CN115861325B - Suspension spring defect detection method and system based on image data - Google Patents
Suspension spring defect detection method and system based on image data Download PDFInfo
- Publication number
- CN115861325B CN115861325B CN202310181114.XA CN202310181114A CN115861325B CN 115861325 B CN115861325 B CN 115861325B CN 202310181114 A CN202310181114 A CN 202310181114A CN 115861325 B CN115861325 B CN 115861325B
- Authority
- CN
- China
- Prior art keywords
- defect
- gray
- value
- pixel points
- pixel point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000007547 defect Effects 0.000 title claims abstract description 225
- 239000000725 suspension Substances 0.000 title claims abstract description 125
- 238000001514 detection method Methods 0.000 title claims abstract description 30
- 230000000694 effects Effects 0.000 claims abstract description 46
- 238000011156 evaluation Methods 0.000 claims abstract description 34
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000012216 screening Methods 0.000 claims abstract description 12
- 238000004364 calculation method Methods 0.000 claims description 25
- 238000004590 computer program Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 4
- 238000012545 processing Methods 0.000 abstract description 5
- 230000002950 deficient Effects 0.000 description 16
- 230000008569 process Effects 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000005299 abrasion Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 7
- 230000011218 segmentation Effects 0.000 description 7
- 230000000087 stabilizing effect Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000005489 elastic deformation Effects 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000035945 sensitivity 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
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to a suspension spring defect detection method and system based on image data. Firstly, sliding a sliding window on a suspension spring surface image, and screening out an initial defect area; obtaining a surface difference factor of the pixel point according to the color characteristic factor and the gray value of the pixel point in the initial defect area; randomly classifying pixel points in any initial defect area, and obtaining two categories by random classification each time; for random classification at any time, calculating the inter-class variance between two classes and the difference degree of gray histograms corresponding to the two classes; and obtaining a division effect evaluation value according to the inter-class variance and the degree of difference, and obtaining a defect class and a corresponding suspension spring defect area according to the division effect evaluation value. The method and the device construct corresponding dividing effect evaluation values by combining the characteristics of the surface defects of the suspension springs, and adjust the dividing effect through the dividing effect evaluation values, so that defect areas are obtained, and the accuracy of defect detection is improved.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a suspension spring defect detection method and system based on image data.
Background
Suspension springs are a critical component commonly found in automotive suspension systems, by which mechanical components of the suspension spring convert jounce energy into other forms of energy during travel. If the mechanical part such as the spring is used, when the automobile runs on a bumpy and uneven road surface according to excellent elastic deformation of the spring, the mechanical energy of the system is converted into elastic deformation energy of the spring, so that stability and safety of the automobile in a running process are ensured. It follows that the suspension springs play a significant role in smooth and safe operation of the overall vehicle system. However, as a mechanical component for high frequency use, the surface of the suspension spring wears to some extent during frequent use. When the surface of the suspension spring is worn seriously, the deformation of the whole spring is abnormal, so that the stability and safety experience in the running process of the vehicle are further influenced, related traffic accidents can be caused seriously, serious economic property loss is caused, and life safety is threatened.
At present, common methods for detecting defects of a suspension spring are as follows: and acquiring a suspension spring image, and performing threshold segmentation on the suspension spring image to obtain a final defect region. The traditional threshold segmentation algorithm needs to set an experience threshold, and when the experience threshold is set inaccurately, the spring defect detection result obtained through the traditional threshold segmentation algorithm is affected, and the result of suspension spring defect detection is affected.
Disclosure of Invention
In order to solve the technical problem that when a traditional threshold segmentation algorithm needs to set a proper experience threshold, otherwise, the detection result of the suspension spring defect is affected, the invention aims to provide a suspension spring defect detection method and system based on image data, and the adopted technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a suspension spring defect detection method based on image data, the method including the steps of:
acquiring an image of the surface of the suspension spring;
sliding a sliding window on the suspension spring surface image, and screening out an initial defect area according to the gray value of a pixel point in the area corresponding to the sliding window; forming color characteristic factors according to the values of all channels of the pixel points in the initial defect area in the LAB color space; for each pixel point in the initial defect area, obtaining a corresponding surface difference factor according to the color characteristic factor and the gray value;
Randomly classifying the pixel points in each initial defect area at least twice, and obtaining two categories by random classification each time; for two categories obtained by random classification at any time, calculating the inter-category variance between the two categories according to the surface difference factors of the pixel points in the two categories, the gray values corresponding to the pixel points and the occurrence frequencies of the gray values; calculating the difference degree of two gray histograms corresponding to pixel points in two categories;
obtaining a classification effect evaluation value according to the inter-class variance and the degree of the difference, and obtaining the class with larger gray average value from two classes corresponding to the maximum classification effect evaluation value as a defect class; and forming a suspension spring defect area by pixel points in each defect category.
Preferably, for each pixel point in the initial defect area, the obtaining a corresponding surface difference factor according to the color feature factor and the gray value includes:
selecting any pixel point in the initial defect area as a target pixel point, and calculating a gray level difference value of gray level values of the target pixel point and the pixel points in the eight neighborhood; calculating the sum of absolute values of gray differences between the pixel points in the eight adjacent domains and the target pixel point; the ratio of the gray level difference value to the sum of the absolute values is a weight stabilizing coefficient of the target pixel point;
Calculating the product of the gray difference value of any pixel point in the eight neighborhood and the weight stabilizing coefficient to be used as a first product; taking the sum of the first products of the pixel points in the eight neighborhoods as an initial difference factor;
and calculating the product of the color characteristic factor and the initial difference factor to be used as a second product, and taking the normalized second product as a surface difference factor.
