereby determining the areas corresponding to the different areas. Comparing the areas corresponding to the different areas with an area threshold, if the areas are larger than the area threshold, the roundness of the bearing retainer to be detected currently does not meet the production requirement, and simultaneously, displaying the circumferential size information corresponding to the bearing retainer to be detected. The circumference size information at least comprises coordinates corresponding to pixel points of different areas of the two, areas corresponding to the different areas and comparison results with area thresholds.
According to the method and the device for detecting the roundness of the bearing retainer to be detected, the roundness detection efficiency can be improved through image identification, and the accuracy of roundness detection can be improved through image identification. In addition, the problem of high cost required by manual detection can be solved through image recognition.
Step 102, obtaining a second image corresponding to the bearing retainer to be detected, and performing filtering restoration on the second image based on the primary wiener filtering restoration result corresponding to the second image and the size of the second image.
In one embodiment of the present application, the second image is restored based on a preset wiener filter, so as to obtain a first wiener filter restoration result corresponding to the second image. And determining a second wiener filtering restored image result corresponding to the second image according to the first wiener filtering restored result, the preset point spread function and the preset additive noise function. And determining the noise corresponding to the second image based on the first wiener filtering restoration result and the original image of the second image. And determining the mean square value of the noise corresponding to the second image according to the first wiener filtering restoration result, the second wiener filtering restoration image result, the noise corresponding to the second image, the preset additive noise function and the size corresponding to the second image. And performing filtering restoration on the second image based on the noise mean square value corresponding to the second image and the first wiener filtering restoration result.
Specifically, a plurality of second images corresponding to the bearing holder to be detected are acquired based on a plurality of monitoring devices provided around the bearing holder to be detected. Wherein each second image is a side view of the bearing retainer to be detected, and each second image corresponds to a different orientation of the side of the bearing retainer to be detected, for example, may correspond to four orientations of the front side, the left side, the right side and the rear side of the bearing retainer to be detected, respectively.
Further, the second image of the bearing retainer to be detected is restored for the first time based on the preset wiener filter in the prior art, and a first wiener filter restoration result corresponding to the second image is obtained.
Based on the function:
and obtaining a second wiener filtering restoration image result corresponding to the second image. Wherein, restoring the result for the second wiener filtering, < >>Restoring the result for the first wiener filtering, +.>Fourier transform of the preset point spread function, +.>For presetting an additive noise function->And performing Fourier transform on the pixel points in the second image.
Further, after the first wiener filtering restoration result of the second image is subjected to Fourier transformation, square calculation is performed on the amplitude of the transformed result, and ratio calculation is performed on the square calculation result and the length-width product of the second image, so that the power spectral density of the second image is obtained.
Further, according to the first wiener filtering restoration result of the second image, the noise is summed, and the ratio of the sum to the length-width product of the second image is calculated, so that the noise mean square value corresponding to the second image is obtained. Specifically, a product calculation is performed on a first wiener filter restoration result of a second image and a preset point spread function to obtain a first numerical value, a noise value is obtained based on a difference value between the second wiener filter restoration result of the second image and the first numerical value, a summation calculation is performed on the obtained noise value, and a ratio calculation is performed on the obtained summation numerical value and a length-width product of the second image to obtain a mean square value of noise corresponding to the second image.
Further, based on the function:
obtaining a ratio result between the noise mean square value corresponding to the second image and the image power spectral density corresponding to the second image; wherein, ratio results; />The mean square value of the noise corresponding to the second image; />Is the length of the second image; />Is the width of the second image; />Performing Fourier transform on the first wiener filtering recovery result to obtain data;
based on the function:
filtering and restoring the second image; wherein, The filtering recovery result is obtained; />Fourier transform of the preset point spread function; />Fourier transforming the degraded image; />For the ratio result, ++>And performing Fourier transform on the pixel points in the second image.
According to the embodiment of the application, the first filtering image result is used as input, the image is filtered again, and the filtering effect of the image can be improved, so that the labeling of the apexes of the window holes in the side view of the bearing retainer to be detected is more accurate. The extraction of the texture features in the contralateral view is more accurate, thereby improving the accuracy of the detection of the bearing holder.
