CN116168349B - Big data-based die-casting visual monitoring system and method - Google Patents
Big data-based die-casting visual monitoring system and method Download PDFInfo
- Publication number
- CN116168349B CN116168349B CN202310451150.3A CN202310451150A CN116168349B CN 116168349 B CN116168349 B CN 116168349B CN 202310451150 A CN202310451150 A CN 202310451150A CN 116168349 B CN116168349 B CN 116168349B
- Authority
- CN
- China
- Prior art keywords
- die casting
- image
- die
- data
- castings
- 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
- 238000004512 die casting Methods 0.000 title claims abstract description 223
- 238000012544 monitoring process Methods 0.000 title claims abstract description 41
- 230000000007 visual effect Effects 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000004458 analytical method Methods 0.000 claims abstract description 38
- 230000002159 abnormal effect Effects 0.000 claims description 29
- 238000004364 calculation method Methods 0.000 claims description 20
- 238000007405 data analysis Methods 0.000 claims description 17
- 230000008859 change Effects 0.000 claims description 13
- 238000007781 pre-processing Methods 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 12
- 238000003860 storage Methods 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 238000004519 manufacturing process Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 238000004140 cleaning Methods 0.000 claims description 9
- 238000013500 data storage Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 101100134058 Caenorhabditis elegans nth-1 gene Proteins 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 230000007613 environmental effect Effects 0.000 abstract description 4
- 238000005266 casting Methods 0.000 description 12
- 239000013307 optical fiber Substances 0.000 description 8
- 230000001629 suppression Effects 0.000 description 8
- 230000009471 action Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 239000000047 product Substances 0.000 description 3
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 2
- 229910045601 alloy Inorganic materials 0.000 description 2
- 239000000956 alloy Substances 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- FYYHWMGAXLPEAU-UHFFFAOYSA-N Magnesium Chemical compound [Mg] FYYHWMGAXLPEAU-UHFFFAOYSA-N 0.000 description 1
- 229910001128 Sn alloy Inorganic materials 0.000 description 1
- ATJFFYVFTNAWJD-UHFFFAOYSA-N Tin Chemical compound [Sn] ATJFFYVFTNAWJD-UHFFFAOYSA-N 0.000 description 1
- HCHKCACWOHOZIP-UHFFFAOYSA-N Zinc Chemical compound [Zn] HCHKCACWOHOZIP-UHFFFAOYSA-N 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- LQBJWKCYZGMFEV-UHFFFAOYSA-N lead tin Chemical compound [Sn].[Pb] LQBJWKCYZGMFEV-UHFFFAOYSA-N 0.000 description 1
- 229910052749 magnesium Inorganic materials 0.000 description 1
- 239000011777 magnesium Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000005058 metal casting Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 229910052725 zinc Inorganic materials 0.000 description 1
- 239000011701 zinc Substances 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a large data-based die-casting visual monitoring system and a large data-based die-casting visual monitoring method, and belongs to the field of die-casting visual monitoring. According to the invention, images of adjacent die castings are subjected to comparison analysis according to the acquired data information, the errors of the die castings are judged according to the adjacent images, and when the errors are large, relevant technicians are reminded, so that the method is not influenced by environmental factors, the images do not need to be repeatedly called, the analysis speed is high, and the monitoring efficiency is high.
Description
Technical Field
The invention relates to the field of visual monitoring of die casting, in particular to a visual monitoring system and method of die casting based on big data.
Background
Die casting is a metal casting process and is characterized in that high pressure is applied to molten metal by utilizing the inner cavity of a die. The mold is typically machined from a stronger alloy. Most die cast castings are free of iron, such as zinc, copper, aluminum, magnesium, lead, tin, and lead-tin alloys and alloys thereof. Casting equipment and molds are expensive, so die casting processes are generally used only for mass production of large quantities of products. The manufacture of die-cast parts is relatively easy, which generally requires only four main steps, with low single cost increments. Die casting is particularly suitable for manufacturing a large number of small and medium-sized castings, and is therefore one of the most widely used of various casting processes. Compared with other casting technologies, the die-casting surface is smoother, and has higher dimensional consistency.
