CN115861320B - Intelligent detection method for automobile part machining information - Google Patents
Intelligent detection method for automobile part machining information Download PDFInfo
- Publication number
- CN115861320B CN115861320B CN202310171997.6A CN202310171997A CN115861320B CN 115861320 B CN115861320 B CN 115861320B CN 202310171997 A CN202310171997 A CN 202310171997A CN 115861320 B CN115861320 B CN 115861320B
- Authority
- CN
- China
- Prior art keywords
- area
- target
- pixel points
- gray
- target area
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 20
- 238000003754 machining Methods 0.000 title claims description 3
- 230000007547 defect Effects 0.000 claims abstract description 180
- 239000006247 magnetic powder Substances 0.000 claims abstract description 66
- 238000004220 aggregation Methods 0.000 claims abstract description 43
- 230000002776 aggregation Effects 0.000 claims abstract description 43
- 238000000034 method Methods 0.000 claims abstract description 27
- 230000002950 deficient Effects 0.000 claims abstract description 25
- 238000009826 distribution Methods 0.000 claims abstract description 22
- 238000002347 injection Methods 0.000 claims abstract description 5
- 239000007924 injection Substances 0.000 claims abstract description 5
- 238000013507 mapping Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 description 6
- 239000000758 substrate Substances 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000005242 forging Methods 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 238000005204 segregation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000004080 punching Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 239000002344 surface layer Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 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 Processing (AREA)
- Investigating Or Analyzing Materials By The Use Of Magnetic Means (AREA)
Abstract
The invention relates to the technical field of image data processing, in particular to an intelligent detection method for automobile part processing information. The method comprises the steps of obtaining a gray image of the bearing surface after fluorescent magnetic powder injection; obtaining a suspected defect area in the gray level image through a watershed algorithm; obtaining a pixel point corresponding to the maximum gray value in each suspected defect area as a target pixel point, clustering the pixel points in each suspected defect area, and taking the area of a cluster where the target pixel point is located as a target area; acquiring the aggregation degree of fluorescent magnetic powder of each target area according to the area of the suspected defect area corresponding to the target area and the position distribution and gradient distribution of the pixel points in the target area; acquiring a gray level difference between a target area and a gray level image as a target difference; obtaining the similarity between the target area and the background area according to the aggregation degree of the fluorescent magnetic powder and the target difference; and determining a defect area in the gray image according to the similarity. The efficiency of detecting defective areas is improved.
Description
Technical Field
The invention relates to the technical field of image data processing, in particular to an intelligent detection method for automobile part processing information.
Background
The bearing is a very critical component in automobile parts, and the quality of the bearing relates to the running safety of the automobile. In the production and processing process of the bearing, defects can be generated due to the fact that the forging heating temperature is too high or the heat preservation time is too long, and grain boundaries are oxidized and even melted in severe cases. The overburden bearing is forged under the defect state, and is subjected to forging, punching and rolling by a heavy hammer, so that the defect part can be torn to form larger defects.
In the prior art, defects on the surface of a bearing are detected by a fluorescent magnetic powder method, the fluorescent magnetic powder is attached to the defect area on the surface of the bearing to form a fluorescent area, and the fluorescent area can be segmented by a watershed algorithm commonly used in the field of image segmentation. However, the fine fluorescent magnetic powder is attached to the uneven part of the surface of the bearing to form a fluorescent region, so that when the bearing is segmented by a watershed algorithm, the area of the suspected defect region reserved after flooding is larger due to the larger fluorescent region, and the suspected defect region comprises the uneven region with partial defects and the defect region, so that the segmented defect region is inaccurate, the identification of the defect region is interfered, and the detection of the defect region is not facilitated.
Disclosure of Invention
In order to solve the technical problem of inaccurate detection of a defect area caused by interference of the uneven area of the bearing to the defect area, the invention aims to provide an intelligent detection method for processing information of an automobile part, and the adopted technical scheme is as follows:
the embodiment of the invention provides an intelligent detection method for automobile part processing information, which comprises the following steps:
obtaining a gray image of the bearing surface after fluorescent magnetic powder injection;
obtaining a suspected defect area in the gray level image through a watershed algorithm; obtaining pixel points corresponding to the maximum gray value in each suspected defect area as target pixel points, clustering the pixel points in each suspected defect area, and taking the area of the cluster where the target pixel points are located as a target area;
acquiring the aggregation degree of fluorescent magnetic powder of each target area according to the area of the suspected defect area corresponding to the target area and the position distribution and gradient distribution of pixel points in the target area;
acquiring a gray scale difference between the target area and the gray scale image as a target difference; obtaining the similarity between the target area and the background area according to the aggregation degree of the fluorescent magnetic powder and the target difference;
and merging the suspected defect area corresponding to the target area with the background area according to the similarity, and determining the defect area in the gray level image.