Preferably, the calculating the inter-class variance between the two classes according to the surface difference factor of the pixel points in the two classes, the gray value corresponding to the pixel points, and the occurrence frequency of the gray value includes:
calculating the defect probability of the category according to the surface difference factor of the pixel points in the category and the occurrence probability of the gray value corresponding to the pixel points; calculating the probability mean value of the class according to the surface difference factor of the pixel points in the class, the gray value corresponding to the pixel points and the occurrence frequency of the gray value;
taking the defect probability of the two categories as a weight, and carrying out weighted summation on the probability average values of the two categories to obtain a regional probability average value of the initial defect region;
respectively calculating squares of differences between the probability average values of the two categories and the regional probability average value to serve as an average value difference factor; and taking the defect probability of the two categories as a weight, and carrying out weighted summation on the mean difference factors corresponding to the two categories to obtain the inter-category variance between the two categories.
Preferably, the calculating the defect probability of the category according to the surface difference factor of the pixel point in the category and the occurrence probability of the gray value corresponding to the pixel point includes:
for any pixel point in the category, calculating the product of the surface difference factor of the pixel point and the occurrence probability of the gray value corresponding to the pixel point as a first defect factor; and calculating the sum of first defect factors of all pixel points in the category as the defect probability of the category.
Preferably, the calculating the probability average value of the category according to the surface difference factor of the pixel points in the category, the gray value corresponding to the pixel points and the occurrence frequency of the gray value includes:
subtracting the gray value corresponding to the pixel point from the preset maximum gray value to obtain an inverted gray value;
for any pixel point in the category, calculating a product of a surface difference factor of the pixel point, an inverted gray value corresponding to the pixel point and the occurrence frequency of the gray value as a second defect factor, and calculating a sum of the second defect factors of all the pixel points in the category; the ratio of the sum of the second defect factors to the defect probability is taken as the probability average of the category.
Preferably, the calculating the difference degree of the two gray histograms corresponding to the pixel points in the two categories includes:
Calculating the absolute value of the difference value of the frequencies of the same gray value in the two gray histograms to be used as an initial difference value; the sum of the initial difference values corresponding to the gray values in the gray histogram is used as the difference degree of the gray histogram corresponding to the two categories.
Preferably, the screening the initial defect area according to the gray value of the pixel point in the area corresponding to the sliding window includes:
and acquiring the gray average value of the outermost pixel point in the area corresponding to the sliding window, and taking the area corresponding to the sliding window as an initial defect area when the pixel value of any outermost pixel point in the area corresponding to the sliding window is smaller than the gray average value.
Preferably, the forming a color feature factor according to the values of each channel of the pixel point in the initial defect area in the LAB color space includes:
for any pixel point in the initial defect area, obtaining the sum of the channel values of the pixel point in the LAB color space; and opening the sum of the channel values to the power, wherein the obtained result value is a color characteristic factor corresponding to the pixel point.
Preferably, the obtaining the classification effect evaluation value from the inter-class variance and the degree of difference includes:
and taking the normalized product of the inter-class variance and the difference degree as a division effect evaluation value.
In a second aspect, an embodiment of the present invention provides an image data-based suspension spring defect detection system, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the image data-based suspension spring defect detection method described above when executing the computer program.
The embodiment of the invention has at least the following beneficial effects:
according to the method, firstly, the sliding window is slid on the obtained suspension spring surface image, the initial defect area is screened out, the defect area on the suspension spring surface image is subjected to preliminary screening, corresponding calculation cost is reduced in the follow-up defect identification process compared with other traditional global defect identification algorithms, and the real-time effect of the defect detection of the whole suspension spring is improved; obtaining a surface difference factor corresponding to the pixel point according to the color characteristic factor and the gray value of the pixel point in the initial defect area; the color characteristic factors reflect the color characteristics of the pixel points so as to facilitate the subsequent distinction of the defect area and the normal area on the surface image of the suspension spring, and the color characteristic factors and the gray values are combined so as to facilitate the subsequent more accurate judgment of whether the pixel points are in the defect area; randomly classifying pixel points in any initial defect area, and obtaining two categories by random classification each time; for random classification at any time, calculating the inter-class variance between two classes and the difference degree of gray histograms corresponding to the two classes, obtaining a classification effect evaluation value according to the inter-class variance and the difference degree, reflecting the difference condition of pixel points in the two classes according to the inter-class variance and the difference degree between the two classes, and when the difference of the pixel points in the two classes is larger, correspondingly reflecting the better classification effect of the two classes so as to be convenient for directly obtaining a defect area formed by defect classes in an initial defect area. According to the embodiment of the invention, the corresponding dividing effect evaluation value index is obtained by combining the characteristic construction of the surface defects of the suspension spring, and the final defect category is obtained according to the dividing effect evaluation value, so that the dividing effect is adjusted by designing the related dividing effect evaluation value, the complicated process that the threshold value needs to be set according to actual experience in the traditional threshold value segmentation algorithm is effectively avoided, and the effectiveness and accuracy of the whole scheme are improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method and a system for detecting a suspension spring defect based on image data according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description refers to the specific implementation, structure, characteristics and effects of the suspension spring defect detection method and system based on image data according to the present invention with reference to the accompanying drawings and the preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The embodiment of the invention provides a suspension spring defect detection method and a system specific implementation method based on image data. The scene detects the situation that the surface of the automobile suspension spring is defective and damaged due to high-frequency vibration, shrinkage and friction, and the CCD camera is used for collecting images of the suspension spring so as to conveniently detect the subsequent defects of the suspension spring. In order to solve the technical problem that the detection result of the suspension spring defect is affected when a proper experience threshold is required to be set by a traditional threshold segmentation algorithm. The method combines the characteristic construction of the surface defects of the suspension springs to obtain the corresponding index of the evaluation value of the dividing effect, obtains the final defect category according to the evaluation value of the dividing effect, realizes the adjustment of the dividing effect by designing the related evaluation value of the dividing effect, effectively avoids the complicated process of setting the threshold according to actual experience in the traditional threshold segmentation algorithm, and improves the effectiveness and accuracy of the whole scheme.
