CN115511884A - Punching compound die surface quality detection method based on computer vision - Google Patents
Punching compound die surface quality detection method based on computer vision Download PDFInfo
- Publication number
- CN115511884A CN115511884A CN202211420216.4A CN202211420216A CN115511884A CN 115511884 A CN115511884 A CN 115511884A CN 202211420216 A CN202211420216 A CN 202211420216A CN 115511884 A CN115511884 A CN 115511884A
- Authority
- CN
- China
- Prior art keywords
- matching
- closed
- die
- hole site
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20076—Probabilistic image processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
-
- 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
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Geometry (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of data processing, in particular to a punching compound die surface quality detection method based on computer vision, which comprises the following steps: acquiring a gray image of the surface of the punching compound die and carrying out edge detection to obtain all closed edges; acquiring the central point of each closed edge and the central point of each hole site in the standard image of the surface of the punching compound die to obtain hole site offset; respectively obtaining closed boundary images of an upper die and a lower die of the punching compound die, carrying out characteristic point analysis on each closed edge to obtain a corresponding characteristic point sequence, obtaining a matching matrix according to the characteristic point sequence of the corresponding closed edge, and further obtaining the matching degree between the corresponding closed edges; obtaining all hole site matching pairs in the upper die and the lower die according to the matching degree; and obtaining the abnormal probability according to the number of all hole site matching pairs in the upper die and the lower die, and performing quality judgment based on the abnormal probability, thereby effectively improving the accuracy of surface quality detection of the counter-punching composite die.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a punching compound die surface quality detection method based on computer vision.
Background
The common production processing machinery device of life such as car, boats and ships, lathe needs multiple mechanical parts in production assembling process, these spare parts have different requirements to material and structure according to in the scene of difference and device, in order to satisfy the spare part requirement of different sizes, traditional spare part processing production process need treat the coarse fodder of processing and carry out multiple processes such as blanking, turn-ups, punch a hole, every process all needs specific mold processing, but link up production between the multiple process and probably lead to the material position transform, cause the processing skew and then influence the spare part quality.
In the modern processing process, the machining of mechanical parts is usually realized by using a composite die, and the production and manufacturing efficiency is higher; the punching composite die comprises an upper die and a lower die, wherein groove hole sites between the dies are in one-to-one correspondence to enable processed parts to meet production quality requirements, but due to the complexity of the production process and errors of production mechanical devices, material position deviation and even damage to the hole sites of the punching composite die or rigid deformation of the hole sites can be caused, so that the punching composite die needs to be periodically detected, but the quality is often judged based on the comparison of the current dies and standards, the influence of deformation in different degrees is ignored, and the accuracy of a judgment result is low.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a punching compound die surface quality detection method based on computer vision, and the adopted technical scheme is as follows:
one embodiment of the invention provides a punching compound die surface quality detection method based on computer vision, which comprises the following steps: acquiring a gray image of the surface of the punching compound die;
performing edge detection on the gray level image to obtain all closed edges in the gray level image; acquiring a central point of each closed edge and a central point of each hole site in a standard image of the surface of the punching compound die, and acquiring hole site offset according to the Euclidean distance between the central point of each closed edge and the central point of the corresponding hole site in the standard image;
respectively acquiring closed boundary images of an upper die and a lower die of the punching composite die, and respectively performing characteristic point analysis on corresponding closed edges in the closed boundary images of the upper die and the lower die to obtain a first characteristic point sequence and a second characteristic point sequence; acquiring a matching matrix according to the first characteristic point sequence and the second characteristic point sequence, acquiring a matching value between the first characteristic point and the second characteristic point based on the matching matrix, and acquiring a matching degree between corresponding closed edges according to the matching value;
obtaining all hole site matching pairs in the upper die and the lower die according to the matching degree; and counting the number of all hole site matching pairs in the upper die and the lower die to obtain the abnormal probability of the punching composite die, and judging the quality based on the abnormal probability.
Preferably, the step of obtaining a matching value between the first feature point and the second feature point based on the matching matrix includes:
acquiring an HSV image corresponding to the gray image, and acquiring the brightness value of each pixel point based on the HSV image;
and constructing a mapping coefficient based on the matching matrix and the coordinate of the first characteristic point, acquiring a first ratio of the abscissa and the ordinate of the second characteristic point and a second ratio of the brightness value of the first characteristic point and the brightness value of the second characteristic point, and obtaining a matching value between the first characteristic point and the second characteristic point by the product of the first ratio, the second ratio and the mapping coefficient.
Preferably, the step of obtaining the matching degree between the corresponding closed edges according to the matching value includes:
judging whether the matching value of the first feature point and the corresponding second feature point is within a preset matching range, if so, successfully matching the first feature point with the second feature point;
and calculating the cumulative sum of the matching values between the first characteristic points and the second characteristic points which are successfully matched on the corresponding closed edges, and normalizing the cumulative sum to obtain the matching degree between the corresponding closed edges.
Preferably, the step of obtaining a matching matrix according to the first feature point sequence and the second feature point sequence includes:
correspondingly selecting F first characteristic points and second characteristic points from the first characteristic point sequence and the second characteristic point sequence, wherein F is a positive integer;
and respectively constructing a first matrix and a second matrix by using the F first characteristic points and the F second characteristic points, taking a multiplication result of the matching matrix and the first matrix as the second matrix, and obtaining each element in the matching matrix based on matrix operation.
Preferably, the step of constructing a mapping coefficient based on the matching matrix and the coordinates of the first feature point includes:
the expression of the mapping coefficient is:
wherein, the first and the second end of the pipe are connected with each other,coordinates representing the first feature point;respectively representing elements in the matching matrix.