Further, the method for obtaining the aggregation degree of the fluorescent magnetic powder comprises the following steps:
acquiring coordinates of pixel points corresponding to the maximum gray value in the target area as first coordinates, calculating the distance between the coordinates corresponding to each pixel point in the target area and the first coordinates, and acquiring the average distance between the pixel points in the target area; carrying out negative correlation mapping and normalization on the average distance, and taking the obtained result as a first result;
acquiring a gradient mean value of edge pixel points in the target area as an average gradient;
acquiring the number of pixel points in the suspected defect area corresponding to the target area, namely the area of the suspected defect area, as a first area;
and calculating the product of the first result, the average gradient and the first area as the aggregation degree of the fluorescent magnetic powder of the target area.
Further, the method for acquiring the target difference comprises the following steps:
calculating an average gray value in the target area as a first value;
calculating an average gray value in the gray image as a second value;
and taking the difference value between the first value and the second value as a target difference between the target area and the gray scale image.
Further, the method for obtaining the similarity comprises the following steps:
and taking the result of carrying out negative correlation mapping and normalization on the product of the aggregation degree of the fluorescent magnetic powder and the target difference as the similarity between the target area and the background area.
Further, the method for determining a defect area in the gray scale image according to the similarity comprises the following steps:
setting a similarity threshold, and merging the suspected defect area with the background area when the similarity is larger than the similarity threshold and the suspected defect area corresponding to the target area is not defective; the non-merged region is the defective region, and the defective region is verified.
Further, the method for clustering the pixel points in each suspected defect area includes:
and clustering the pixel points in the suspected defect area according to the positions and gray values of the pixel points by using a DBSCAN density clustering algorithm.
The invention has the following beneficial effects:
the suspected defect area in the gray level image of the bearing surface is obtained through a watershed algorithm, and then only the suspected defect area is analyzed, so that the defect area detection efficiency is improved; the pixel point corresponding to the maximum gray value in each suspected defect area is obtained as a target pixel point, so that the subsequent analysis of the specific characteristics of the suspected defect areas is facilitated; therefore, the pixel points in each suspected defect area are clustered, the area of the cluster where the target pixel points are located is taken as a target area, the pixel points in the target area are the pixel points with larger gray value peaks in the suspected defect areas, and the attachment condition of fluorescent magnetic powder in the corresponding suspected defect areas can be reflected, so that the fluorescent magnetic powder aggregation degree of each target area is obtained according to the area of the suspected defect areas corresponding to the target areas and the position distribution and gradient distribution of the pixel points in the target areas, and the defect areas are primarily judged; in order to further determine the defect area, acquiring a gray level difference between the target area and the gray level image as a target difference, and indirectly determining the difference degree between the suspected defect area where the target area is located and the background area; and then, the similarity between the target area and the background area is obtained according to the aggregation degree of the fluorescent magnetic powder and the target difference, and the suspected defect area corresponding to the target area is combined with the background area according to the similarity, so that the interference on defect area detection caused by uneven bearing surface is removed, the over-segmentation area caused by a watershed algorithm is eliminated, the defect area in the gray level image is determined, and the accuracy of defect area detection is 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 schematic flow chart of an intelligent detection method for processing information of automobile parts according to an embodiment of the invention;
fig. 2 is a distribution diagram of gray value sizes in a suspected defect area according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of specific implementation, structure, characteristics and effects of an intelligent detection method for processing information of automobile parts according to the invention with reference to the accompanying drawings and 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 following specifically describes a specific scheme of the intelligent detection method for the processing information of the automobile parts, which is provided by the invention, with reference to the accompanying drawings.
Referring to fig. 1, a flow chart of an intelligent detection method for processing information of an automobile part according to an embodiment of the invention is shown, and the method comprises the following steps:
step S1: and obtaining a gray image of the bearing surface after fluorescent magnetic powder injection.
Specifically, the purpose of the embodiment of the invention is to detect the defects on the surface of a bearing by a fluorescent magnetic powder method, firstly, a certain amount of fluorescent magnetic powder is sprayed on the surface of the bearing by using a Saifu flaw detector, then, an image of the surface of the bearing after the fluorescent magnetic powder is sprayed is collected by a high-definition camera, the obtained image is subjected to denoising treatment, and the obtained image is subjected to denoising by using a mean filtering algorithm, and then, the denoised image is subjected to graying treatment, so that a gray image is obtained.