The die casting is a precise casting method, the dimensional tolerance of the die casting cast through die casting is very small, and the surface precision is very high, so that the die casting is required to be monitored in the die casting process, and the error of the die casting is reduced, however, in the prior art, the die casting image is compared with the standard image through collecting the die casting image, and the production monitoring of a large number of die castings, so that the environment where the die casting and the standard image are located is difficult to keep consistent, the influence of environmental factors is large, the analysis speed is low, the efficiency is low, the visual monitoring of a large number of die castings is difficult to be performed in time, and large error is easy to generate.
From this, how to carry out visual monitoring to die casting in time, not influenced by environmental factor, it is very necessary to improve quality monitoring's efficiency. Therefore, a need exists for a visual die casting monitoring system and method based on big data.
Disclosure of Invention
The invention aims to provide a die casting visual monitoring system and method based on big data, which are used for collecting basic data information, collecting die casting image information through camera equipment, analyzing and processing the collected image information, comparing and analyzing adjacent die casting images, classifying when abnormal die castings appear according to the analysis result, and displaying and reminding a user by voice so as to solve the problems in the background technology.
In order to solve the technical problems, the invention provides the following technical scheme: a visual die casting monitoring system based on big data, the visual monitoring system comprising: the system comprises a data acquisition module, a database, a data analysis module and a user reminding module;
the data acquisition module is connected with the database, the database is connected with the data analysis module, and the data analysis module is connected with the user reminding module; the data acquisition module is used for acquiring basic data information, acquiring die casting image information through the camera equipment, the database is used for carrying out encryption storage on acquired data and analysis results, the data analysis module is used for carrying out analysis processing on the acquired data, and the user reminding module is used for classifying when abnormal die castings appear according to the analysis results and reminding a user.
Further, the data acquisition module comprises a basic data input unit and an image acquisition unit, wherein the basic data input unit is used for inputting basic data information of die casting equipment and die castings, and the image acquisition unit is used for acquiring images of the die castings through the camera equipment, so that the state of each die casting can be clearly known.
Further, the database includes a data storage unit, a data encryption unit and a data cleaning unit, the data storage unit is used for storing collected data and analysis results through SAN technology, the SAN is a storage area network, the SAN is a high-speed special subnet connecting storage devices such as disk arrays, tapes and related servers through connection devices such as optical fiber hubs, optical fiber routers and optical fiber switches, the SAN provides higher throughput capacity for connection between the servers and the storage devices by means of optical fiber channels, supports farther distance and more reliable connection, the SAN can be a switched network or a shared network, the data encryption unit encrypts the whole process through a homomorphic encryption algorithm so as to ensure data security, prevent leakage of produced die casting information, the homomorphic encryption refers to an encryption algorithm meeting homomorphic operation properties of ciphertext, namely, after the data is homomorphic encrypted, specific calculation is performed on ciphertext, the obtained plaintext after corresponding homomorphic decryption is performed on the ciphertext calculation results is equivalent to direct identical calculation on plaintext data, so that the ciphertext is visible, and if the homomorphic calculation algorithm is in a homomorphic form of any encryption algorithm is supported in a homomorphic form; if the ciphertext is supported to be calculated in a partial form, for example, only addition, only multiplication or limited times of addition and multiplication are supported, the ciphertext is called semi-homomorphic encryption or partial homomorphic encryption, and the data cleaning unit is used for automatically cleaning the die casting data information meeting the requirements after the die casting production is completed, so that redundant data are cleaned, the storage space is saved, and the cost of data storage is reduced.
Further, the data analysis module comprises an image preprocessing unit and an error analysis unit, the image preprocessing unit is used for preprocessing and analyzing images according to collected image data, extracting outlines of die castings in the images, comparing adjacent die casting images, and because the number of die castings needing to be detected is extremely large, standard data images do not need to be repeatedly adjusted through comparison of adjacent images, complexity of an image comparison process is reduced, the realization mode is simple and quick, the stability of a system is improved, data information of each die casting can be quickly known, meanwhile, environment changes of adjacent images are small, interference generated is small, images are convenient to extract, errors are small, the error analysis unit is used for analyzing and processing errors between the identified images, monitoring quality of the die castings according to the continuous changes of the errors, and avoiding error accumulation of the adjacent die casting images, so that overall errors are large.