Preferably, the step of obtaining the hole site offset according to the euclidean distance between the center point of each closed edge and the center point of the corresponding hole site in the standard image includes:
constructing straight lines in different directions through the center point of the closed edge, wherein each straight line and the closed edge are intersected at two points to form a corresponding line segment, and acquiring the length of the line segment formed by the straight lines in different directions;
selecting the maximum value and the minimum value of the lengths of all line segments, and obtaining an offset factor according to the difference value of the maximum value and the minimum value, wherein the offset factor and the difference value are in positive correlation;
and taking the product of the Euclidean distance between the central point of the closed edge and the central point of the corresponding hole site in the standard image and the offset factor as the hole site offset of the closed edge.
Preferably, the step of counting the number of all the hole site matching pairs in the upper die and the lower die to obtain the abnormal probability of the punching composite die includes:
acquiring the number of all unmatched feature points on the two closed edges in each hole position matching pair and the sum of the number of all feature points on the two closed edges in the hole position matching pair, and taking the ratio result of the number of all unmatched feature points on the two closed edges and the sum of the number of all feature points on the two closed edges as an unmatched ratio;
calculating the product result between the unmatched ratio and the reciprocal of the matching degree of the hole site matching pairs, wherein the sum of the product results corresponding to all the hole site matching pairs is an abnormal factor;
and acquiring the sum of hole site offsets of all closed edges in all the hole site matching pairs, calculating the product of the sum of the hole site offsets and the abnormal factor, and performing normalization processing to obtain the abnormal probability.
Preferably, the step of determining the quality based on the abnormal probability includes:
and when the abnormal probability is greater than a preset abnormal threshold value, the surface quality of the punching compound die is poor.
Preferably, the step of acquiring all closed edges in the grayscale image includes:
traversing search is carried out on each edge pixel point in the gray level image, and the traversing search direction is the eight neighborhood direction of the edge pixel point;
when the edge pixel point of the nth traversal search and the edge pixel point of the second traversal search are the same pixel point, and the edge pixel point of the (n-1) th traversal search and the edge pixel point of the first traversal search are also the same pixel point, the traversal search edge is a closed edge; wherein n is a positive integer.
Preferably, the closed boundary image is an image including all closed edges.
The invention has the following beneficial effects: in the embodiment of the invention, through analyzing the gray level image on the surface of the punching compound die, all closed edges in the gray level image are obtained firstly so as to screen out the interference of other false edges and reduce the calculated amount; calculating the Euclidean distance between the position of each central point on the closed edge and the position of the corresponding hole site in the standard image to obtain the hole site offset, and quantizing the offset degree through actual data to make analysis more intuitive; and then the corresponding closed edges in the upper die and the lower die of the punching composite die are analyzed, a corresponding characteristic point sequence is constructed through the characteristic points on each closed edge, a matching matrix is further obtained for auxiliary analysis, so that a matching value between every two corresponding characteristic points is obtained, and then the matching degree of the closed edges is obtained.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flowchart of a method for detecting the surface quality of a punching compound die based on computer vision according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a closed edge and a non-closed edge according to an embodiment of the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the surface quality inspection method of a punching compound die based on computer vision according to the present invention with reference to the accompanying drawings and preferred embodiments shows the following detailed descriptions. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
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 method and the device are suitable for detecting the surface quality of the punching compound die, and the obtained abnormal probability is more accurate through the hole site offset of each closed edge in the corresponding gray level image of the punching compound die and the matching degree combination analysis between the corresponding hole sites of the upper die and the lower die of the punching compound die, and the surface quality based on more accurate abnormal probability analysis is more accurate.
The following describes a specific scheme of the punching compound die surface quality detection method based on computer vision in detail with reference to the accompanying drawings.
Referring to fig. 1, a flow chart of a method for detecting surface quality of a punching compound die based on computer vision according to an embodiment of the present invention is shown, the method including the following steps:
and S100, acquiring a gray image of the surface of the punching compound die.
When the punching compound die works, the quality of each hole site is the basis for ensuring normal use and production, and the quality of the punching compound die is poor due to the fact that the hole sites on the punching compound die are not matched or deformed due to the problems of mechanical devices or the errors of operators, and a large quality problem exists for produced parts, so that the surface quality of the punching compound die needs to be detected.
Considering that the CCD camera has the advantages of high sensitivity, clear imaging and no smear, the CCD camera is adopted to collect images of the punching compound die in the embodiment of the invention to obtain RGB images corresponding to the punching compound die, and the punching compound die is divided into an upper die and a lower die, so the RGB images of the punching compound die collected in the embodiment of the invention are the RGB images of the upper die and the RGB images of the lower die; in order to facilitate subsequent analysis of the image, the RGB image is processed by using a maximum method to obtain a corresponding gray-scale image, and means for obtaining the gray-scale image by using the maximum method is known in the art and is not described in detail.
In order to improve the accuracy of subsequent analysis based on the gray level image and reduce the influence of noise errors in the image, the method of gaussian filtering is used as a common means for removing noise points in the gray level image, and is not described in detail, and the gray level image after gaussian filtering is processed when the gray level image is analyzed and calculated subsequently.
Further, in order to facilitate the analysis of different color information on the surface of the counter-punching composite die, the gray level image is converted into an HSV color space to obtain a corresponding HSV image, and a brightness value corresponding to each pixel point in the gray level image can be obtained according to the HSV image; meanwhile, in order to facilitate the comparative analysis of the punching compound die, the embodiment of the invention also correspondingly obtains a standard CAD image of the punching compound die, wherein the standard CAD image comprises the standard position and the standard shape of each punching hole of the punching compound die; and the standard CAD image is consistent with the gray level image in size, and the standard CAD image is recorded as a standard image for subsequent analysis.