And 103, determining coordinates corresponding to a plurality of window holes of the bearing retainer to be detected in the filtered and restored second image, and determining a plurality of window hole size change values based on the coordinates corresponding to the plurality of window holes, so as to determine window hole size information corresponding to the bearing retainer to be detected according to the plurality of window hole size change values.
In one embodiment of the present application, the filtered restored second image is input to a preset fenestration vertex detection model to label a plurality of fenestration vertices in the filtered restored second image based on the preset fenestration vertex detection model. And determining coordinates corresponding to the plurality of window vertices respectively based on the relative positions of the filtered and restored second image edge vertices and the plurality of window vertices. And grouping coordinates corresponding to the window apexes, and determining size information corresponding to each window according to the coordinates corresponding to the window apexes in each group. And determining the size change values of the plurality of window holes based on the size information corresponding to the two adjacent window holes.
Specifically, the embodiment of the application is provided with a preset window hole vertex detection model for detecting and labeling the vertices of the window holes in the image. The training process is that a bearing retainer side image sample set is taken as input, a bearing retainer side image sample set of a standard window hole vertex is taken as output, and a preset neural network model is trained to obtain the preset window hole vertex detection model.
Further, the second image after the current filtering and restoration is input into a preset window vertex detection model, so that the window vertices in the current second image are marked through the preset window vertex detection model.
For example, an intersection point between a straight line where the left edge of the second image is located and a straight line where the lower edge is located is used as an origin, a straight line where the left edge of the second image is located is used as a y axis, and a straight line where the lower edge of the second image is located is used as an x axis. And determining the coordinates of the left upper vertex, the left lower vertex, the right upper vertex and the right lower vertex of the second image, connecting the four vertices of the second image with the currently marked window hole vertexes respectively, and determining the coordinates of the currently marked window hole vertexes based on the length and the angle of the connecting line.
Further, each window of the bearing retainer to be detected corresponds to four vertexes, the four vertexes are divided into a group by taking the four vertexes of the middle position of the current second image as the center, and the area formed by the four vertexes is one window nearest to the monitoring device for shooting the second image. Wherein, four window apexes in each group are adjacent four apexes, and the four apexes can form a rectangular area. Based on the coordinates of the four vertexes in each group, the length and width dimensions of the rectangular areas corresponding to the four vertexes can be determined, and the length and width dimensions of the bearing retainer window holes corresponding to the four vertexes can be obtained.
Further, the size of the central window is determined by taking a window closest to the monitoring device for capturing the second image as the center, and then the sizes of the plurality of windows are sequentially determined on the left and right sides. Since the aperture in the captured image is gradually smaller in the aperture area on both the left and right sides with the center aperture as the center, the aperture size on both the left and right sides obtained from the image is gradually smaller. Therefore, the change value between the sizes of the plurality of window holes sequentially detected leftward and the change value between the sizes of the plurality of window holes sequentially detected rightward with the center window hole as the center in the current second image can be obtained. Wherein the variation value may be a ratio between a length of a previous aperture and a length of a next aperture, and a ratio between a width of the previous aperture and a width of the next aperture.
In one embodiment of the application, obtaining the size change value of the window hole of the second image corresponding to different orientations of the bearing retainer to be detected; the second images corresponding to the different directions are respectively shot by the monitoring devices uniformly arranged at the different directions of the bearing retainer to be detected. And comparing the window size change values of the second images corresponding to different orientations with the preset reference window size change values corresponding to the corresponding orientations respectively, so as to determine the size defect information of the window hole of the bearing retainer to be detected based on the comparison result.
Specifically, in the embodiment of the application, a plurality of monitoring devices are uniformly arranged on the side surface of the bearing holder to be detected, each monitoring device can shoot a bearing holder side surface image of a certain angle, for example, in the case of arranging four monitoring devices, each monitoring device can shoot at least a 90-degree bearing holder side surface image. Thereby the side of the bearing retainer to be detected is completely shot through four monitoring devices.
Further, each monitoring device corresponds to a preset reference window size change value, and the window size change value of the second image corresponding to the bearing retainer to be detected at present is compared with the preset reference window size change value, so that the size defect information of the window hole of the bearing retainer to be detected is determined based on the comparison result.