The mean filtering algorithm and the graying process are all in the prior art, and are not described herein.
Step S2: obtaining a suspected defect area in the gray level image through a watershed algorithm; and acquiring a pixel point corresponding to the maximum gray value in each suspected defect area as a target pixel point, clustering the pixel points in each suspected defect area, and taking the area of the cluster where the target pixel point is located as a target area.
Specifically, the bearing surface is marked by a fluorescent magnetic powder method, different flatness of the bearing surface can form different bright and dark areas, for example, when the degree of concavity and convexity of the bearing surface is large, fine fluorescent magnetic powder can adhere to the uneven part of the bearing surface to form an integral fluorescent area; the defects on the surface of the bearing are oxidized and cracked to form sharp corners on the surface layer metal grain boundaries under microscopic observation, if the segregation of the components in the metal is serious, the grain boundaries also start to melt, sharp corner-shaped cavities can be formed when the segregation is serious, and fluorescent magnetic powder can be adsorbed in the defect areas to form an integral fluorescent area. Analysis of the fluorescent area in the bearing surface indicates that the larger the fluorescent area, the more likely the fluorescent area is to be a larger uneven or defective area in the bearing surface. And (3) in the gray image, the gray value corresponding to the fluorescent region is larger, the gray value corresponding to the background region which is the other region except the fluorescent region is smaller, and the initial segmentation region in the gray image, namely the suspected defect region, is obtained through a watershed algorithm. The watershed algorithm finds minima in the gray image and then starts to expand, if there are too many tiny areas in the gray image to create many small catchment basins, this can lead to the gray image being over-segmented. Therefore, when the water diversion ridge algorithm is used for segmentation, the fluorescence area is larger, so that the area of the suspected defect area reserved after flooding is larger, the suspected defect area comprises not only an uneven area with partial defects but also a defect area, and the segmented defect area is inaccurate. Therefore, the obtained suspected defect area needs to be processed, the suspected defect area without defects is removed, and then the rest suspected defect areas are identified, so that the defect area is determined. According to the embodiment of the invention, the flooding position of the watershed algorithm is optimized according to the characteristics of the defect area, the optimal flooding position is obtained, and the interference of the uneven area without defects caused by marking the fluorescent magnetic powder is removed, so that the reserved suspected defect area is the defect area.
The watershed algorithm is a well-known technique, and will not be described in detail here.
The suspected defect areas obtained through the watershed algorithm are judged directly according to a threshold value, and the suspected defect areas are uneven areas and defect areas, so that the specific characteristics of each suspected defect area are required to be analyzed to merge the suspected defect areas, and the suspected defect areas belonging to the normal uneven areas are merged with the background areas, namely, when the suspected defect areas are segmented through the watershed algorithm, the process of flooding is not only needed, but also the suspected defect areas are required to be judged, and the suspected defect areas belonging to the normal uneven areas are also made to be flooded objects.
Analyzing specific characteristics of the suspected defect areas, and acquiring target areas in each suspected defect area, wherein the specific operation of acquiring the target areas is as follows:
and obtaining a pixel point corresponding to the maximum gray value in each suspected defect area as a target pixel point, and if at least two pixels corresponding to the maximum gray value in one suspected defect area exist, arbitrarily selecting the pixel point corresponding to the maximum gray value as the target pixel point. And clustering the pixel points in the suspected defect area according to the positions and gray values of the pixel points by using a DBSCAN density clustering algorithm, wherein the DBSCAN density clustering algorithm clusters according to the positions of the pixel points, and selects which pixel points are clustering objects through the gray values. And acquiring the area of the cluster where the target pixel point is located as a target area. The embodiment of the invention sets the neighborhood radius in the DBSCAN density clustering algorithm to be 3, the minimum neighborhood pixel point number to be 3, the threshold value is an empirical threshold value, and an implementer can set the neighborhood radius according to different implementation environments. The DBSCAN density clustering algorithm is a well-known technique, and will not be described in detail here.