Step S200, carrying out edge detection on the gray level image to obtain all closed edges in the gray level image; and acquiring the central point of each closed edge and the central point of each hole site in the standard image of the surface of the punching compound die, and acquiring the hole site offset according to the Euclidean distance between the central point of each closed edge and the central point of the corresponding hole site in the standard image.
In order to analyze each punched hole in the gray level image on the surface of the punching composite die, the approximate position of each punched hole is firstly obtained to divide the position of each punched hole for facilitating subsequent analysis and processing, so that the Canny operator is adopted to carry out edge detection on the gray level image of the punching composite die in the embodiment of the invention to obtain all edge pixel points and all edge contours in the gray level image, and the edge image obtained by edge detection of the gray level image is a binary image only containing the edge pixel points and a black background area.
However, errors may exist in all edges in the gray-scale image detected by the Canny operator, false edge information exists, and the amount of subsequent analysis calculation is increased, so that all the edges in the gray-scale image are preliminarily screened: all hole sites on the punching compound die are generally circular and are closed-shaped areas, so that the edges corresponding to the hole sites also need to be closed edges, and all pixel points in the binary images of the edges are subjected to traversal search based on the closed edges to judge whether all the edges are closed edges or not, so that preliminary screening is completed; the method for judging the closed edge specifically comprises the following steps:
taking any edge pixel point in the binary image as a starting point, searching the edge pixel points in the eight neighborhood directions of the edge pixel point one by one, and assuming that the edge pixel point is reached after n times of searchingIf the edge pixel point is searched at the momentAnd the second edge pixel point traversed at the beginning of the searchIf the pixel point is the same pixel point, the edge pixel point in the n-1 searching is judgedWhether or not to be correlated with the starting point at the time of searchIf the searched edge pixels are the same pixel, the searched edge pixels are the edge pixels which finally return to the initial position, namely the edge formed by the edge pixels is a closed edge; referring to FIG. 2, a schematic diagram of a closed edge and a non-closed edge is shown.
By analogy, traversing search is carried out on all edge pixel points in the binary image, whether the edge formed by the edge pixel points is a closed edge is judged according to a judging method of the closed edge, all closed edges in the binary image are obtained, and correspondingly, all closed edges in the gray image are obtained based on the positions of all the closed edges in the binary image; edges which do not meet the requirement of closed edges are considered as false edges, edge pixel points corresponding to all the false edges are abandoned, and subsequent analysis and calculation are not participated.
All closed edges in the gray level image are obtained, and each closed edge corresponds to different hole positions on the punching compound die; in the actual working process of the punching compound die, the punching compound die can be improperly operated when certain mechanical devices are processed, so that the hole positions are damaged or deformed, the original shapes of the hole positions can be damaged, but the number of the hole positions cannot be changed, and the number of the obtained closed edges is also the number of the hole positions; all hole sites on the upper die and the lower die of the punching compound die are generally regular geometric shapes, so that the shape of the closed edge is analyzed to preliminarily judge whether the shape of each hole site on the punching compound die deviates.
Specifically, a central point of each hole site shape in the standard image obtained in step S100 is obtained, and the position of the central point is marked; because the position of the hole site on the punching compound die is fixed, although the hole site can be fixed in the actual working processThe positions of the hole sites are fixed and unchanged when the hole sites deform, so that different hole sites are respectively marked in the standard image, the gray image corresponds to the position of each hole site in the standard image, the mark of each hole site in the gray image is correspondingly obtained, and the positions of the central points of the shapes of all the hole sites in the obtained standard image are respectively marked as the central points of the shapes of all the hole sites in the standard imageWherein s represents the number of hole sites in the standard image; because the number of hole sites cannot be changed, the coordinates of the central points of all the closed edges in the gray scale image are further acquired, and the coordinates of the central points of all the closed edges are respectively recorded asWherein s represents the number of all closed edges in the gray level image, and is consistent with the number of hole sites in the standard image; and analyzing and calculating according to the coordinates of the center point of each closed edge and the position of the center point of each hole site in the standard image, so that the hole site offset of each closed edge with deviation is quantized, and the result is more visual.
Specifically, the method for obtaining the hole site offset of each closed edge comprises the following steps:
considering that the hole site represented by the punching composite die is regular in shape, the circular hole site is defaulted in the embodiment of the invention, so that when a technician operates improperly or the parameters of a processing device set error in the processing and manufacturing process, the surface of the punching composite die is subjected to high-strength mechanical external force, rigid deformation occurs, and the deformed position has great influence on the hole site on the surface of the punching composite die; when the hole position deforms, the hole position can generate large deformation offset in different directions, so that straight lines in different directions are constructed by passing through the center point of each closed edge, the direction in the embodiment of the invention is that every 1 degree in 360 degrees is one direction, 360 straight lines are obtained according to the center point, two intersection points exist between each straight line and the closed edge, the length of a line segment between every two intersection points is obtained, the length of the line segment in the embodiment of the invention is also the Euclidean distance between the two intersection points, the maximum value and the minimum value of the lengths of the line segments corresponding to the straight lines in all directions are selected, the offset factor of the closed edge is judged according to the difference between the maximum value and the minimum value, and the calculation method of the offset factor comprises the following steps:
wherein the content of the first and second substances,is shown asDeviation of individual closure edges;the length of a line segment in the 1 st direction corresponding to the central point is represented;representing the length of the line segment in the 360 th direction corresponding to the central point;represents a maximum function;represents a minimum function;denotes the firstThe perimeter of each closed edge and the perimeter acquisition are known technologies, which are not described in detail, and the deviation of each closed edge is uniformly quantified by the perimeter.