According to the method and the device, the size change value of the plurality of window holes is determined according to the coordinates corresponding to the plurality of window holes, so that shooting of each window hole is not needed, and the size of each window hole is not needed to be measured independently to determine whether the size meets the generation requirement. And then shoot less images through less monitoring devices, every image can detect a plurality of fenestrations simultaneously, improves the speed of detecting the window size.
And 104, dividing the filtered and restored second image into a plurality of rectangular areas, and determining a plurality of adjacent areas based on gray values of all pixel points in the rectangular areas so as to determine surface roughness information corresponding to the bearing retainer to be detected based on texture features corresponding to the adjacent areas.
In one embodiment of the present application, image segmentation is performed on each of the plurality of rectangular areas to obtain bearing retainer side images corresponding to each of the plurality of rectangular areas. And acquiring pixel point coordinates corresponding to the side images of the plurality of bearing retainers, and determining adjacent points corresponding to the pixel points respectively so as to determine a plurality of adjacent areas based on gray values of the adjacent points. And determining texture features corresponding to the rectangular areas respectively based on the gray values and the coordinate information of the adjacent areas. And comparing the texture features corresponding to the rectangular areas with texture features of corresponding preset reference rectangular areas, and determining the surface roughness information corresponding to the bearing retainer to be detected based on the comparison result.
Specifically, the filtered restored second image is divided into a plurality of rectangular areas. And carrying out image segmentation on each rectangular area to determine the side image of the corresponding bearing retainer to be detected in each area. And secondly, determining a pixel point coordinate set in each rectangular region, and determining adjacent points corresponding to each pixel point respectively, wherein the adjacent points can be 8 pixel points of the current pixel point, namely upper, lower, left and right, upper left, lower right and lower right. And determining the gray value of each adjacent point based on 8 adjacent points, dividing the adjacent points with approximate gray values into the same group, and dividing each rectangular area again, wherein each rectangular area can be divided into a plurality of adjacent areas, and the gray values of one or more pixel points in the same adjacent area are approximately the same.
Further, a plurality of adjacent areas with approximate gray values in each rectangular area are communicated, distribution conditions corresponding to pixels with different gray values can be obtained, and texture features corresponding to the current rectangular area are obtained based on the gray value distribution conditions of the communicated areas and the distribution conditions of the pixels of the communicated areas. Wherein, each rectangular region can have one or more connected regions, and the gray value distribution of each connected region is different.
In one embodiment of the application, determining a corresponding texture type based on texture features, and determining a preset confidence value corresponding to the texture type in a preset texture feature confidence table; the texture feature confidence list comprises a plurality of texture feature types and confidence values corresponding to the texture feature types. And determining the area of the region corresponding to each different texture type, so as to determine the area occupation ratio of the texture type based on the area of the region, the total area corresponding to each of the plurality of rectangular regions and the confidence value. And under the condition that the area occupation ratio is larger than an area occupation ratio threshold value corresponding to texture features of a preset reference rectangular area, determining the surface roughness information corresponding to the bearing retainer to be detected based on the comparison result.
Specifically, the embodiment of the application is provided with a texture feature confidence table, wherein the texture feature confidence table comprises a plurality of texture feature types and preset confidence values corresponding to the texture feature types respectively. Wherein, different texture feature types correspond to different gray value distribution conditions. And determining the corresponding texture feature type, and determining a confidence value corresponding to the texture feature type based on the texture feature confidence.
Further, determining the area of the connected domain corresponding to the texture type which does not accord with the production condition currently, determining the area of the rectangular area where the connected domain is located, obtaining a connected domain reference area value based on the product between the area of the connected domain and the confidence value, and calculating the ratio of the connected domain reference area value to the area of the rectangular area, so that the area occupation ratio corresponding to the current texture type can be obtained. If the area occupation ratio is larger than the area occupation ratio threshold value corresponding to the texture features of the preset reference rectangular area, the problem of surface roughness of the area is indicated, and the surface roughness information corresponding to the bearing retainer to be detected needs to be given. The surface roughness information at least comprises texture feature types corresponding to the current rectangular area, area occupation ratios corresponding to different texture features respectively, and whether the different area occupation ratios exceed an area occupation ratio threshold value.