Further, the user reminds the module to include die casting classification unit, screen display unit and pronunciation to remind the unit, die casting classification unit is used for separating unusual die casting according to analysis result, and relevant technicians of being convenient for overhauls, improves die casting finished product's correct rate, screen display unit is used for showing unusual die casting condition to relevant technicians through screen display equipment, pronunciation reminds relevant technicians through pronunciation, and the relevant technicians of being convenient for in time inspect the maintenance to the die casting.
A visual monitoring method for die casting based on big data comprises the following steps:
s1, inputting basic data information of die casting equipment and die castings, and acquiring images of the die castings through camera equipment;
s2, preprocessing and analyzing the image according to the acquired image information, extracting the outline of the die casting in the image, and comparing adjacent die casting images;
s3, carrying out corresponding matching on the characteristic points according to the preprocessed image, and carrying out error analysis;
s4, classifying when abnormal die castings appear according to analysis results, and reminding related technicians through display equipment and voice.
Further, in step S2, according to the acquired die casting image information, extracting a die casting contour in the image, and comparing adjacent images;
s201, performing binarization processing on the acquired image;
s202, performing two-dimensional Gaussian filtering on the image;
placing the image in a coordinate system which can be set by the related technicians, wherein the image isWhere x and y represent pixel values of the image, a two-dimensional gaussian filter function f is calculated by the following formula:
wherein,,expressed as the variance of the gaussian function, two one-dimensional rank filters are obtained by decomposition:
s203, eliminating noise interference of the image through convolution operation;
two one-dimensional row-column filters are respectively connected with the imagePerforming convolution calculation to obtain output:
wherein,,represented as convolution operator, implements low-pass filtering to eliminate noise interference;
s204, extracting edge points of the die casting in the image according to the image gradient, and connecting to form the outline of the die casting;
the gradient line direction is determined by the following formulaAnd (3) performing calculation: />;
For gradient amplitudePerforming non-maximum suppression when gradient amplitude of pixel point on image +.>Less than along the gradientThe gradient amplitude of two adjacent pixel points in the line direction is judged to be a non-edge point, and the gradient amplitude is +.>Setting to 0 and vice versa to 1, non-maximum suppression, as the name implies, is suppression of elements that are not maxima, which can be understood as local maximum searching. The local representation is a neighborhood, the neighborhood has two variable parameters, one is the dimension of the neighborhood, and the other is the size of the neighborhood, and the essence is to search local maxima and inhibit non-maxima elements; performing double thresholding separation on non-maximum suppression images, and setting a threshold valueAnd->Wherein->By a high threshold->Dividing to obtain image->From low threshold->Dividing to obtain image->In the form of imagesBased on, image->In order to supplement edge connection, extracting the outline of the die casting, processing the image by using a high threshold value and a low threshold value, and reserving more detail information in the image so that the image comparison result is more accurate;
first, theContour image data extracted from the individual die castings form a vector set +.>Contour image data extracted from the nth die casting form a vector set +.>Image contrast index +.>And (3) performing calculation:
wherein N is expressed as the number of die castings, and a threshold value is setWhen->When the comparison of the images is similar, the error meets the standard, the die casting meets the requirement, and when +.>And when the comparison error of the image is large, the image indicates that the die casting does not meet the requirement, the mark is an abnormal die casting, and the step S4 is directly executed to remind relevant technicians of checking and maintaining.