The following specifically describes a specific scheme of the suspension spring defect detection method and system based on image data provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart illustrating a method and a system for detecting a suspension spring defect based on image data according to an embodiment of the invention is shown, the method includes the following steps:
step S100, a suspension spring surface image is acquired.
The quality of the acquired image can greatly influence the accuracy of the detection of the defects of the subsequent suspension springs, and the quality of the acquired image is limited by corresponding image shooting and acquisition equipment. Compared with other CMOS electronic component cameras, the CCD camera has the advantages of sensitivity to light perception, obvious representation of related details of images acquired by the CCD camera, clear imaging quality and no smear. Meanwhile, compared with the area-array CCD camera, the area-array CCD camera is mature and reliable in related technology, and errors are not prone to occurring in the corresponding image acquisition process, so that in order to acquire suspension spring images with clear and reliable quality, the area-array CCD camera is selected for shooting and acquiring the surfaces of the suspension springs.
For the initial suspension spring surface image under RGB color space which can be obtained by shooting and collecting, the accuracy of the defect identification of the subsequent suspension springs is improved, and the corresponding calculation cost is reduced. And obtaining a suspension spring surface gray image by a weighted average method from the initial suspension spring surface image under the RGB color space obtained after shooting and acquisition.
The random natural noise possibly existing in the shooting and collecting process is considered to cause a certain degree of interference on the accuracy degree of the follow-up suspension spring surface defect identification. In order to reduce or even eliminate the interference of random noise on the identification of the surface defects of the suspension spring, a guiding filter is used for carrying out noise reduction treatment on the gray level image of the surface of the suspension spring, so that a first suspension spring surface image after noise reduction is obtained.
Meanwhile, in order to avoid the phenomenon that brightness distribution of different areas is uneven due to uneven illumination during shooting and acquisition to cause larger influence on the detection of the defects of the follow-up suspension springs, a histogram equalization algorithm is used for processing the surface images of the first suspension springs to obtain the surface images of the suspension springs, and the influence of the uneven illumination on the detection of the defects of the follow-up suspension springs is reduced.
Finally, in order to facilitate extraction of color characteristic information at different positions of the suspension spring surface, the suspension spring surface image needs to be converted to obtain an LAB suspension spring image in a corresponding LAB color space.
Thus, the obtained suspension spring surface image and the LAB suspension spring image under the corresponding LAB color space can be obtained.
Step S200, sliding a sliding window on the surface image of the suspension spring, and screening out an initial defect area according to the gray value of the pixel point in the area corresponding to the sliding window; forming color characteristic factors according to the values of all channels of the pixel points in the initial defect area in the LAB color space; and for each pixel point in the initial defect area, obtaining a corresponding surface difference factor according to the color characteristic factor and the gray value.
And analyzing and calculating the acquired suspension spring surface image, and judging the suspension spring surface image. The method comprises the following steps: (1) And sliding a sliding window on the suspension spring surface image, and screening out an initial defect area according to the gray value of the pixel point in the area corresponding to the sliding window. (2) And forming color characteristic factors according to the values of all channels of the pixel points in the initial defect area in the LAB color space. (3) And for any pixel point in the initial defect area, obtaining a corresponding surface difference factor according to the color characteristic factor and the gray value.
The specific development is as follows:
(1) And sliding a sliding window on the suspension spring surface image, and screening out an initial defect area according to the gray value of the pixel point in the area corresponding to the sliding window.
In order to perform further relevant defect detection analysis on the suspension spring, it is necessary to perform preliminary rough judgment on the surface image of the suspension spring to determine whether a defect exists, further analyze and calculate the relevant defect on the surface of the suspension spring to be detected if the relevant defect exists, and not further calculate the relevant defect otherwise, so that the overall real-time response effect of the system is improved.
A sliding window of 9*9 is arranged, the sliding window is slid on the surface image of the suspension spring, the step length of each sliding is 1, namely, each pixel point is taken as the center for sliding, and the step length of each sliding is 1. It is noted that in other embodiments the practitioner may adjust the step size of the sliding as the case may be. Counting the gray values of the pixel points at corresponding positions of the pixel points in the area corresponding to the whole sliding window according to the clockwise direction, and screening out an initial defect area according to the gray values of the pixel points in the area corresponding to the sliding window, wherein the gray values are specific: acquiring the gray average value of the outermost pixel point in the area corresponding to the sliding window, and when the pixel value of any outermost pixel point in the area corresponding to the sliding window is smaller than the gray average value The area corresponding to the sliding window is used as an initial defect area. That is, the encoding is given to the outermost pixel point in the region corresponding to the sliding window, and since the size of the sliding window is 9*9, the encoding of the outermost pixel point in the region corresponding to the sliding window is from 1 to 36. And constructing a suspension spring discrimination line graph, wherein the horizontal axis of the suspension spring discrimination line graph is the code of the outer-most pixel point arranged clockwise from the pixel point at the upper left corner in the area corresponding to the sliding window, and the vertical axis is the gray value of the pixel point corresponding to the code. In the embodiment of the invention, an empirical gray scale quantization value is adopted, and the gray scale value of the corresponding pixel point is positioned in the interval
And (3) upper part. Acquiring the gray average value of the pixel points at the outermost ring in the area corresponding to the sliding window, constructing a gray average value line taking the gray average value as an ordinate, and considering that a valley appears when a point lower than the gray average value line exists in the suspension spring judging line graph, and calling the point lower than the gray average value line as the valley point; and taking the area of the sliding window corresponding to the suspension spring discrimination line graph with the low valley point as an initial defect area.
The initial defect area can be screened out by carrying out preliminary detection and judgment on whether the surface of the suspension spring has defects or not through the corresponding suspension spring judging line diagram constructed as described above.