And 105, obtaining defect information of the bearing retainer to be detected based on the circumferential size information, the window size information and the surface roughness information.
Comparing the size change value of the window hole of the second image corresponding to the current bearing retainer to be detected with the size change value of the preset reference window hole, and determining that the window hole of the current bearing retainer to be detected has a size defect under the condition that the comparison result is larger than a preset threshold value. And presents window size information corresponding to the bearing retainer to be inspected. The window size information may include at least coordinates corresponding to windows that do not meet the production conditions, and a comparison result with a preset threshold.
Further, if the area of the connected domain in the exclusive or graph is larger than the area threshold, the difference between the first image and the preset template image is larger, that is, the circumference of the current bearing retainer to be detected does not meet the production condition, and at the moment, circumference size defect information corresponding to the bearing retainer to be detected is generated. The circumferential defect information at least comprises a pixel point set of a connected domain in the exclusive or diagram and a difference value between the pixel point set and the area threshold value.
Determining surface roughness defect information corresponding to the bearing retainer to be detected based on the surface roughness information corresponding to the bearing retainer to be detected; the surface roughness defect information at least comprises a surface roughness grade, a surface defect type and processing equipment related to the surface defect type.
Further, according to the obtained surface roughness information corresponding to the bearing retainer to be detected, the surface roughness defect information corresponding to the bearing retainer to be detected is determined. The surface roughness defect information at least comprises a surface roughness grade, a surface defect type and processing equipment related to the surface defect type. The surface defect types can be rust, crack, abrasion, scratch, crush injury, burr, burn and the like, and the production process which can possibly generate the defect can be determined through the determined surface defect types, so that the production equipment which can possibly generate the defect is determined, and the corresponding equipment of workers is reminded to carry out detection maintenance.
Fig. 2 is a schematic structural diagram of a bearing retainer detection device based on image recognition according to an embodiment of the present application. As shown in fig. 2, the bearing-holder detecting apparatus 200 based on image recognition 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, the instructions being executable by the at least one processor 201 to enable the at least one processor 201 to: acquiring a first image corresponding to a bearing retainer to be detected, performing image segmentation on the first image, performing image matching based on a plurality of edge points corresponding to the segmented image, and performing exclusive-or operation on the matched image to obtain circumferential size information corresponding to the bearing retainer to be detected; acquiring a second image corresponding to the bearing retainer to be detected, and performing filtering restoration on the second image based on a primary wiener filtering restoration result corresponding to the second image and the size of the second image; determining coordinates corresponding to a plurality of window holes of the bearing retainer to be detected in the filtered and restored second image, and determining the size change values of the window holes based on the coordinates corresponding to the window holes, so as to determine window hole size information corresponding to the bearing retainer to be detected according to the size change values of the window holes; dividing the filtered and restored second image into a plurality of rectangular areas, determining a plurality of adjacent areas based on gray values of pixel points in the rectangular areas, and determining surface roughness information corresponding to the bearing retainer to be detected based on texture features corresponding to the adjacent areas; and obtaining defect information of the bearing retainer to be detected based on the circumferential size information, the window size information and the surface roughness information.
The embodiments also provide a non-volatile computer storage medium storing computer executable instructions configured to: acquiring a first image corresponding to a bearing retainer to be detected, performing image segmentation on the first image, performing image matching based on a plurality of edge points corresponding to the segmented image, and performing exclusive-or operation on the matched image to obtain circumference size information corresponding to the bearing retainer to be detected; acquiring a second image corresponding to the bearing retainer to be detected, and performing filtering restoration on the second image based on a primary wiener filtering restoration result corresponding to the second image and the size of the second image; determining coordinates corresponding to a plurality of window holes of the bearing retainer to be detected in the filtered and restored second image, and determining a plurality of window hole size change values based on the coordinates corresponding to the plurality of window holes, so as to determine window hole size information corresponding to the bearing retainer to be detected according to the plurality of window hole size change values; dividing the filtered and restored second image into a plurality of rectangular areas, and determining a plurality of adjacent areas based on gray values of pixel points in the rectangular areas so as to determine surface roughness information corresponding to the bearing retainer to be detected based on texture features corresponding to the adjacent areas; and obtaining defect information of the bearing retainer to be detected based on the circumferential size information, the window size information and the surface roughness information.