The reason why the region of the cluster where the target pixel point is located is taken as the target region is as follows:
the fluorescent magnetic powder attached to the defect area is more and uniform, so that the whole gray value of the defect area is larger and the gray fluctuation is larger; the gray value distribution of the pixel points in the defect area is shown as a B area in fig. 2, and the height of each peak in the B area is the gray value of the corresponding pixel point in the defect area; the uneven area of the bearing surface is provided with background areas between suspected defect areas formed by attachment of fluorescent magnetic powder, so that the distance between peaks in the corresponding gray distribution is larger, gray value peaks in the suspected defect areas corresponding to the uneven area are uneven, gray fluctuation of pixel points is large or small, as in an area A in fig. 2, each peak in the area A is a gray value of a corresponding pixel point in the corresponding suspected defect area. According to fig. 2, it can be found that the distribution characteristics of peaks between the suspected defect area and the defect area corresponding to the uneven area are greatly different, the peak interval of the area a is larger, and the peak is high or low; the peaks of the region B are denser and larger, after the watershed algorithm is passed, partial peaks of the region A are smaller than a threshold value, so that the region A cannot be segmented, the peaks of the region B are larger than a segmentation threshold value, the peaks of the region B are segmented, and the aggregation degree of fluorescent magnetic powder of the peaks of the region B is larger. Therefore, the more the fluorescent magnetic powder is gathered, the more likely the suspected defective region is a defective region. The cluster where the target pixel points are located can fully reflect the distribution of the high gray value peaks in the corresponding suspected defect areas, so that the area of the cluster where the target pixel points are located is selected as the target area, and the target area is analyzed. Because the distribution of gray value peaks in the region A is different from that in the region B, the number of pixels in the target region corresponding to each suspected defect region after clustering is different, and the distribution of the pixels is also different, and the aggregation degree of fluorescent magnetic powder in each target region is obtained according to the number and the distribution of the pixels in the target region.
Step S3: and acquiring the aggregation degree of the fluorescent magnetic powder of each target area according to the area of the suspected defect area corresponding to the target area and the position distribution and gradient distribution of the pixel points in the target area.
Specifically, the identification of the defect area by the fluorescent area generated by the uneven area of the bearing surface can be interfered, and the fluorescent area without defects and the defect area cannot be accurately distinguished only by the gray distribution. However, since there is a possibility that a defect exists in the fluorescent-labeled region, a suspected defective region is obtained by dividing the fluorescent region in the gray-scale image, and then the suspected defective region is analyzed to obtain an accurate defective region.
And analyzing the suspected defect area, combining the suspected defect area without defects with the background area, and taking the finally divided area as the defect area. As can be seen from step S2, the suspected defect area can be primarily analyzed according to the aggregation degree of the fluorescent magnetic powder in the target area.
Preferably, the method for obtaining the aggregation degree of the fluorescent magnetic powder comprises the following steps: acquiring coordinates of pixel points corresponding to the maximum gray value in the target area as first coordinates, calculating the distance between the coordinates corresponding to each pixel point in the target area and the first coordinates, and acquiring the average distance between the pixel points in the target area; carrying out negative correlation mapping and normalization on the average distance, and taking the obtained result as a first result; acquiring a gradient mean value of edge pixel points in a target area as an average gradient; acquiring the number of pixel points in a suspected defect area corresponding to a target area, namely the area of the suspected defect area, as a first area; the product of the first result, the average gradient and the first area is calculated as the degree of aggregation of the fluorescent magnetic powder in the target region.
It should be noted that, the fluorescence area of the bearing surface caused by the unevenness only remains the area with more obvious fluorescence after being divided by the watershed algorithm, and the area with less obvious fluorescence is divided into the background area; the defect area is larger in area which is reserved after the defect area is segmented by a watershed algorithm, so that the number of pixels with larger gray values in the defect area is larger, the pixels with larger gray values are peak values of each peak in the B area in fig. 2, the pixels in the corresponding target area are pixels which are similar to the peak value of the maximum peak value in the corresponding suspected defect area, and the pixels in the corresponding target area are a certain larger defect area on the bearing surface. The average distance of the pixel points in the target area is calculated to represent the aggregation degree of the fluorescent magnetic powder in the target area, and the greater the aggregation degree of the fluorescent magnetic powder is, the greater the possible degree that the suspected defect area corresponding to the target area is the defect area is. The gradient change of the edge pixel points is used for correcting the clustering result, the density clustering result is that the pixel points close to the peak value are clustered into one type, the areas corresponding to the clustered pixel points are clustered into one type, the peaks of the defect areas are clustered relatively, the gray value is larger, the larger the gradient of the edge pixel points indicates that the more obvious the edge characteristics generated by the aggregation of fluorescent magnetic powder in the areas are, and the greater the aggregation degree of the fluorescent magnetic powder is. Different peaks are clustered into one region, namely a target region, so that the larger the area of the target region, namely the larger the number of pixel points in the target region, the more the fluorescent magnetic powder at the corresponding position of the target region is clustered, the larger the fluorescent magnetic powder aggregation degree is, and the greater the possible degree that the corresponding suspected defect region is a defect region is.