When a defect occurs on the surface of the suspension spring, the corresponding defect position should be darker than the surrounding pixels. Therefore, three situations can occur in the region corresponding to the sliding window constructed by the embodiment of the invention. Case one: the region contains complete abrasion defects, the gray value changes sharply at two sides of the edge of the complete abrasion defect of the suspension spring, and two valleys appear in the corresponding suspension spring judging line diagram; and a second case: the region contains partial wear defect, and a valley appears in the corresponding suspension spring judging line diagram; and a third case: when the region does not contain abrasion defects, no valley appears in the corresponding suspension spring fold line judging graph. In the case of the suspension spring discrimination line graphs of different situations, marking the sliding window corresponding to the valley as 1, wherein the sliding window is used as an initial defect area, namely the area of the sliding window corresponding to the suspension spring discrimination line graph with the low valley point is used as the initial defect area, namely when the pixel value of any pixel point at the outermost ring in the area corresponding to the sliding window is smaller than the gray level average value, the area corresponding to the sliding window is used as the initial defect area; whereas the sliding window area where no valley occurs is marked as 0.
If the marks of all sliding windows are 0 finally, the possibility of abrasion defect on the surface of the suspension spring is low; otherwise, if the mark of the final sliding window is 1, it is indicated that the sliding window has a relatively high possibility of abrasion defect in the corresponding position area, and further analysis and calculation are needed, so that the suspension spring is subjected to relatively accurate defect detection, namely, the initial defect area is subjected to further defect detection.
(2) And forming color characteristic factors according to the values of all channels of the pixel points in the initial defect area in the LAB color space.
According to the analysis and calculation in the step (1), the surface image of the suspension spring can be subjected to preliminary screening, and if relevant abrasion defects exist, the surface image of the suspension spring can be further analyzed and calculated. And obtaining corresponding window areas of the defect positions according to the windows marked with 1, and further judging and analyzing the window areas.
For a size of
And the sliding window on the surface image of the suspension spring with the abrasion defect marked as 1, namely for the initial defect area, calculating the numerical value of the corresponding surface difference factor according to the following formula for each pixel point in the initial defect area. Since the color at the position where the defect occurs on the surface of the suspension spring will have a certain difference in color from the surrounding normal pixel points, in order to calculate the difference, the corresponding color characteristic information of the surface of the suspension spring is constructed. Specific: and forming color characteristic factors according to the values of all channels in the LAB suspension spring image of the pixel points in the initial defect area under the LAB color space. The method for acquiring the color characteristic factors comprises the following steps: for any pixel point in the initial defect area, obtaining the sum of the channel values of the pixel point in the LAB color space; for the said The sum of the channel values is square, and the obtained result value is the color characteristic factor corresponding to the pixel point.
The calculation formula of the color characteristic factor is as follows:
wherein ,
is given by the coordinates
Color feature factors of the pixel points of (a);
is given by the coordinates
The value of the L channel of the pixel point of (2) in the LAB color space;
is given by the coordinates
The number of the A channel of the pixel point of (2) in the LAB color space;
is given by the coordinates
The number of B channels in LAB color space.
For the pixel point on the suspension spring surface image, when the pixel point belongs to a defective pixel point, the numerical value of each channel of the pixel point in the LAB color space is larger than that of each channel of a normal pixel point, so that the numerical values of the LAB channels corresponding to the pixel point are directly summed and squared again to reflect the color characteristic factors corresponding to the pixel point. The larger the probability that the pixel is a defective pixel, the larger the value of each channel of the corresponding LAB, and the larger the color characteristic factor of the corresponding pixel, otherwise, the larger the probability that the pixel is a normal pixel, the smaller the value of each channel of the corresponding LAB, and the smaller the color characteristic factor of the corresponding pixel.
(3) And for any pixel point in the initial defect area, obtaining a corresponding surface difference factor according to the color characteristic factor and the gray value.
The numerical value of the color characteristic information of the pixel points at different positions on the surface of the suspension spring under the corresponding sliding window can be calculated through a calculation formula of the color characteristic factors. Specifically, when the calculated position is at the suspension spring surface defect position, the smaller the calculated color feature value is, the greater the probability of defect occurrence at the corresponding pixel point position. And a scientific judgment basis is provided for the subsequent further division of different pixel points on the surface of the suspension spring. Further, for any pixel point in the initial defect area, a corresponding surface difference factor is obtained according to the color characteristic factor and the gray value. Specific: selecting any pixel point in the initial defect area as a target pixel point, and calculating a gray level difference value of gray level values of the target pixel point and the pixel points in the eight neighborhood; calculating the sum of absolute values of gray differences between the pixel points in the eight adjacent domains and the target pixel point; the ratio of the gray level difference value to the sum of the absolute values is a weight stabilizing coefficient of the target pixel point; calculating the product of the gray difference value of any pixel point in the eight neighborhood and the weight stabilizing coefficient to be used as a first product; taking the sum of the first products of the pixel points in the eight neighborhoods as an initial difference factor; and calculating the product of the color characteristic factor and the initial difference factor to be used as a second product, and taking the normalized second product as a surface difference factor.
The calculation formula of the surface difference factor is as follows:
wherein ,
is given by the coordinates
A surface difference factor of the pixel points of (a);
is a normalization function;
is given by the coordinates
Color feature factors of the pixel points of (a);
is given by the coordinates
Gray value of the ith pixel point in the eight adjacent areas corresponding to the pixel point;
is given by the coordinates
Gray values of pixels of (a). In the embodiment of the invention, the normalization is carried out by adopting a range normalization method.