By analogy, the offset factor corresponding to each closed edge is obtained, and the larger the difference between the line segments formed by the intersection points of the straight lines constructed by the center points of the closed edges and the closed edges is, the larger the difference in the shape of the closed edges is, the larger the deviation degree from the circle is, the larger the corresponding offset factor is, and the smaller the regularity of the shape corresponding to the closed edges is.
Further, calculating the difference between the coordinates of the center point of the closed edge and the position of the center point of the corresponding hole site in the standard image, wherein the larger the difference is, the shape difference between the shape of the closed edge and the shape of the corresponding hole site in the standard image is, and the Euclidean distance between the coordinates of the center point of the closed edge and the position of the center point of the corresponding hole site in the standard image is taken as the measurement of the difference, and the calculation of the Euclidean distance is a common means and is not described in detail; will be firstThe Euclidean distance between the coordinate of the central point of each closed edge and the position of the central point of the corresponding hole site in the standard image is recorded asAccording to the firstObtaining the offset factor of each closed edge and the Euclidean distance between the coordinate of the center point of the closed edge and the position of the center point of the corresponding hole site in the standard imageHole site offset for each closed edge, the hole site offset calculated as:
wherein the content of the first and second substances,is shown asHole site offset of each closed edge;denotes the firstAn offset factor for each closed edge;is shown asCoordinates of the center point of each closed edgeThe position of the central point of the corresponding hole site in the standard imageThe euclidean distance between them.
The larger the offset factor of the closed edge is, and the larger the Euclidean distance between the closed edge and the central point between the corresponding hole sites in the standard image is, the larger the shape deformation degree and the offset degree of the closed edge are, and the more serious the hole site deformation is, so the larger the hole site offset corresponding to the closed edge is.
Step S300, respectively acquiring closed boundary images of an upper die and a lower die of the punching compound die, and respectively performing characteristic point analysis on corresponding closed edges in the closed boundary images of the upper die and the lower die to obtain a first characteristic point sequence and a second characteristic point sequence; and acquiring a matching matrix according to the first characteristic point sequence and the second characteristic point sequence, acquiring a matching value between the first characteristic point and the second characteristic point based on the matching matrix, and acquiring the matching degree between corresponding closed edges according to the matching value.
In the process of machining mechanical parts by using the punching compound die, a coarse material to be machined is generally placed between an upper die and a lower die of the punching compound die, and then is subjected to punching forging on a punch press; therefore, the hole site matching between the upper die and the lower die of the punching composite die has a great influence on parts in the processing process, when the hole site deforms due to a mechanical external force, the hole site between the corresponding upper die and the corresponding lower die may not be complete enough or have poor shape consistency, the quality of processing by using the punching composite die at the moment cannot be guaranteed, and therefore, the corresponding analysis needs to be performed on the closed edges in the upper die and the lower die of the punching composite die.
Since the grayscale images acquired in step S100 include the grayscale images of the upper die and the lower die of the stamping compound die, the closed boundary image of the grayscale image of the upper die and the closed boundary image of the grayscale image of the lower die, which are the images marked with all the closed edges, can be obtained based on the same method as in step S200; the standard image also includes a standard image of the upper die and a standard image of the lower die, and the standard image includes a mark for each hole position, it should be noted that, since the upper die and the lower die are matched in the working process, the mark of each hole position of the upper die and the mark of the same hole position of the lower die in the standard image are in a corresponding matching relationship, and then the mark of each closed edge in the upper die is obtained according to the mark of each hole position in the standard image of the upper die and recorded asCorrespondingly, the corresponding mark number of each hole position in the lower die is obtained according to the mark number of each hole position in the standard image of the lower dieClosing edgeAnd closing edgeA closed edge corresponding to the matched hole site; the characteristic shapes of the matched closed edges are analyzed to determine the correlation matching degree between the two closed edges at the moment, and whether the hole positions corresponding to the two closed edges can be well matched or not is further analyzed to work.
Specifically, feature point analysis is performed on each closed edge in the closed boundary image corresponding to the upper die to obtain feature points of the closed edge, and the feature points corresponding to each closed edge in the closed boundary image corresponding to the upper die form a first feature point sequence; correspondingly, performing feature point analysis on each closed edge in the closed boundary image corresponding to the lower die to obtain feature points of the closed edge, and forming a second feature point sequence by using the feature points corresponding to each closed edge in the closed boundary image corresponding to the lower die; in the method for extracting the feature points in the embodiment of the invention, the existing SURF algorithm is adopted, and different feature point extraction methods can be selected in other embodiments, so that a first feature point sequence of each closed edge in the closed boundary image of the upper die and a second feature point sequence of each closed edge in the closed boundary image of the lower die are obtained.
When the upper die and the lower die are normally matched to work, a certain relationship exists between the hole sites of the upper die and the hole sites of the lower die, for example, the hole sites of the upper die and the hole sites of the lower die are equal or are enlarged and reduced in equal proportion, so that a corresponding relationship exists between the hole sites of the upper die and the hole sites of the lower die, and the matching degree between the hole sites of the upper die and the hole sites of the lower die is analyzed based on the relationship between the corresponding hole sites of the upper die and the lower die, namely the matching degree between the closed edge in the closed boundary image of the upper die and the corresponding closed edge in the closed boundary image of the lower die is analyzed.