After the watershed algorithm is used for dividing, each suspected defect area is equivalent to an isolated area, the number of pixel points in the suspected defect area corresponding to the defect area is definitely large, so that the probability that the suspected defect area with the larger number of pixel points is the defect area is larger, the suspected defect area without defects which is divided by unevenness is equivalent to scattered points, the number of pixel points in the corresponding suspected defect area is small, and the probability that the suspected defect area is the defect area is smaller. The number of pixels in the suspected defective region, i.e., the area of the suspected defective region, is thus taken as one component of the degree of aggregation of the fluorescent magnetic powder.
Taking the target region j as an example, the coordinates of the pixel point corresponding to the maximum gray value in the target region j are obtained as the first coordinates, if the maximum gray value in the target region j corresponds toAt least two corresponding pixel points are provided, and the coordinate of one pixel point is arbitrarily selected as a first coordinate. And acquiring the distance between each pixel point in the target area j and the pixel point corresponding to the first coordinate according to the coordinate of each pixel point in the target area j. By passing through
The operator obtains the gradient of each pixel point in the target region j, and obtains the gradient of the edge pixel point at the boundary of the region, wherein,
operators are known techniques, and will not be described in detail here. Obtaining the aggregation degree of the fluorescent magnetic powder of the target region j according to the distance between each pixel point in the target region j and the pixel point corresponding to the first coordinate and the gradient of each pixel point, and obtaining the aggregation degree of the fluorescent magnetic powder of the target region j
The formula of (2) is:
wherein,,
the aggregation degree of the fluorescent magnetic powder in the target area j;
the number of pixel points in the target region j;
the coordinate of the pixel point corresponding to the maximum gray value in the target region j is the first coordinate;
is the first in the target region j
Coordinates of the individual pixel points;
is the first in the target region j
Gradients of the edge pixels;
the area of the suspected defect area where the target area j is located is the first area;
is a natural constant.
Average distance of the two
The smaller, the more concentrated the position distribution of the pixel points in the target region j is,
the larger; average gradient
The larger the gray scale change of the edge pixel point in the target region j is, the more obvious the edge characteristic generated by the aggregation of fluorescent magnetic powder is,
the larger;
the larger the area of the suspected defect area where the target area j is located, that is, the larger the number of pixels, the more pixels satisfying the clustering condition, the better the clustering effect,
the larger; thus, the first and second substrates are bonded together,
the larger the target region j, the higher the aggregation degree of the fluorescent magnetic powder is, and the more probable the suspected defect region where the target region j is located is a defect region.
According to the method for acquiring the aggregation degree of the fluorescent magnetic powder of the target area j, the aggregation degree of the fluorescent magnetic powder of each target area is acquired. The larger the aggregation degree of the fluorescent magnetic powder in the target area is, the more uneven the target area is, and the more obvious the marking effect is after the uneven area is marked by the fluorescent magnetic powder method; the smaller the degree of aggregation of the fluorescent magnetic powder in the target region, the more the target region is indicated to be a normal region remained due to the uneven bearing surface. The correlation of the target areas is inaccurate through the positions of the pixel points, the clustering result is corrected according to the gradient change among the peaks, and whether the corresponding suspected defect areas can be combined with the background areas is judged according to the aggregation degree of fluorescent magnetic powder of each target area.
Step S4: acquiring a gray scale difference between the target area and the gray scale image as a target difference; and obtaining the similarity between the target area and the background area according to the aggregation degree of the fluorescent magnetic powder and the target difference.
Specifically, in order to determine the optimal flooding position of the watershed algorithm, the suspected defect area without defects is combined with the background area, so that an accurate defect area is obtained, and the similarity between the target area and the background area is obtained according to the aggregation degree of fluorescent magnetic powder of the target area.
Preferably, the method for obtaining the similarity between the target area and the background area is as follows: calculating an average gray value in the target area as a first value; calculating an average gray value in the gray image as a second value; the difference between the first value and the second value is taken as a target difference between the target area and the gray scale image. And taking the result of carrying out negative correlation mapping and normalization on the product of the aggregation degree of the fluorescent magnetic powder and the target difference as the similarity between the target area and the background area.
The suspected defect area with high similarity is combined with the background area by calculating the similarity between the suspected defect area and the background area, so that the suspected defect area with partial non-defects and the judgment of the defect area is eliminated. When the segmentation is performed by the watershed algorithm, excessive interference of suspected defect areas without defects is avoided, so that the defect areas in the gray level image are segmented more accurately.