Wherein, in the calculation formula of the surface difference factor
And the noise pixel points in the sliding window in the calculation process are prevented from causing abnormal influence on the whole numerical value by introducing the weight stabilizing coefficient as the weight stabilizing coefficient. According to the analysis, corresponding gray amplitude changes can occur at the positions of the defect edges in the surface image of the suspension spring, and corresponding surface difference factors can be calculated for the pixel points at each different position in the initial defect area screened out in the surface image of the suspension spring
To judge the difference between the pixel points at different positions and the adjacent pixel points, and further judge whether the pixel points are positioned at the defect positions of the surface of the suspension spring. When the pixel point is positioned at the defect position of the surface of the suspension springWhen the method is put, the calculated surface difference factors are correspondingly calculated 
The value of (2) will be correspondingly smaller; otherwise, when the pixel point is positioned at the normal area position of the surface of the suspension spring, the calculated surface difference factor is correspondingly calculated
The value of (2) will be larger.
Step S300, carrying out random classification on pixel points in each initial defect area at least twice, and obtaining two categories by each random classification; for two categories obtained by random classification at any time, calculating the inter-category variance between the two categories according to the surface difference factors of the pixel points in the two categories, the gray values corresponding to the pixel points and the occurrence frequencies of the gray values; and calculating the difference degree of the two gray histograms corresponding to the pixel points in the two categories.
And dividing the pixel points in the initial defect area by combining the characteristic information of the pixel points in the initial defect area on the surface image of the suspension spring.
For a size of
The pixel points can be divided into two categories, one category is the category of the normal pixel points of the suspension spring; the other is the pixel point class of the suspension spring defect area. Based on the above, the pixel points in the window can be divided by combining the corresponding pixel point characteristic information obtained by the analysis and calculation, so as to obtain the corresponding suspension spring surface defect area.
First, at the corresponding size of
Randomly selecting a part of pixels in a suspension spring window area, marking the selected pixels as initial defective area pixels, and marking the number of the pixels in a corresponding area as
At the same time, the number of the pixel points of the other part is recorded as
. Namely, randomly classifying pixel points in any initial defect area, and obtaining two categories by random classification each time;
the following correlation calculation and analysis is performed for the two categories obtained by the initial random division. First, the occurrence frequency corresponding to each gray value in the initial defect area is obtained. The calculation formula of the occurrence frequency is as follows:
, wherein ,
the occurrence frequency of the gray value corresponding to the ith pixel point;
the number of the pixels is the same as the gray value of the ith pixel;
is the total number of pixels in the entire initial defective area. Since the size of the sliding window is 9*9 in the embodiment of the present invention, the total number of pixels in the initial defective area is 81, and in other embodiments, the size of the sliding window set by the embodiment adapts to the total number of pixels in the initial defective area. And carrying out subsequent distribution on the pixel points at all different positions in the initial defect area. And has the following relation 
、
。
Further, for two categories obtained by random classification at any time, calculating the inter-category variance between the two categories according to the surface difference factors of the pixel points in the two categories, the gray values corresponding to the pixel points and the occurrence frequencies of the gray values; and then constructing gray histograms corresponding to pixel points in the two categories, and calculating the difference degree of the gray histograms corresponding to the two categories. To further calculate the classification effect evaluation values of the two categories obtained by the random classification. The greater the evaluation value of the dividing effect is, the clearer the corresponding classification of the two categories is, wherein the probability that all the categories are pixel points in the defect area is higher, and the probability that the other category is pixel points in the normal area is higher.
The method for obtaining the inter-class variance comprises the following steps: calculating the defect probability of the category according to the surface difference factor of the pixel points in the category and the occurrence probability of the gray value corresponding to the pixel points; and calculating the probability mean value of the class according to the surface difference factor of the pixel points in the class, the gray value corresponding to the pixel points and the occurrence frequency of the gray value. The method for acquiring the defect probability of the category comprises the following steps: for any pixel point in the category, calculating the product of the surface difference factor of the pixel point and the occurrence probability of the gray value corresponding to the pixel point as a first defect factor; and calculating the sum of first defect factors of all pixel points in the category as the defect probability of the category.
The calculation formula of the defect probability is as follows:
wherein ,
the defect probability corresponding to the nth category;
the number of pixels in the nth class;
the surface difference factor corresponding to the ith pixel point;
the frequency of occurrence of the gray value corresponding to the i-th pixel point.
According to the corresponding calculated surface difference factor when the pixel point is positioned at the defect position of the surface of the suspension spring
The numerical values of (2) will be phaseShould be small; otherwise, when the pixel point is positioned at the normal area position of the surface of the suspension spring, the calculated surface difference factor is correspondingly calculated
The larger the relation of the numerical value of the pixel point, the more the surface difference factor of the pixel point is, the larger the probability that the pixel point belongs to the pixel point in the defect area.
The calculation formula of the probability mean value of the category is as follows: subtracting the gray value corresponding to the pixel point from the preset maximum gray value to obtain the inverted gray value. For any pixel point in the category, calculating a product of a surface difference factor of the pixel point, an inverted gray value corresponding to the pixel point and the occurrence frequency of the gray value as a second defect factor, and calculating a sum of the second defect factors of all the pixel points in the category; the ratio of the sum of the second defect factors to the defect probability is taken as the probability average of the category.
The calculation formula of the probability mean value is as follows:
wherein ,
the probability mean value corresponding to the nth category;
the number of pixels in the nth class;
the inverse gray value corresponding to the ith pixel point is obtained;
the surface difference factor corresponding to the ith pixel point;
the occurrence frequency of the gray value corresponding to the ith pixel point;
and the defect probability corresponding to the nth category.
Further, after the defect probability corresponding to the category is calculated, the probability average value corresponding to the category is calculated by combining the defect probability corresponding to the category, the denominator of the calculation formula of the probability average value is the probability average value, and the numerator is the occurrence frequency corresponding to the gray value and the surface difference factor used for calculating the probability average value, and the inverse gray value of the pixel point is further combined, so that the probability average value corresponding to the category is calculated. The purpose of subtracting the gray value from the preset maximum gray value to obtain the inverted gray value is to make the inverted gray value of the pixel with larger defect probability be in direct proportion to the situation that the pixel is the defective pixel when the probability average value is calculated later, that is, the larger the inverted gray value of the pixel in the category is, the larger the probability average value corresponding to the category is. The probability average reflects the probability that the pixel points in the category are defective pixel points to a certain extent. In the embodiment of the present invention, the preset maximum gray value is 255, and in other embodiments, the practitioner can adjust the value according to the actual situation.