Further, in step S3, according to the image of the die casting extraction contour, when the image contour of the die casting meets the requirement, the feature points of the comparison image are correspondingly matched, and the error is analyzed;
the extracted contour image data forms a setWherein n is the number of die castings, the thExtraction of individual die castingsThe variance of the contour image of (2) is +.>First->The variance of the profile image extracted from the individual die castings is +.>K is expressed as the scale ratio between two adjacent images, then +.>Convolution of the image of the individual die-cast parts with the Gaussian kernel L and +.>Convolution of the image of the individual die cast parts with the Gaussian kernel>The method comprises the following steps:
the scale index E is calculated by the following formula:
taylor expansion is carried out on the two-dimensional Gaussian filter function f through the following formula to obtain an expanded function:
Wherein,,values denoted as extreme points>For scale space coordinates, T is denoted transpose, when located at the extreme point +.>At the time of Taylor expansion function->Setting the threshold value as +.>When->When the point contrast is low, deleting from the candidate points;
converting the image into a real symmetric matrix Z, and screening the characteristic points through the following formula:
wherein,,expressed as the sum of matrix eigenvalues, +.>Represented as determinant of matrix>Expressed as the ratio of the maximum characteristic value to the minimum characteristic value, when the characteristic points meet the inequality, the characteristic points meeting the requirements are the corresponding characteristic points, and the gradient amplitude is combined +.>And gradient line direction->Realizing the characteristics of adjacent contrast imagesThe sign points are correspondingly matched;
after the first die casting can be confirmed by relevant technicians whether the first die casting meets the requirements, the acquired image provides a basis for subsequent visual monitoring, and the same characteristic point change vector is obtained according to the corresponding matching of the characteristic points of the adjacent images to form a setWherein n is the number of die castings, and the included angle set is +.>In the first placeAt each die casting, the characteristic point change vector is +.>The included angle is->At the n-1 th die casting, the characteristic point variation vector is +.>The included angle is->In->At the die casting part, the characteristic point change vector is +.>The limiting distance of the nth-1 die casting is +.>And (3) performing calculation:
then associate the included angleThe related included angle is the connection line of the n-1 th feature point and the first feature point and the n-1 th feature point and the +.>Included angles of the connecting lines of the characteristic points;
the following formula is used for the firstLimiting distance of individual die castings->And (3) performing calculation:
the feature point position of the first image is acquired asThe error range is set by taking the center of the point as R: />When limiting distance->When the characteristic points of the die castings meet the requirements, reminding is not carried out; when limiting distance +>When the die castings are abnormal, relevant technicians are reminded, the situation that the integral errors caused by image comparison of adjacent die castings are gradually accumulated and become large is avoided, and the accuracy of data analysis is improved.
Further, in step S4, when an abnormal die casting occurs according to the analysis result, the die casting is classified, the abnormal die casting is displayed to the related technician through the display device, and meanwhile, the voice is used for reminding the related technician, so that the related technician can conveniently and rapidly overhaul the abnormal die casting, and the working efficiency of the related technician is improved.
Compared with the prior art, the invention has the following beneficial effects:
according to the invention, basic data information is acquired, the image information of the die castings is acquired through the image pickup equipment, the acquired image information is analyzed and processed, and the adjacent die casting images are subjected to comparison analysis, so that the time interval for image acquisition and comparison of the adjacent images is short, the speed is high, the images are not influenced by environmental factors, the operation of repeatedly calling the images is avoided, the monitoring can be performed in any scene, the effect of rapid analysis can be realized even if the number of the die castings is large, and the range of the die casting visual monitoring scene is enlarged. According to the analysis result, when abnormal die castings appear, the die castings are classified, relevant technicians are displayed and reminded by voice, the relevant technicians can quickly know corresponding basic information of the die castings and the specific problems, the visual monitoring efficiency is improved, the working efficiency of the relevant technicians is improved, and the system robustness is improved.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
FIG. 1 is a schematic diagram of the module composition of a visual die casting monitoring system based on big data according to the present invention;
FIG. 2 is a flow chart of the steps of a visual die casting monitoring method based on big data according to the present invention;
FIG. 3 is a schematic diagram of error analysis of a visual die casting monitoring method based on big data according to the present invention;
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1-2, the present invention provides the following technical solutions: a visual die casting monitoring system based on big data, the visual monitoring system comprising: the system comprises a data acquisition module, a database, a data analysis module and a user reminding module;
the data acquisition module is connected with the database, the database is connected with the data analysis module, and the data analysis module is connected with the user reminding module;
the data acquisition module is used for acquiring basic data information, the image information of the die castings is acquired through the camera equipment, the data acquisition module comprises a basic data input unit and an image acquisition unit, the basic data input unit is used for inputting basic data information of the die castings and the die castings, such as equipment numbers, production quantity, production order numbers and the like, and the image acquisition unit is used for acquiring images of the die castings through the camera equipment, such as installation of a high-definition camera or use of a radar and the like, so that the state of each die casting can be clearly known.