Taking the target region j in the step S3 as an example, according to the gray value of each pixel point in the target region j, the average gray value of the target region j, that is, the first value is obtained; according to the gray value of each pixel point in the gray image, the average gray value of the gray image, namely the second value, is obtained, because the pixel points in the target area j are all the pixel points marked by fluorescent magnetic powder, the gray value is larger, namely the first value is larger, and the pixel points in the gray image also comprise the pixel points in the background area, therefore the second value is smaller than the first value, and the difference value between the first value and the second value, namely the target difference is a positive number. Acquiring target difference of target region j
The formula of (2) is:
wherein,,
the target difference between the target region j and the gray level image;
the average gray value of the target region j is the first value;
the number of pixels in the gray level image;
is the first in the gray level image
Gray values of individual pixels.
It should be noted that the number of the substrates,
the larger the difference between the average gray value of the target region j and the average gray value of the gray image is, the larger the difference is; by fluorescent magnetic powder methodMarking, wherein the gray value of the defect area is larger than that of the normal area, and the larger the gray value is, the more obvious the defect is; thus, the first and second substrates are bonded together,
the larger the difference between the target region j and the background region is explained to be larger.
According to the aggregation degree of fluorescent magnetic powder
Difference from the target
Obtaining the similarity between the target region j and the background region
The formula of (2) is:
wherein,,
similarity between the target region j and the background region;
the aggregation degree of the fluorescent magnetic powder in the target area j;
target variance for target region j;
is a natural constant.
It should be noted that the number of the substrates,
the larger the suspected defective area where the target area j is located, the more likely it is a defective area,
the smaller;
the larger the difference between the gray value in the target region j and the gray value in the background region, the less similar the target region j and the background region, the more likely the target region j is a defective region,
the smaller; thus, the first and second substrates are bonded together,
the smaller the target region j, the less similar the background region, the more likely the suspected defect region in which the target region j is located is a defect region.
And obtaining the similarity of each target area and the background area according to the method for obtaining the similarity of the target area j and the background area. And setting a proper segmentation threshold value in a watershed algorithm according to the similarity, so that the final segmentation area is a defect area.
Step S5: and merging the suspected defect area corresponding to the target area with the background area according to the similarity, and determining the defect area in the gray level image.
Setting a similarity threshold, and merging the suspected defect area with the background area when the similarity is larger than the similarity threshold and the suspected defect area corresponding to the target area is not defective; the non-merged region is the defective region, and the defective region is verified.
The embodiment of the invention sets the similarity threshold to 0.32, and an implementer can adjust the similarity threshold according to different actual conditions. When the similarity is larger than the similarity threshold, the similarity between the target area and the background area is large, and the suspected defect area where the target area is located is not defective, the suspected defect area is combined with the background area, and the suspected defect area is not segmented when the gray image is segmented through a watershed algorithm.
After the suspected defect areas are combined, the suspected defect areas without defects are eliminated, the rest of the non-combined suspected defect areas are the defect areas, and the defect areas are marked, so that the defect areas can be verified manually, and the defect detection efficiency is improved.
The embodiment of the invention ends up.
In summary, the embodiment of the invention obtains the gray image of the bearing surface after fluorescent magnetic powder injection; obtaining a suspected defect area in the gray level image through a watershed algorithm; obtaining a pixel point corresponding to the maximum gray value in each suspected defect area as a target pixel point, clustering the pixel points in each suspected defect area, and taking the area of a cluster where the target pixel point is located as a target area; acquiring the aggregation degree of fluorescent magnetic powder of each target area according to the area of the suspected defect area corresponding to the target area and the position distribution and gradient distribution of the pixel points in the target area; acquiring a gray level difference between a target area and a gray level image as a target difference; obtaining the similarity between the target area and the background area according to the aggregation degree of the fluorescent magnetic powder and the target difference; and determining a defect area in the gray image according to the similarity. The efficiency of detecting defective areas is improved.