And taking the defect probabilities of the two categories as weights, and carrying out weighted summation on the probability average values of the two categories to obtain the regional probability average value of the initial defect region.
The calculation formula of the regional probability mean value is as follows:
wherein ,
the regional probability average value of the initial defect region;
the defect probability corresponding to the first category in the initial defect area is determined;
the probability average value corresponding to the first category in the initial defect area is obtained;
the defect probability corresponding to the second category in the initial defect area;
the probability average value corresponding to the second category in the initial defect area.
The regional probability average value is obtained by weighted summation of the defect probability and the probability average value of two categories corresponding to the initial defect region, wherein the defect probability and the regional probability average value are in positive correlation, and when the defect probability corresponding to one category in the initial defect region is larger, the regional probability average value corresponding to the region is larger as a whole, and similarly, when the defect probability corresponding to one category is smaller, the regional probability average value corresponding to the region is smaller as a whole. When the probability average value corresponding to one category in the initial defect area is larger, the area is taken as a whole, the probability average value of the corresponding area is larger, and similarly, when the probability average value corresponding to one category is smaller, the area is taken as a whole, and the probability average value of the corresponding area is smaller.
Namely, for the two classes obtained by initial division, the defect probability corresponding to the first class and the corresponding probability average value, and the defect probability of the second class and the numerical value of the corresponding probability average value can be calculated by the formula,
the size of the regional probability mean value corresponding to the whole initial defect region. Further, respectively calculating squares of differences between the probability average values of the two categories and the regional probability average value to serve as an average value difference factor; and taking the defect probability of the two categories as a weight, and carrying out weighted summation on the mean difference factors corresponding to the two categories to obtain the inter-category variance between the two categories.
The calculation formula of the variance among the classes is as follows:
wherein ,
the inter-class variance corresponding to the initial defect area;
the regional probability average value of the initial defect region;
the defect probability corresponding to the first category in the initial defect area is determined;
the probability average value corresponding to the first category in the initial defect area is obtained;
the defect probability corresponding to the second category in the initial defect area;
the probability average value corresponding to the second category in the initial defect area.
The magnitude of the value corresponding to the inter-class variance between the two classes can be calculated according to the calculation formula of the inter-class variance, and when the magnitude of the calculated inter-class variance is larger, the larger the difference between the corresponding defective pixel point class and the normal pixel point class in the two classes is, namely the clearer the corresponding two classes are divided, the pixel points in the defective area are all in one class, and the larger the probability of the pixel points in the normal area is the other class, the better the clustering effect is.
Respectively constructing two different categories obtained by dividing to obtain gray histograms corresponding to the different categories; and calculating the difference degree of the gray level histograms corresponding to the two categories. It should be noted that, constructing the gray level histogram corresponding to the category according to the pixel points in the category is a well known technique of those skilled in the art, and will not be described herein. Calculating the difference degree of gray histograms corresponding to two categories, and specifically: calculating the absolute value of the difference value of the frequencies of the same gray value in the two gray histograms to be used as an initial difference value; the sum of the initial difference values corresponding to the gray values in the gray histogram is used as the difference degree of the gray histogram corresponding to the two categories.
The calculation formula of the difference degree is as follows:
wherein ,
the degree of difference for the initial defect region;
the maximum gray value in the gray histograms corresponding to the two categories is obtained;
the frequency corresponding to the gray value k in the first category;
is the corresponding frequency for the gray value k in the second category.
The difference degree of the histograms corresponding to two different categories can be calculated by the calculation formula of the difference degree
Is a numerical value of (a). The difference degree of the histograms corresponding to the two categories reflects the classification condition of the two categories, and the better the classification effect is, the greater the difference degree is. That is, when the effect of the two classification is better, the difference in the corresponding gray values should be larger, the degree of difference of the calculated histograms 
The larger the value of (2) should be, whereas when the two classes are less effective, the corresponding gray value isThe smaller the gap should be, the smaller the value of the degree of difference in the histograms of the two categories calculated.
Step S400, obtaining a classification effect evaluation value according to the inter-class variance and the degree of difference, and obtaining the class with larger gray average value in the two classes corresponding to the maximum classification effect evaluation value as a defect class; and forming a suspension spring defect area by pixel points in each defect category.
Obtaining a division effect evaluation value by the inter-class variance and the difference degree, and specifically: taking the product of the normalized inter-class variance and the difference degree as the dividing effect evaluation value. Namely multiplying the inter-class variance and the degree of difference, normalizing the multiplied result value, and taking the normalized value as the dividing effect evaluation value.
The calculation formula of the evaluation value of the dividing effect is as follows:
wherein ,
the evaluation value is divided into effect;
is the inter-class variance;
is the degree of difference;
is a normalization function.
The difference between the two classes and the difference degree corresponding to the two classes reflect the difference condition of the two classes, and when the difference between the two classes is larger, the better the two classes divide the defective pixel point and the normal pixel point, the smaller the difference corresponding to the two classes reflects that the two classes obtained by dividing in this way do not divide the defective pixel point and the normal pixel point well.
The current division can be calculated to obtain corresponding two categories of divisionsScore evaluation value
According to the above calculation and analysis process, when the evaluation value of the division effect is maximum, the division of the defective pixel point and the normal pixel point in the corresponding window achieves the best effect. Therefore, two categories corresponding to the maximum value of the evaluation value of the dividing effect are obtained as final categories, and the category with the larger gray value mean value in the final categories is the defect category. And then find out and divide the best effect evaluation value of effect and correspond to two kinds, and distinguish the correspondent defect classification, the pixel point in the defect classification is defective pixel point, form the defective area of the suspension spring by the pixel point in the defect classification, finish the defect detection to the suspension spring.