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 apparatus, devices, non-volatile computer storage medium embodiments, the description is relatively simple, as it is substantially similar to method embodiments, with reference to the section of the method embodiments being relevant.
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. Such modifications and changes do not depart from the spirit and scope of the embodiments of the present application.
Claims (10)
1. A method of detecting a bearing retainer based on image recognition, the method comprising:
acquiring a first image corresponding to a bearing retainer to be detected, performing image segmentation on the first image, performing image matching based on a plurality of edge points corresponding to the segmented image, and performing exclusive-or operation on the matched image to obtain circumferential size information corresponding to the bearing retainer to be detected;
Acquiring a second image corresponding to the bearing retainer to be detected, and performing filtering restoration on the second image based on a primary wiener filtering restoration result corresponding to the second image and the size of the second image;
determining coordinates corresponding to a plurality of window holes of the bearing retainer to be detected in the filtered and restored second image, and determining size change values of the window holes based on the coordinates corresponding to the window holes to be detected, so as to determine window hole size information corresponding to the bearing retainer to be detected according to the size change values of the window holes;
dividing the filtered and restored second image into a plurality of rectangular areas, determining a plurality of adjacent areas based on gray values of pixel points in the rectangular areas, and determining surface roughness information corresponding to the bearing retainer to be detected based on texture features corresponding to the adjacent areas;
and obtaining defect information of the bearing retainer to be detected based on the circumferential size information, the window size information and the surface roughness information.
2. The method for detecting a bearing retainer based on image recognition according to claim 1, wherein the determining coordinates corresponding to a plurality of windows of the bearing retainer to be detected in the second image after the filtering and restoring, and determining the size change values of the plurality of windows based on the coordinates corresponding to the plurality of windows, comprises:
Inputting the filtered and restored second image into a preset window aperture vertex detection model to mark a plurality of window aperture vertices in the filtered and restored second image based on the preset window aperture vertex detection model;
determining coordinates corresponding to the plurality of window hole vertexes respectively based on relative positions between the filtered and restored second image edge vertexes and the plurality of window hole vertexes;
grouping coordinates corresponding to the window apexes, and determining size information corresponding to each window according to the coordinates corresponding to the window apexes in each group;
and determining the size change values of the plurality of window holes based on the size information corresponding to the two adjacent window holes.
3. The method for detecting the bearing retainer based on image recognition according to claim 1, wherein the image segmentation is performed on the first image, image matching is performed based on a plurality of edge points corresponding to the segmented image, and exclusive-or operation is performed on the matched image to obtain the circumferential size information corresponding to the bearing retainer to be detected, and the method specifically comprises:
preprocessing the first image to obtain a gray level image corresponding to the first image;
Image segmentation is carried out on the first image based on a gray value histogram corresponding to the gray map, so that a bearing retainer image is obtained;
a plurality of edge points are selected at will on the bearing retainer image, diameters corresponding to the edge points are determined, and a center point of the bearing retainer image is determined based on the intersection points among the diameters;
matching the bearing retainer image with a preset template image based on the center point of the bearing retainer image and the center point of the preset template image; the preset template image is a binary image corresponding to the bearing retainer with the standard size;
performing exclusive-or operation on the two matched images to obtain an exclusive-or image;
and determining a pixel point set corresponding to the connected domain in the exclusive or graph, determining the area of the connected domain based on the pixel point set, comparing the area of the connected domain with an area threshold value, and determining the circumferential size information corresponding to the bearing retainer to be detected based on the comparison result.