Closed edges in the closed boundary image of the above moldClosed edge in closed boundary image with lower dieFor example, assume a closed edgeThe corresponding first characteristic point sequence isWherein,Indicating a closed edgeThe first 1 st feature point is set as the feature point,indicating a closed edgeGo to the firstThe number of the characteristic points is one,to close the edgeThe number of all the characteristic points; closure edgeThe corresponding second characteristic point sequence isWherein, in the step (A),indicating a closed edgeThe first 1 st characteristic point is set as the characteristic point,indicating a closed edgeTo go toThe number of the characteristic points is one,indicating a closed edgeThe number of all the characteristic points; if the surface of the punching compound die is normal, the closed edge of the upper dieClosing edge with lower dieIf there is a corresponding relationship between the two, the characteristic point sequence corresponding to the closed edge is used for representation, that is, the closed edge isFirst feature point sequence and closed edge ofThe following affine transformation relationship exists between the second feature point sequences:
wherein, the first and the second end of the pipe are connected with each other,indicating a closed edgeThe first sequence of feature points of (1);indicating a closed edgeSecond feature of (2)A sequence of characteristic points;a matching matrix is represented.
It should be noted that, in the embodiment of the present invention, analyzing the feature points is based on the coordinate information of the feature points, and a matching matrix is set in the embodiment of the present inventionIs 3*3, then the matrix is matchedCan be expressed as:
Then, each matching coefficient in the matching matrix needs to be fitted to obtain, F first characteristic points and F second characteristic points are correspondingly selected from the first characteristic point sequence and the second characteristic point sequence, wherein F is a positive integer; and respectively constructing a first matrix and a second matrix by using the F first characteristic points and the F second characteristic points, taking a multiplication result of the matching matrix and the first matrix as the second matrix, and obtaining each element in the matching matrix based on matrix operation.
In the embodiment of the invention, the coordinate positions of continuous 9 characteristic points in the first characteristic point sequence and the second characteristic point sequence are respectively selected to form a first matrix and a second matrix, and a corresponding matching matrix is obtained based on the first matrix and the second matrixEach of the matching coefficients of (a) to (b),the specific calculation method is matrix operation in mathematics and is not described in detail; and obtaining each matching coefficient in the matching matrix according to the first matrix corresponding to the first characteristic point sequence and the second matrix corresponding to the second characteristic sequence.
Then, a closed edge may be calculated based on the matching matrixAnd closing edgeThe matching degree between the first characteristic point and the second characteristic point is obtained by multiplying the first ratio, the second ratio and the mapping coefficient. Judging whether the matching value of the first feature point and the corresponding second feature point is within a preset matching range, if so, successfully matching the first feature point and the second feature point; and calculating the cumulative sum of the matching values between the first characteristic points and the second characteristic points which are successfully matched on the corresponding closed edges, and normalizing the cumulative sum to obtain the matching degree between the corresponding closed edges.
First calculate the closure edgeAnd closing edgeMatching values between corresponding feature points, and obtaining a closed edge by summing the matching values between a plurality of pairs of successfully matched feature pointsAnd closing edgeThe degree of matching between the feature points, and the formula for calculating the matching value between the feature points such asThe following:
wherein the content of the first and second substances,indicating the closing edge of the upper mouldsMiddle first characteristic pointClosing edge with lower dieMiddle second characteristic pointA match value therebetween;first characteristic point of upper moldThe brightness value of (a);second characteristic point of lower dieThe brightness value of (a);first characteristic point of upper moldThe position coordinates of (a);second characteristic point of lower dieThe position coordinates of (a);respectively, represent the matching coefficients in the matching matrix.
The brightness values of the positions of the pixel points on the punching compound die are relatively consistent, and when the punching compound die is worn or deformed, the color characteristics are different, so that the difference of the color characteristics is reflected by the ratio of the brightness values of the characteristic points, namely the second ratioThe closer the value of (1) is, the closer the color characteristic between the two characteristic points is; from the above affine transformation relationship between the first feature point and the second feature point, the coordinates of the first feature point are foundAnd coordinates of the second feature pointIf the first feature point corresponds to the second feature point, the coordinate relationship between the feature points should conform to the affine transformation relationship, and in order to conform to the operation between the matrices, the coordinate vector is setAndrespectively complementing 0 and transposing to obtain the coordinate matrix of the first characteristic pointAnd a coordinate matrix of the second feature pointThe transposition process is the prior known technology and is not described again; then the coordinate matrix for the first feature pointAnd a coordinate matrix of the second feature pointThe following requirements should be satisfied:
then, according to the calculation principle of the matrix, it can be deduced that:,(ii) a Therefore, the product of the first ratio and the mapping coefficient in the matching value calculation formulaWhen the value of (A) is 1, that is to say,Two equations are true, thenWhen the matching value is calculatedThe larger the difference between the value of (1) and (1), the more dissimilar and unmatched the characteristics between the two characteristic points.