In summary, the present invention relates to the technical field of image processing, and the method includes firstly acquiring an image of a surface of a suspension spring; sliding a sliding window on the surface image of the suspension spring, and screening out an initial defect area according to the gray value of the pixel point in the area corresponding to the sliding window; forming color characteristic factors according to the values of all channels of pixel points in the initial defect area in the LAB color space; for any pixel point in the initial defect area, a corresponding surface difference factor is obtained according to the color characteristic factor and the gray value; randomly classifying pixel points in any initial defect area, and obtaining two categories by random classification each time; for two categories obtained by random classification at any time, calculating the inter-category variance between the two categories according to the surface difference factors of the pixel points in the two categories, the gray values corresponding to the pixel points and the occurrence frequencies of the gray values; calculating the difference degree of two gray histograms corresponding to pixel points in two categories; obtaining a classification effect evaluation value according to the inter-class variance and the difference degree, and obtaining the class with larger gray average value from two classes corresponding to the maximum classification effect evaluation value as a defect class; the suspension spring defect area is formed by pixel points in the defect category.
The embodiment of the invention also provides a suspension spring defect detection system based on image data, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the steps of the method when executing the computer program. Since the detailed description of the suspension spring defect detection method based on the image data is given above, the detailed description is omitted.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings 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.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (6)
1. The suspension spring defect detection method based on the image data is characterized by comprising the following steps of:
acquiring an image of the surface of the suspension spring;
Sliding a sliding window on the suspension spring surface image, and screening out an initial defect area according to the gray value of a pixel point in the area corresponding to the sliding window; forming color characteristic factors according to the values of all channels of the pixel points in the initial defect area in the LAB color space; for each pixel point in the initial defect area, obtaining a corresponding surface difference factor according to the color characteristic factor and the gray value;
randomly classifying the pixel points in each initial defect area at least twice, and obtaining two categories by random classification each time; for two categories obtained by random classification at any time, calculating the inter-category variance between the two categories according to the surface difference factors of the pixel points in the two categories, the gray values corresponding to the pixel points and the occurrence frequencies of the gray values; calculating the difference degree of two gray histograms corresponding to pixel points in two categories;
obtaining a classification effect evaluation value according to the inter-class variance and the degree of the difference, and obtaining the class with larger gray average value from two classes corresponding to the maximum classification effect evaluation value as a defect class; forming a suspension spring defect area by pixel points in each defect category;
the method comprises the steps of calculating the inter-class variance between two classes according to the surface difference factors of the pixel points in the two classes, the gray values corresponding to the pixel points and the occurrence frequencies of the gray values, wherein the steps are as follows: calculating the defect probability of the category according to the surface difference factor of the pixel points in the category and the occurrence probability of the gray value corresponding to the pixel points; calculating the probability mean value of the class according to the surface difference factor of the pixel points in the class, the gray value corresponding to the pixel points and the occurrence frequency of the gray value; taking the defect probability of the two categories as a weight, and carrying out weighted summation on the probability average values of the two categories to obtain a regional probability average value of the initial defect region; respectively calculating squares of differences between the probability average values of the two categories and the regional probability average value to serve as an average value difference factor; taking the defect probability of the two categories as a weight, and carrying out weighted summation on the mean difference factors corresponding to the two categories to obtain an inter-category variance between the two categories;
The method for acquiring the color characteristic factors comprises the following steps: for any pixel point in the initial defect area, obtaining the sum of the channel values of the pixel point in the LAB color space; the sum of the channel values is divided to the power, and the obtained result value is a color characteristic factor corresponding to the pixel point;
the calculation formula of the surface difference factor is as follows:
wherein ,
for the coordinates +.>
A surface difference factor of the pixel points of (a); />
Is a normalization function; />
For the coordinates +.>
Color feature factors of the pixel points of (a); />
For the coordinates +.>
Gray value of the ith pixel point in the eight adjacent areas corresponding to the pixel point; />
For the coordinates +.>
Gray values of the pixels of (a);
the calculation formula of the evaluation value of the dividing effect is as follows:
wherein ,
the evaluation value is divided into effect; />
Is the inter-class variance; />
Is the degree of difference; />
Is a normalization function.
2. The method for detecting defects of a suspension spring based on image data according to claim 1, wherein the calculating the defect probability of the category according to the surface difference factor of the pixel points in the category and the occurrence probability of the gray value corresponding to the pixel points comprises:
for any pixel point in the category, calculating the product of the surface difference factor of the pixel point and the occurrence probability of the gray value corresponding to the pixel point as a first defect factor; and calculating the sum of first defect factors of all pixel points in the category as the defect probability of the category.
3. The method for detecting defects of a suspension spring based on image data according to claim 1, wherein calculating a probability mean value of a class according to a surface difference factor of pixel points in the class, gray values corresponding to the pixel points, and occurrence frequencies of the gray values, comprises:
subtracting the gray value corresponding to the pixel point from the preset maximum gray value to obtain an inverted gray value;
for any pixel point in the category, calculating a product of a surface difference factor of the pixel point, an inverted gray value corresponding to the pixel point and the occurrence frequency of the gray value as a second defect factor, and calculating a sum of the second defect factors of all the pixel points in the category; the ratio of the sum of the second defect factors to the defect probability is taken as the probability average of the category.
4. The method for detecting defects of a suspension spring based on image data according to claim 1, wherein the calculating the degree of difference between two gray histograms corresponding to pixels in two categories comprises:
calculating the absolute value of the difference value of the frequencies of the same gray value in the two gray histograms to be used as an initial difference value; the sum of the initial difference values corresponding to the gray values in the gray histogram is used as the difference degree of the gray histogram corresponding to the two categories.