4. The method for detecting a bearing retainer based on image recognition according to claim 1, wherein the determining a plurality of adjacent areas based on gray values of pixels in the plurality of rectangular areas, so as to determine the surface roughness information corresponding to the bearing retainer to be detected based on texture features corresponding to the adjacent areas, specifically comprises:
Image segmentation is respectively carried out on the rectangular areas so as to obtain side images of the bearing retainer, which correspond to the rectangular areas respectively;
acquiring pixel point coordinates corresponding to the side images of the bearing retainers respectively, and determining adjacent points corresponding to the pixel points respectively so as to determine a plurality of adjacent areas based on gray values of the adjacent points;
determining texture features corresponding to the rectangular areas respectively based on the gray values and coordinates of the adjacent areas;
and comparing the texture features corresponding to the rectangular areas with texture features of corresponding preset reference rectangular areas, and determining the surface roughness information corresponding to the bearing retainer to be detected based on the comparison result.
5. The method for detecting the bearing retainer based on image recognition according to claim 4, wherein the comparing the texture features corresponding to the rectangular regions with the texture features of the corresponding preset reference rectangular regions, and determining the surface roughness information corresponding to the bearing retainer to be detected based on the comparison result, specifically comprises:
determining a corresponding texture type based on the texture features, and determining a preset confidence value corresponding to the texture type in a preset texture feature confidence table; the texture feature confidence list comprises a plurality of texture feature types and confidence values respectively corresponding to the texture feature types;
Determining the areas of adjacent areas corresponding to different texture types respectively, so as to determine the area occupation ratio of the texture types based on the areas of the adjacent areas, the total areas corresponding to a plurality of rectangular areas respectively and the confidence value;
and under the condition that the area occupation ratio is larger than an area occupation ratio threshold value corresponding to texture features of a preset reference rectangular area, determining the surface roughness information corresponding to the bearing retainer to be detected based on a comparison result.
6. The method for detecting the bearing retainer based on image recognition according to claim 1, wherein determining window size information corresponding to the bearing retainer to be detected according to the plurality of window size variation values specifically comprises:
acquiring window size variation values of second images corresponding to different orientations of the bearing retainer to be detected respectively; the second images corresponding to the different directions are respectively shot by the monitoring devices uniformly arranged at the different directions of the bearing retainer to be detected;
and comparing the window size change values of the second images corresponding to the different orientations with the preset reference window size change values corresponding to the corresponding orientations respectively, so as to determine window size information corresponding to the bearing retainer to be detected based on the comparison result.
7. The method for detecting a bearing retainer based on image recognition according to claim 1, wherein the filtering recovery of the second image based on the primary wiener filtering recovery result corresponding to the second image and the size of the second image specifically includes:
restoring the second image based on a preset wiener filter to obtain a first wiener filter restoration result corresponding to the second image;
determining a second wiener filtering restoration image result corresponding to the second image according to the first wiener filtering restoration result, a preset point spread function and a preset additive noise function;
determining noise corresponding to a second image based on the size corresponding to the second image of the first wiener filtering restoration result;
determining a noise mean square value corresponding to the second image according to the first wiener filtering restoration result, the second wiener filtering restoration image result, the noise corresponding to the second image, the preset additive noise function and the size corresponding to the second image;
and performing filtering restoration on the second image based on the noise mean square value corresponding to the second image and the first wiener filtering restoration result.
8. The method for detecting a bearing retainer based on image recognition according to claim 7, wherein the filtering and restoring the second image based on the mean square value of noise corresponding to the second image as the result of the first wiener filtering and restoring specifically comprises:
function-based
Obtaining a ratio result between the noise mean square value corresponding to the second image and the image power spectral density corresponding to the second image; wherein, the ratio results; />The mean square value of the noise corresponding to the second image; />Is the length of the second image; />Is the width of the second image; />To be the instituteThe first wiener filtering recovery result is subjected to Fourier transformation to obtain data;
function-based
9. An image recognition based bearing retainer inspection apparatus, characterized in that the apparatus comprises a memory for storing computer program instructions and a processor for executing the program instructions, wherein the computer program instructions, when executed by the processor, trigger the apparatus to perform the method of any of claims 1-8.