The data base is used for carrying out encryption storage on collected data and analysis results, the data base comprises a data storage unit, a data encryption unit and a data cleaning unit, the data storage unit is used for carrying out data storage on the collected data and the analysis results through SAN technology, the SAN is a storage area network, the data base is a high-speed special sub-network for connecting storage equipment such as a disk array, a tape and the like with related servers through connecting equipment such as an optical fiber hub, an optical fiber router and an optical fiber switch, the data base is used for providing higher throughput capacity by means of connection between the optical fiber channel and the storage equipment, supporting farther distance and more reliable connection, the SAN can be a switched network or a shared network, the data encryption unit is used for carrying out data encryption on the whole process through a homomorphic encryption algorithm, so that the safety of the data is guaranteed, the leakage of produced die casting information is prevented, the homomorphic encryption algorithm is used for meeting the homomorphic operation property of ciphertext, namely, after the data is subjected to homomorphic encryption, the specific calculation is carried out on ciphertext, the obtained plaintext after corresponding homomorphic decryption is equivalent to the same as the plaintext, the same as the data is directly carried out the same calculation, the homomorphic encryption algorithm is realized, and if the homomorphic encryption algorithm is different from any homomorphic encryption algorithm is realized; if the ciphertext is supported to be calculated in a partial form, for example, only addition, only multiplication or limited times of addition and multiplication are supported, the ciphertext is called semi-homomorphic encryption or partial homomorphic encryption, and the data cleaning unit is used for automatically cleaning the die casting data information meeting the requirements after the die casting production is completed, so that redundant data are cleaned, the storage space is saved, and the cost of data storage is reduced.
The data analysis module is used for analyzing and processing collected data, the data analysis module comprises an image preprocessing unit and an error analysis unit, the image preprocessing unit is used for preprocessing and analyzing images according to the collected image data, extracting outlines of pressure castings in the images, comparing adjacent pressure casting images, and because the number of the pressure castings needing to be detected is extremely large, standard data images do not need to be repeatedly adjusted through the comparison of the adjacent images, the complexity of the image comparison process is reduced, the realization mode is simple and quick, the stability of the system is improved, the data information of each pressure casting can be quickly known, meanwhile, the environment change of the adjacent images is small, the generated interference is small, the images are convenient to extract, the error is small, the error analysis unit is used for analyzing and processing errors between the identified images, monitoring the quality of the pressure castings according to the continuous change of the errors, and avoiding the image error accumulation of the adjacent pressure castings, and the integral error is large.
The user reminding module is used for classifying when abnormal die castings appear according to the analysis result, and reminding the user, and comprises a die casting classification unit, a screen display unit and a voice reminding unit, wherein the die casting classification unit is used for separating abnormal die castings according to the analysis result, facilitating maintenance of related technicians, improving the accuracy of die casting finished products, the screen display unit is used for displaying abnormal die casting conditions for the related technicians through the screen display device, and the voice reminding unit is used for reminding the related technicians through voice, so that the related technicians can check and maintain the die castings in time.