5. The method for detecting defects of a suspension spring based on image data according to claim 1, wherein the step of screening out an initial defect area according to gray values of pixel points in an area corresponding to a sliding window comprises the steps of:
and acquiring the gray average value of the outermost pixel point in the area corresponding to the sliding window, and taking the area corresponding to the sliding window as an initial defect area when the pixel value of any outermost pixel point in the area corresponding to the sliding window is smaller than the gray average value.
6. The suspension spring defect detection system based on image data comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the steps of the suspension spring defect detection method based on image data according to any one of claims 1-5 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310181114.XA CN115861325B (en) | 2023-03-01 | 2023-03-01 | Suspension spring defect detection method and system based on image data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310181114.XA CN115861325B (en) | 2023-03-01 | 2023-03-01 | Suspension spring defect detection method and system based on image data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115861325A CN115861325A (en) | 2023-03-28 |
| CN115861325B true CN115861325B (en) | 2023-06-20 |
Family
ID=85659417
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310181114.XA Active CN115861325B (en) | 2023-03-01 | 2023-03-01 | Suspension spring defect detection method and system based on image data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115861325B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116205906B (en) * | 2023-04-25 | 2023-07-18 | 青岛豪迈电缆集团有限公司 | Nondestructive testing method for production abnormality in cable |
| CN116246174B (en) * | 2023-04-26 | 2023-08-08 | 山东金诺种业有限公司 | Sweet potato variety identification method based on image processing |
| CN116320467B (en) * | 2023-05-19 | 2023-07-25 | 山东中科冶金矿山机械有限公司 | Geological survey data compression storage method |
| CN116760952B (en) * | 2023-08-17 | 2023-10-20 | 山东欣晖电力科技有限公司 | Unmanned aerial vehicle-based electric power iron tower maintenance inspection method |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115311292A (en) * | 2022-10-12 | 2022-11-08 | 南通创铭伊诺机械有限公司 | Strip steel surface defect detection method and system based on image processing |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105740844A (en) * | 2016-03-02 | 2016-07-06 | 成都翼比特自动化设备有限公司 | Insulator cracking fault detection method based on image identification technology |
| CN110414538B (en) * | 2019-07-24 | 2022-05-27 | 京东方科技集团股份有限公司 | Defect classification method, defect classification training method and device thereof |
| CN112200800B (en) * | 2020-10-30 | 2022-10-28 | 福州大学 | Electrowetting display defect detection method based on gray level histogram |
| CN115294158A (en) * | 2022-09-26 | 2022-11-04 | 江苏万森绿建装配式建筑有限公司 | Hot continuous rolling strip steel image segmentation method based on machine vision |
| CN115272335A (en) * | 2022-09-29 | 2022-11-01 | 江苏万森绿建装配式建筑有限公司 | Metallurgical metal surface defect detection method based on significance detection |
| CN115351598B (en) * | 2022-10-17 | 2024-01-09 | 安徽金锘轴承制造有限公司 | Method for detecting bearing of numerical control machine tool |
-
2023
- 2023-03-01 CN CN202310181114.XA patent/CN115861325B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115311292A (en) * | 2022-10-12 | 2022-11-08 | 南通创铭伊诺机械有限公司 | Strip steel surface defect detection method and system based on image processing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115861325A (en) | 2023-03-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115861325B (en) | Suspension spring defect detection method and system based on image data | |
| CN114842017B (en) | HDMI cable surface quality detection method and system | |
| WO2007101269A1 (en) | Approaches and apparatus for eye detection in a digital image | |
| CN116758083B (en) | Quick detection method for metal wash basin defects based on computer vision | |
| CN108563979B (en) | Method for judging rice blast disease conditions based on aerial farmland images | |
| CN116385450B (en) | PS sheet wear resistance detection method based on image processing | |
| CN115239738B (en) | Intelligent detection method for automobile part configuration | |
| CN114387273A (en) | Environmental dust concentration detection method and system based on computer image recognition | |
| CN112734674A (en) | Image denoising method | |
| CN116012357A (en) | Automobile circular trailer hole air line defect detection method based on machine vision | |
| CN117095004B (en) | Excavator walking frame main body welding deformation detection method based on computer vision | |
| CN115909256B (en) | Road disease detection method based on road visual image | |
| CN116883408B (en) | Integrating instrument shell defect detection method based on artificial intelligence | |
| CN115018785A (en) | Hoisting steel wire rope tension detection method based on visual vibration frequency identification | |
| CN117522719B (en) | Bronchoscope image auxiliary optimization system based on machine learning | |
| CN110378271B (en) | Gait recognition equipment screening method based on quality dimension evaluation parameters | |
| CN115100510B (en) | Tire wear degree identification method | |
| CN115512310A (en) | Vehicle type recognition method and system based on face features under video monitoring | |
| CN111611930B (en) | Parking space line detection method based on illumination consistency | |
| CN115457489A (en) | Vehicle type identification method based on video monitoring | |
| CN114373147A (en) | Detection method for low-texture video license plate | |
| US8699761B2 (en) | Method for evaluating quality of image representing a fingerprint pattern | |
| CN116597432B (en) | License plate recognition system based on improved yolov5 algorithm | |
| CN116958182B (en) | Quick concrete crack detection method based on image data | |
| TW202008215A (en) | Object attribute analyzing method and related object attribute analyzing device |
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 | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Suspension Spring Defect Detection Method and System Based on Image Data Effective date of registration: 20230630 Granted publication date: 20230620 Pledgee: China postal savings bank, Surabaya County branch, Limited by Share Ltd. Pledgor: Shandong Zhongke Metallurgical Mining Machinery Co.,Ltd.|Jining Zhongke Mining Machinery Factory Registration number: Y2023980046498 |