By analogy, the matching value between the feature points in the corresponding closed edge in the closed boundary image of the upper die and the closed boundary image of the lower die is obtainedIf the matching value is larger, it is indicated that the two feature points are more likely to be corresponding feature points and the feature information is more similar, and for more intuitive analysis, the matching range is set in this embodiment, and when the matching value between the two feature points is largerWhen the matching range is within the matching range, the two feature points are successfully matched, and as a preferred example, the matching range is set as。
Counting the number of all successfully matched feature points between the two closed edges, and accumulating the matching values of all successfully matched feature points to obtain the matching degree between the corresponding closed edges, namely the matching degree between the corresponding hole positions of the upper die and the lower die; in order to facilitate subsequent analysis and quantification, the matching degree is subjected to normalization processing, and the normalization processing method is the existing means and is not explained; the lower the matching degree is, the more difficult the matching work among hole positions is, and the more possible problems occur to the quality of the punching compound die; therefore, in the embodiment of the present invention, the hole locations with the matching degree greater than the preset threshold are recorded as successful matching hole locations, that is, the matching edge pair is obtained according to the matching degree between the closing edge of the upper die and the closing edge of the corresponding lower die, and the two successfully matched closing edges are one matching edge pair.
As a preferable example, the preset threshold value is set to 0.8 in the embodiment of the present invention.
S400, obtaining all hole site matching pairs in the upper die and the lower die according to the matching degree; and counting the number of all hole site matching pairs in the upper die and the lower die to obtain the abnormal probability of the punching composite die, and judging the quality based on the abnormal probability.
In step S300, the matching degree of the closed edge in the closed boundary image corresponding to the upper die and the closed edge in the closed boundary image corresponding to the lower die is obtained, and a plurality of groups of matched edge pairs are correspondingly obtained, where each group of matched edge pairs is also a group of hole site matched pairs, and the quality condition of the hole punching composite die is further analyzed according to all the hole site matched pairs.
Acquiring the number of all unmatched feature points on two closed edges in each hole site matching pair and the sum of the number of all feature points on two closed edges in each hole site matching pair, and taking the ratio result of the number of all unmatched feature points on the two closed edges and the sum of the number of all feature points on the two closed edges as an unmatched ratio; calculating the product result between the unmatched ratio and the reciprocal of the matching degree of the hole site matching pairs, wherein the sum of the product results corresponding to all the hole site matching pairs is an abnormal factor; and acquiring the sum of hole site offsets of all closed edges in all hole site matching pairs, calculating the product of the sum of the hole site offsets and the abnormal factor, and performing normalization processing to obtain the abnormal probability.
Specifically, the number of the feature points which are not successfully matched on each closed edge in all hole site matching pairs of the punching compound die is counted, and the number of the first feature points which are not successfully matched on the closed edge corresponding to the upper die in the hole site matching pairs is recorded as the number of the first feature points which are not successfully matched on the closed edge corresponding to the upper die in the hole site matching pairsCorrespondingly, the number of the second feature points of the closing edge corresponding to the lower die which are not successfully matched is recorded asObtaining abnormal factors according to the number of all unmatched feature points in all hole site matching pairs in the upper die and the lower die and the matching degree corresponding to the hole site matching pairs, wherein the abnormal factors are specifically calculated as follows:
wherein the content of the first and second substances,representing abnormal factors of the punching compound die;is shown asMatching degree of the hole position matching pairs;is shown asThe number of the first characteristic points of the upper die of the group hole site matching pair which are not successfully matched is larger than the number of the first characteristic points of the upper die of the group hole site matching pair which are not successfully matched;is shown asThe number of the second characteristic points of the group hole position matching pair which are not successfully matched on the lower die is larger than the number of the second characteristic points;representing the number of first characteristic points of the hole site matching pairs on the upper die;representing the second characteristic point quantity of the hole position matching pairs on the lower die;the number of all hole site matching pairs on the punching compound die is shown.
The larger the value of (A) is, the smaller the matching degree of the hole site matching pair is, the smaller the matching degree between two hole sites is, and the quality problem may exist, so the larger the corresponding abnormal factor is;the larger the value of (A) is, the more the number of feature points which are not successfully matched in the corresponding two hole sites is, although the matching degree meets the preset thresholdThe larger the value, but the more likely there may be a difference between actual hole locations.
Further, the relative offset of the punching compound die is obtained according to the hole site offset of each closed edge, for convenience of analysis, the relative offset is obtained only according to the hole site offset of the hole site matching pair successfully matched in the embodiment of the invention, and the specific relative offset is calculated as follows:
wherein the content of the first and second substances,representing the relative offset of the punching compound die;representing the number of hole site matching pairs of an upper die and a lower die of the punching compound die;is shown asHole site offset of each closure edge.
When the relative offset corresponding to the punching compound die is larger, the larger the hole site offset corresponding to each closed edge is, the larger the hole site offset is, the larger the deformation condition of the closed edge is, and the worse the quality is, so that the larger the relative offset of the punching compound die is, the worse the quality of the punching compound die is.
According to the relative offset and the abnormal factor of the punching compound die, the quality evaluation index of the punching compound die is constructed, in the embodiment of the invention, the abnormal probability is used as the basis of quality evaluation, and the calculation method of the abnormal probability comprises the following steps:
wherein,Representing the abnormal probability of the punching compound die;represents an anomaly factor;represents a relative offset;expressing a range normalization function in order to make the result of the anomaly probability。
The larger the abnormal factor is, the worse the quality of the corresponding punching compound die is, the larger the relative offset is, the more serious the deformation of the punching compound die is, and the worse the quality is, so that the larger the value of the abnormal probability is, the worse the quality of the punching compound die is, and the existence of defects of the punching compound die is judged; in the embodiment of the invention, an abnormal threshold value is set for judgment, and when the abnormal probability is greater than the abnormal threshold value, the quality of the punching compound die is judged to be poor; as a preferable example, the abnormality threshold value is set to 0.85 in the present embodiment.