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 (4)
1. The intelligent detection method for the automobile part machining information is characterized by comprising the following steps of:
obtaining a gray image of the bearing surface after fluorescent magnetic powder injection;
obtaining a suspected defect area in the gray level image through a watershed algorithm; obtaining pixel points corresponding to the maximum gray value in each suspected defect area as target pixel points, clustering the pixel points in each suspected defect area, and taking the area of the cluster where the target pixel points are located as a target area;
acquiring the aggregation degree of fluorescent magnetic powder of each target area according to the area of the suspected defect area corresponding to the target area and the position distribution and gradient distribution of pixel points in the target area;
acquiring a gray scale difference between the target area and the gray scale image as a target difference; obtaining the similarity between the target area and the background area according to the aggregation degree of the fluorescent magnetic powder and the target difference;
combining the suspected defect area corresponding to the target area with the background area according to the similarity, and determining a defect area in the gray level image;
the method for acquiring the aggregation degree of the fluorescent magnetic powder comprises the following steps:
acquiring coordinates of pixel points corresponding to the maximum gray value in the target area as first coordinates, calculating the distance between the coordinates corresponding to each pixel point in the target area and the first coordinates, and acquiring the average distance between the pixel points in the target area; carrying out negative correlation mapping and normalization on the average distance, and taking the obtained result as a first result;
acquiring a gradient mean value of edge pixel points in the target area as an average gradient;
acquiring the number of pixel points in the suspected defect area corresponding to the target area, namely the area of the suspected defect area, as a first area;
calculating the product of the first result, the average gradient and the first area as the aggregation degree of the fluorescent magnetic powder of the target area;
the target difference acquisition method comprises the following steps:
calculating an average gray value in the target area as a first value;
calculating an average gray value in the gray image as a second value;
and taking the difference value between the first value and the second value as a target difference between the target area and the gray scale image.
2. The intelligent detection method for processing information of automobile parts according to claim 1, wherein the method for obtaining the similarity comprises the following steps:
and taking the result of carrying out negative correlation mapping and normalization on the product of the aggregation degree of the fluorescent magnetic powder and the target difference as the similarity between the target area and the background area.
3. The intelligent detection method for automotive part processing information according to claim 1, wherein the method for determining the defective area in the grayscale image according to the similarity comprises:
setting a similarity threshold, and merging the suspected defect area with the background area when the similarity is larger than the similarity threshold and the suspected defect area corresponding to the target area is not defective; the non-merged region is the defective region, and the defective region is verified.
4. The intelligent detection method for automotive part processing information according to claim 1, wherein the method for clustering the pixels in each suspected defective area comprises:
and clustering the pixel points in the suspected defect area according to the positions and gray values of the pixel points by using a DBSCAN density clustering algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310171997.6A CN115861320B (en) | 2023-02-28 | 2023-02-28 | Intelligent detection method for automobile part machining information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310171997.6A CN115861320B (en) | 2023-02-28 | 2023-02-28 | Intelligent detection method for automobile part machining information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115861320A CN115861320A (en) | 2023-03-28 |
| CN115861320B true CN115861320B (en) | 2023-05-12 |
Family
ID=85659149
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310171997.6A Active CN115861320B (en) | 2023-02-28 | 2023-02-28 | Intelligent detection method for automobile part machining information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115861320B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116091499B (en) * | 2023-04-07 | 2023-06-20 | 山东中胜涂料有限公司 | Abnormal paint production identification system |
| CN116630815B (en) * | 2023-07-25 | 2023-09-22 | 济南玖通志恒信息技术有限公司 | Intelligent agricultural pest detection method |
| CN116990993B (en) * | 2023-09-26 | 2023-12-22 | 深圳市柯达科电子科技有限公司 | LCD display panel quality detection method |
| CN117291937B (en) * | 2023-11-27 | 2024-03-05 | 山东嘉达装配式建筑科技有限责任公司 | Automatic plastering effect visual detection system based on image feature analysis |
| CN118115499B (en) * | 2024-04-29 | 2024-07-12 | 深圳市金利源绝缘材料有限公司 | Visual detection method for production quality of electric power copper bar |