10. A non-transitory computer storage medium storing computer executable instructions, wherein the computer executable instructions are capable of performing the method of any one of claims 1-8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310348401.5A CN116128873A (en) | 2023-04-04 | 2023-04-04 | Bearing retainer detection method, device and medium based on image recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310348401.5A CN116128873A (en) | 2023-04-04 | 2023-04-04 | Bearing retainer detection method, device and medium based on image recognition |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116128873A true CN116128873A (en) | 2023-05-16 |
Family
ID=86299313
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310348401.5A Pending CN116128873A (en) | 2023-04-04 | 2023-04-04 | Bearing retainer detection method, device and medium based on image recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116128873A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116523922A (en) * | 2023-07-05 | 2023-08-01 | 上海圣德曼铸造海安有限公司 | Bearing surface defect identification method |
CN116586925A (en) * | 2023-07-19 | 2023-08-15 | 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 |
Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102297761A (en) * | 2011-06-15 | 2011-12-28 | 西安交通大学 | Bolt faying face supersonic wave detection apparatus and data processing method thereof |
CN104346773A (en) * | 2013-07-23 | 2015-02-11 | 中国科学院长春光学精密机械与物理研究所 | Push-scan type space camera image on-orbit real-time restoration method |
CN105118033A (en) * | 2015-08-19 | 2015-12-02 | 杭州电子科技大学 | Wiener filtering image restoration method using adaptive noise estimation |
CN106651791A (en) * | 2016-11-21 | 2017-05-10 | 云南电网有限责任公司电力科学研究院 | Recovery method for single motion blur image |
CN110458901A (en) * | 2019-06-26 | 2019-11-15 | 西安电子科技大学 | A kind of optimum design method of overall importance based on the photo electric imaging system for calculating imaging |
CN110728253A (en) * | 2019-07-22 | 2020-01-24 | 中南大学 | Texture feature measurement method based on particle roughness |
CN110728677A (en) * | 2019-07-22 | 2020-01-24 | 中南大学 | Texture roughness defining method based on sliding window algorithm |
CN111583243A (en) * | 2020-05-09 | 2020-08-25 | 中南大学 | Method for reconstructing adjacent point cell for detecting cellular regularity |
CN111583237A (en) * | 2020-05-09 | 2020-08-25 | 中南大学 | Neighborhood window recursion cell reconstruction method for detecting cellular regularity |
CN111652811A (en) * | 2020-04-14 | 2020-09-11 | 中国人民解放军国防科技大学 | Motion blurred image restoration method based on edge function and optimal window wiener filtering |
CN114897881A (en) * | 2022-06-10 | 2022-08-12 | 合肥工业大学 | Crystal grain defect detection method based on edge characteristics |
CN114937000A (en) * | 2022-03-30 | 2022-08-23 | 中国人民解放军总医院第二医学中心 | Computer-implemented method for determining a stenosis rate of a tubular and related product |
CN115018828A (en) * | 2022-08-03 | 2022-09-06 | 深圳市尹泰明电子有限公司 | Defect detection method for electronic component |
CN115493536A (en) * | 2022-11-16 | 2022-12-20 | 广州粤芯半导体技术有限公司 | Roundness measuring method and device, equipment terminal and readable storage medium |
-
2023
- 2023-04-04 CN CN202310348401.5A patent/CN116128873A/en active Pending
Patent Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102297761A (en) * | 2011-06-15 | 2011-12-28 | 西安交通大学 | Bolt faying face supersonic wave detection apparatus and data processing method thereof |
CN104346773A (en) * | 2013-07-23 | 2015-02-11 | 中国科学院长春光学精密机械与物理研究所 | Push-scan type space camera image on-orbit real-time restoration method |
CN105118033A (en) * | 2015-08-19 | 2015-12-02 | 杭州电子科技大学 | Wiener filtering image restoration method using adaptive noise estimation |
CN106651791A (en) * | 2016-11-21 | 2017-05-10 | 云南电网有限责任公司电力科学研究院 | Recovery method for single motion blur image |
CN110458901A (en) * | 2019-06-26 | 2019-11-15 | 西安电子科技大学 | A kind of optimum design method of overall importance based on the photo electric imaging system for calculating imaging |
CN110728677A (en) * | 2019-07-22 | 2020-01-24 | 中南大学 | Texture roughness defining method based on sliding window algorithm |
CN110728253A (en) * | 2019-07-22 | 2020-01-24 | 中南大学 | Texture feature measurement method based on particle roughness |