In summary, in the embodiment of the present invention, the gray image of the surface of the punching compound die is obtained; carrying out edge detection on the gray level image to obtain all closed edges in the gray level image; acquiring the central point of each closed edge and the central point of each hole site in a standard image of the surface of the punching compound die, and acquiring the hole site offset according to the Euclidean distance between the central point of each closed edge and the central point of the corresponding hole site in the standard image; respectively acquiring closed boundary images of an upper die and a lower die of the punching composite die, and respectively performing characteristic point analysis on corresponding closed edges in the closed boundary images of the upper die and the lower die to obtain a first characteristic point sequence and a second characteristic point sequence; acquiring a matching matrix according to the first characteristic point sequence and the second characteristic point sequence, acquiring a matching value between the first characteristic point and the second characteristic point based on the matching matrix, and acquiring a matching degree between corresponding closed edges according to the matching value; obtaining all hole site matching pairs in the upper die and the lower die according to the matching degree; and counting the number of all hole site matching pairs in the upper die and the lower die to obtain the abnormal probability of the punching composite die, and performing quality judgment based on the abnormal probability to improve the accuracy of surface quality detection of the punching composite die.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
All the embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that are within the spirit of the present invention are intended to be included therein.
Claims (10)
1. A punching compound die surface quality detection method based on computer vision is characterized by comprising the following steps:
obtaining a gray image of the surface of the punching compound die;
performing edge detection on the gray level image to obtain all closed edges in the gray level image; acquiring a central point of each closed edge and a central point of each hole site in a standard image of the surface of the punching compound die, and acquiring hole site offset according to the Euclidean distance between the central point of each closed edge and the central point of the corresponding hole site in the standard image;
respectively acquiring closed boundary images of an upper die and a lower die of the punching composite die, and respectively performing characteristic point analysis on corresponding closed edges in the closed boundary images of the upper die and the lower die to obtain a first characteristic point sequence and a second characteristic point sequence; acquiring a matching matrix according to the first characteristic point sequence and the second characteristic point sequence, acquiring a matching value between the first characteristic point and the second characteristic point based on the matching matrix, and acquiring a matching degree between corresponding closed edges according to the matching value;
obtaining all hole site matching pairs in the upper die and the lower die according to the matching degree; and counting the number of all hole site matching pairs in the upper die and the lower die to obtain the abnormal probability of the punching composite die, and judging the quality based on the abnormal probability.
2. A computer vision-based method for inspecting surface quality of a punching compound die according to claim 1, wherein the step of obtaining a matching value between a first feature point and a second feature point based on the matching matrix comprises:
acquiring an HSV image corresponding to the gray image, and acquiring the brightness value of each pixel point based on the HSV image;
and constructing a mapping coefficient based on the matching matrix and the coordinate of the first characteristic point, acquiring a first ratio of the abscissa and the ordinate of the second characteristic point and a second ratio of the brightness value of the first characteristic point and the brightness value of the second characteristic point, and obtaining a matching value between the first characteristic point and the second characteristic point by the product of the first ratio, the second ratio and the mapping coefficient.
3. A computer vision based method for inspecting surface quality of a punching compound die according to claim 2, wherein the step of obtaining the matching degree between the corresponding closed edges according to the matching value comprises:
judging whether the matching value of the first feature point and the corresponding second feature point is within a preset matching range, if so, successfully matching the first feature point and the second feature point;
and calculating the cumulative sum of the matching values between the first characteristic points and the second characteristic points which are successfully matched on the corresponding closed edges, and normalizing the cumulative sum to obtain the matching degree between the corresponding closed edges.
4. A computer vision-based method for inspecting surface quality of a punching compound die according to claim 1, wherein the step of obtaining a matching matrix according to the first characteristic point sequence and the second characteristic point sequence comprises:
correspondingly selecting F first characteristic points and second characteristic points from the first characteristic point sequence and the second characteristic point sequence, wherein F is a positive integer;
and respectively constructing a first matrix and a second matrix by using the F first characteristic points and the F second characteristic points, taking a multiplication result of the matching matrix and the first matrix as the second matrix, and obtaining each element in the matching matrix based on matrix operation.
5. A punching compound die surface quality detection method based on computer vision according to claim 4, characterized in that the step of constructing mapping coefficients based on the matching matrix and the coordinates of the first feature point comprises:
the expression of the mapping coefficient is:
6. The method of claim 1, wherein the step of obtaining the hole site offset according to the Euclidean distance between the center point of each closed edge and the center point of the corresponding hole site in the standard image comprises:
straight lines in different directions are constructed through the center point of the closed edge, each straight line and the closed edge are intersected at two points to form a corresponding line segment, and the length of the line segment formed by the straight lines in different directions is obtained;
selecting the maximum value and the minimum value of the lengths of all line segments, and obtaining an offset factor according to the difference value of the maximum value and the minimum value, wherein the offset factor and the difference value are in positive correlation;
and taking the product of the Euclidean distance between the central point of the closed edge and the central point of the corresponding hole site in the standard image and the offset factor as the hole site offset of the closed edge.
7. The method for detecting the surface quality of the punching composite die based on the computer vision as claimed in claim 3, wherein the step of counting the number of all the hole site matching pairs in the upper die and the lower die to obtain the abnormal probability of the punching composite die comprises:
acquiring the number of all unmatched feature points on the two closed edges in each hole site matching pair and the sum of the number of all feature points on the two closed edges in the hole site matching pair, and taking the ratio result of the number of all unmatched feature points on the two closed edges and the sum of the number of all feature points on the two closed edges as an unmatched ratio;
calculating the product result between the unmatched ratio and the reciprocal of the matching degree of the hole site matching pairs, wherein the sum of the product results corresponding to all the hole site matching pairs is an abnormal factor;
and acquiring the sum of hole site offsets of all closed edges in all the hole site matching pairs, calculating the product of the sum of the hole site offsets and the abnormal factor, and performing normalization processing to obtain the abnormal probability.