| CN118505686B (en) * | 2024-07-17 | 2024-09-10 | 陕西海鹰汽车部件有限公司 | Punching defect detection method for automobile punching pipe |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101168474A (en) * | 2006-10-27 | 2008-04-30 | 群康科技(深圳)有限公司 | Method for manufacturing polycrystalline silicon thin film at low temperature |
| CN103955913A (en) * | 2014-02-18 | 2014-07-30 | 西安电子科技大学 | SAR image segmentation method based on line segment co-occurrence matrix characteristics and regional maps |
| CN107145896A (en) * | 2017-03-14 | 2017-09-08 | 西南科技大学 | Dysnusia identifying system based on fluorescentmagnetic particle(powder) |
| CN107292882A (en) * | 2017-08-02 | 2017-10-24 | 国网电力科学研究院武汉南瑞有限责任公司 | One kind is based on the adaptive electrical equipment malfunction detection methods of Meanshift |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018133074A1 (en) * | 2017-01-22 | 2018-07-26 | 四川金瑞麒智能科学技术有限公司 | Intelligent wheelchair system based on big data and artificial intelligence |
| CN108595596A (en) * | 2018-04-19 | 2018-09-28 | 厦门启尚科技有限公司 | A kind of image similarity search method |
| CN108629783B (en) * | 2018-05-02 | 2021-05-04 | 山东师范大学 | Image segmentation method, system and medium based on image feature density peak search |
| CN109215020B (en) * | 2018-08-30 | 2022-06-14 | 国网黑龙江省电力有限公司佳木斯供电公司 | High-voltage transmission line fault identification method based on computer vision |
| CN110706236B (en) * | 2019-09-03 | 2023-01-13 | 西人马大周(深圳)医疗科技有限公司 | Three-dimensional reconstruction method and device of blood vessel image |
| CN112131924A (en) * | 2020-07-10 | 2020-12-25 | 国网河北省电力有限公司雄安新区供电公司 | Transformer substation equipment image identification method based on density cluster analysis |
| CN112906550B (en) * | 2021-02-09 | 2022-07-19 | 哈尔滨理工大学 | Static gesture recognition method based on watershed transformation |
| CN113642631B (en) * | 2021-08-10 | 2022-07-12 | 沭阳协润电子有限公司 | Dangerous area electronic fence generation method and system based on artificial intelligence |
| CN114581692B (en) * | 2022-03-06 | 2022-12-02 | 扬州孚泰电气有限公司 | Vibration damper fault detection method and system based on intelligent pattern recognition |
| CN115100212B (en) * | 2022-08-29 | 2022-11-18 | 卡松科技股份有限公司 | Method for detecting pollution degree of lubricating oil |
| 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
- 2023-02-28 CN CN202310171997.6A patent/CN115861320B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101168474A (en) * | 2006-10-27 | 2008-04-30 | 群康科技(深圳)有限公司 | Method for manufacturing polycrystalline silicon thin film at low temperature |
| CN103955913A (en) * | 2014-02-18 | 2014-07-30 | 西安电子科技大学 | SAR image segmentation method based on line segment co-occurrence matrix characteristics and regional maps |
| CN107145896A (en) * | 2017-03-14 | 2017-09-08 | 西南科技大学 | Dysnusia identifying system based on fluorescentmagnetic particle(powder) |
| CN107292882A (en) * | 2017-08-02 | 2017-10-24 | 国网电力科学研究院武汉南瑞有限责任公司 | One kind is based on the adaptive electrical equipment malfunction detection methods of Meanshift |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115861320A (en) | 2023-03-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115861320B (en) | Intelligent detection method for automobile part machining information | |
| CN115311292B (en) | Strip steel surface defect detection method and system based on image processing | |
| CN111179225B (en) | Test paper surface texture defect detection method based on gray gradient clustering | |
| CN109682839B (en) | Online detection method for surface defects of metal arc-shaped workpiece | |
| Aminzadeh et al. | Automatic thresholding for defect detection by background histogram mode extents | |
| CN106780486B (en) | Steel plate surface defect image extraction method | |
| CN115601367A (en) | LED lamp wick defect detection method | |
| CN109685760B (en) | MATLAB-based SLM powder bed powder laying image convex hull depression defect detection method | |
| CN107437243B (en) | Tire impurity detection method and device based on X-ray image | |
| CN115375676A (en) | Stainless steel product quality detection method based on image recognition | |
| CN115861291B (en) | Chip circuit board production defect detection method based on machine vision | |
| CN104700085B (en) | A kind of chip positioning method based on template matches | |
| CN115272321B (en) | Textile defect detection method based on machine vision | |
| CN109540925B (en) | Complex ceramic tile surface defect detection method based on difference method and local variance measurement operator | |
| CN116152242B (en) | Visual detection system of natural leather defect for basketball | |
| CN110648330B (en) | Defect detection method for camera glass | |
| CN114820625B (en) | Automobile top block defect detection method | |
| CN115294159B (en) | Method for dividing corroded area of metal fastener | |
| CN116091499A (en) | Abnormal paint production identification system | |
| CN110060239B (en) | Defect detection method for bottle opening of bottle | |
| CN115239718B (en) | Plastic product defect detection method and system based on image processing | |
| CN115100191A (en) | Metal casting defect identification method based on industrial detection | |
| CN115274486B (en) | Semiconductor surface defect identification method | |
| CN107516315B (en) | Tunneling machine slag tapping monitoring method based on machine vision | |
| CN112927182A (en) | Digital PCR microarray image analysis method |
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 |