CN111652811A (en) * | 2020-04-14 | 2020-09-11 | 中国人民解放军国防科技大学 | Motion blurred image restoration method based on edge function and optimal window wiener filtering |
CN111583243A (en) * | 2020-05-09 | 2020-08-25 | 中南大学 | Method for reconstructing adjacent point cell for detecting cellular regularity |
CN111583237A (en) * | 2020-05-09 | 2020-08-25 | 中南大学 | Neighborhood window recursion cell reconstruction method for detecting cellular regularity |
CN114937000A (en) * | 2022-03-30 | 2022-08-23 | 中国人民解放军总医院第二医学中心 | Computer-implemented method for determining a stenosis rate of a tubular and related product |
CN114897881A (en) * | 2022-06-10 | 2022-08-12 | 合肥工业大学 | Crystal grain defect detection method based on edge characteristics |
CN115018828A (en) * | 2022-08-03 | 2022-09-06 | 深圳市尹泰明电子有限公司 | Defect detection method for electronic component |
CN115493536A (en) * | 2022-11-16 | 2022-12-20 | 广州粤芯半导体技术有限公司 | Roundness measuring method and device, equipment terminal and readable storage medium |
Non-Patent Citations (3)
Title |
---|
张明辉 等: "机器视觉在轴承检测中的研究现状及发展趋势", 《机床与液压》, vol. 47, no. 23, pages 183 - 189 * |
王朝晖 等: "轴承保持架计算机视觉检测方法研究", 《机械设计与制造》, pages 35 - 37 * |
陈向伟 等: "基于计算机视觉的微小轴承表面缺陷检测", 《机床与液压》, vol. 37, no. 9, pages 130 - 132 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116523922A (en) * | 2023-07-05 | 2023-08-01 | 上海圣德曼铸造海安有限公司 | Bearing surface defect identification method |
CN116523922B (en) * | 2023-07-05 | 2023-10-20 | 上海圣德曼铸造海安有限公司 | Bearing surface defect identification method |
CN116586925A (en) * | 2023-07-19 | 2023-08-15 | 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 |
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 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6358351B1 (en) | Surface defect inspection method and surface defect inspection apparatus | |
CN110163853B (en) | Edge defect detection method | |
CN111179225B (en) | Test paper surface texture defect detection method based on gray gradient clustering | |
CN116168026B (en) | Water quality detection method and system based on computer vision | |
CN106934803B (en) | method and device for detecting surface defects of electronic device | |
CN116128873A (en) | Bearing retainer detection method, device and medium based on image recognition | |
CN113109368B (en) | Glass crack detection method, device, equipment and medium | |
CN111353993B (en) | Thread angle measuring method based on machine vision | |
CN109632808B (en) | Edge defect detection method and device, electronic equipment and storage medium | |
CN115222733A (en) | Metal component surface defect detection method based on image recognition | |
CN114693610A (en) | Welding seam surface defect detection method, equipment and medium based on machine vision | |
CN109444169A (en) | A kind of bearing defect detection method and system | |
CN113012157B (en) | Visual detection method and system for equipment defects | |
CN113077437B (en) | Workpiece quality detection method and system | |
CN115953397A (en) | Method and equipment for monitoring process preparation flow of conical bearing retainer | |
CN115375629A (en) | Method for detecting line defect and extracting defect information in LCD screen | |
CN112669265B (en) | Method for realizing surface defect detection based on Fourier transform and image gradient characteristics | |
CN108346138B (en) | Surface defect detection method and system based on image processing | |
CN113516619A (en) | Product surface flaw identification method based on image processing technology | |
CN112950594A (en) | Method and device for detecting surface defects of product and storage medium | |
CN116883987A (en) | Pointer instrument reading identification method for unmanned inspection of transformer substation | |
CN111028215A (en) | Method for detecting end surface defects of steel coil based on machine vision | |
CN115311287A (en) | Method for detecting production abnormity of common rail oil injector | |
CN115239663A (en) | Method and system for detecting defects of contact lens, electronic device and storage medium | |
CN115049641A (en) | Electric data processing method and system for anomaly detection of mechanical parts |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20230516 |
|
RJ01 | Rejection of invention patent application after publication |