8. The method for detecting the surface quality of the punching compound die based on the computer vision as claimed in claim 1, wherein the step of judging the quality based on the abnormal probability comprises:
and when the abnormal probability is greater than a preset abnormal threshold value, the surface quality of the punching compound die is poor.
9. A computer vision based method for inspecting the surface quality of a punching compound die according to claim 1, wherein the step of acquiring all closed edges in the gray image comprises:
traversing search is carried out on each edge pixel point in the gray level image, and the traversing search direction is the eight neighborhood direction of the edge pixel point;
when the edge pixel point of the nth traversal search and the edge pixel point of the second traversal search are the same pixel point, and the edge pixel point of the n-1 traversal search and the edge pixel point of the first traversal search are also the same pixel point, the edge of the traversal search is a closed edge; wherein n is a positive integer.
10. A computer vision based method for inspecting the surface quality of a punching compound die according to claim 1, characterized in that the closed boundary image is an image including all closed edges.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211420216.4A CN115511884B (en) | 2022-11-15 | 2022-11-15 | Punching compound die surface quality detection method based on computer vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211420216.4A CN115511884B (en) | 2022-11-15 | 2022-11-15 | Punching compound die surface quality detection method based on computer vision |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115511884A true CN115511884A (en) | 2022-12-23 |
CN115511884B CN115511884B (en) | 2023-03-07 |
Family
ID=84514595
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211420216.4A Active CN115511884B (en) | 2022-11-15 | 2022-11-15 | Punching compound die surface quality detection method based on computer vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115511884B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116503413A (en) * | 2023-06-29 | 2023-07-28 | 常熟林芝电子技术有限公司 | Filament quality intelligent detection method based on computer vision |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111062927A (en) * | 2019-12-18 | 2020-04-24 | 广东电网有限责任公司 | Method, system and equipment for detecting image quality of unmanned aerial vehicle |
CN112466404A (en) * | 2020-12-14 | 2021-03-09 | 浙江师范大学 | Unsupervised clustering method and unsupervised clustering system for metagenome contigs |
CN115294338A (en) * | 2022-09-29 | 2022-11-04 | 中威泵业(江苏)有限公司 | Impeller surface defect identification method |
-
2022
- 2022-11-15 CN CN202211420216.4A patent/CN115511884B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111062927A (en) * | 2019-12-18 | 2020-04-24 | 广东电网有限责任公司 | Method, system and equipment for detecting image quality of unmanned aerial vehicle |
CN112466404A (en) * | 2020-12-14 | 2021-03-09 | 浙江师范大学 | Unsupervised clustering method and unsupervised clustering system for metagenome contigs |
CN115294338A (en) * | 2022-09-29 | 2022-11-04 | 中威泵业(江苏)有限公司 | Impeller surface defect identification method |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116503413A (en) * | 2023-06-29 | 2023-07-28 | 常熟林芝电子技术有限公司 | Filament quality intelligent detection method based on computer vision |
CN116503413B (en) * | 2023-06-29 | 2023-09-08 | 常熟林芝电子技术有限公司 | Filament quality intelligent detection method based on computer vision |
Also Published As
Publication number | Publication date |
---|---|
CN115511884B (en) | 2023-03-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116721106B (en) | Profile flaw visual detection method based on image processing | |
CN109087274B (en) | Electronic device defect detection method and device based on multi-dimensional fusion and semantic segmentation | |
CN114529549B (en) | Cloth defect labeling method and system based on machine vision | |
CN112330628A (en) | Metal workpiece surface defect image detection method | |
CN115345885A (en) | Method for detecting appearance quality of metal fitness equipment | |
CN114820773B (en) | Silo transport vehicle carriage position detection method based on computer vision | |
CN114549981A (en) | Intelligent inspection pointer type instrument recognition and reading method based on deep learning | |
CN113793337B (en) | Locomotive accessory surface abnormal degree evaluation method based on artificial intelligence | |
CN115830015B (en) | Hardware stamping accessory quality detection method based on computer vision | |
CN115511884B (en) | Punching compound die surface quality detection method based on computer vision | |
CN115311289A (en) | Method for detecting oil stain defects of plain-color cloth | |
CN115272336A (en) | Metal part defect accurate detection method based on gradient vector | |
CN115272350A (en) | Method for detecting production quality of computer PCB mainboard | |
CN116758067B (en) | Metal structural member detection method based on feature matching | |
CN111539951B (en) | Visual detection method for outline size of ceramic grinding wheel head | |
CN117893532A (en) | Die crack defect detection method for die forging rigging based on image processing | |
CN117314901A (en) | Scale-adaptive chip detection neural network system | |
CN116740053A (en) | Management system of intelligent forging processing production line | |
CN116823708A (en) | PC component side mold identification and positioning research based on machine vision | |
CN113643290B (en) | Straw counting method and device based on image processing and storage medium | |
CN115330785A (en) | Textile vamp quality detection method | |
CN115471684A (en) | Injection molding workpiece template matching method, electronic equipment and storage medium | |
CN110827272B (en) | Tire X-ray image defect detection method based on image processing | |
CN114354631A (en) | Valve blank surface defect detection method based on vision | |
CN111854617A (en) | Micro drill bit size detection method based on machine vision |
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 |