A visual monitoring method for die casting based on big data comprises the following steps:
s1, inputting basic data information of die casting equipment and die castings, and acquiring images of the die castings through camera equipment;
s2, preprocessing and analyzing the image according to the acquired image information, extracting the outline of the die casting in the image, and comparing adjacent die casting images;
in the step S2, extracting die casting contours in the images according to the acquired die casting image information, and comparing the adjacent images;
s201, performing binarization processing on the acquired image;
s202, performing two-dimensional Gaussian filtering on the image;
placing the image in a coordinate system which can be set by the related technicians, wherein the image isWhere x and y represent pixel values of the image, a two-dimensional gaussian filter function f is calculated by the following formula:
wherein,,expressed as the variance of the gaussian function, two one-dimensional rank filters are obtained by decomposition:
s203, eliminating noise interference of the image through convolution operation;
two one-dimensional row-column filters are respectively connected with the imagePerforming convolution calculation to obtain output:
wherein,,represented as convolution operator, implements low-pass filtering to eliminate noise interference;
s204, extracting edge points of the die casting in the image according to the image gradient, and connecting to form the outline of the die casting;
for gradient amplitudePerforming non-maximum suppression when gradient amplitude of pixel point on image +.>If the gradient amplitude value is smaller than the gradient amplitude value of two adjacent pixel points along the gradient line direction, judging the pixel point as a non-edge point, and adding the gradient amplitude value +.>Setting to 0 and vice versa to 1, non-maximum suppression, as the name implies, is suppression of elements that are not maxima, which can be understood as local maximum searching. The local representation is a neighborhood, the neighborhood has two variable parameters, one is the dimension of the neighborhood, and the other is the size of the neighborhood, and the essence is to search local maxima and inhibit non-maxima elements; performing double thresholding separation on non-maximum suppression images, and setting a threshold valueAnd->Wherein->By a high threshold->Dividing to obtain image->From low threshold->Dividing to obtain image->In the form of imagesBased on, image->In order to supplement edge connection, extracting the outline of the die casting, processing the image by using a high threshold value and a low threshold value, and reserving more detail information in the image so that the image comparison result is more accurate;
first, theContour image data extracted from the individual die castings form a vector set +.>Contour image data extracted from the nth die casting form a vector set +.>Image contrast index +.>And (3) performing calculation:
wherein N is expressed as the number of die castings, and a threshold value is setWhen->When the comparison of the images is similar, the error meets the standard, the die casting meets the requirement, and when +.>And when the comparison error of the image is large, the image indicates that the die casting does not meet the requirement, the mark is an abnormal die casting, and the step S4 is directly executed to remind relevant technicians of checking and maintaining.
S3, carrying out corresponding matching on the characteristic points according to the preprocessed image, and carrying out error analysis;
in the step S3, according to the image of the die casting extraction outline, when the image outline of the die casting meets the requirement, the characteristic points of the comparison image are correspondingly matched, and the error is analyzed;
the extracted contour image data forms a setWherein n is die castingNumber of pieces, the firstThe variance of the profile image extracted from the individual die castings is +.>First->The variance of the profile image extracted from the individual die castings is +.>K is expressed as the scale ratio between two adjacent images, then +.>Convolution of the image of the individual die-cast parts with the Gaussian kernel L and +.>Convolution of the image of the individual die cast parts with the Gaussian kernel>The method comprises the following steps:
the scale index E is calculated by the following formula:
taylor expansion is carried out on the two-dimensional Gaussian filter function f through the following formula to obtain an expanded function:
Wherein,,values denoted as extreme points>For scale space coordinates, T is denoted transpose, when located at the extreme point +.>At the time of Taylor expansion function->Setting the threshold value as +.>When->When the point contrast is low, deleting from the candidate points;
converting the image into a real symmetric matrix Z, and screening the characteristic points through the following formula:;
wherein,,expressed as the sum of matrix eigenvalues, +.>Represented as determinant of matrix>Expressed as the ratio of the maximum characteristic value to the minimum characteristic value, when the characteristic points meet the inequality, the characteristic points meeting the requirements are the corresponding characteristic points, and the gradient amplitude is combined +.>And gradient line direction->Realizing the corresponding matching of the feature points of the adjacent contrast images;
after the first die casting can be confirmed by relevant technicians whether the first die casting meets the requirements, the acquired image provides a basis for subsequent visual monitoring, and the same characteristic point change vector is obtained according to the corresponding matching of the characteristic points of the adjacent images to form a setWherein n is the number of die castings, and the included angle set is +.>In the first placeAt each die casting, the characteristic point change vector is +.>The included angle is->At the n-1 th die casting, the characteristic point variation vector is +.>The included angle is->In->At each die casting, the characteristic point change vector is +.>The limiting distance of the nth-1 die casting is +.>And (3) performing calculation:
then associate the included angleThe related included angle is the connection line of the n-1 th feature point and the first feature point and the n-1 th feature point and the +.>Included angles of the connecting lines of the characteristic points;
the following formula is used for the firstLimiting distance of individual die castings->And (3) performing calculation:
the feature point position of the first image is acquired asThe error range is set by taking the center of the point as R: />When limiting distance->When the characteristic points of the die castings meet the requirements, reminding is not carried out; when limiting distance +>When the die castings are abnormal, relevant technicians are reminded, the situation that the integral errors caused by image comparison of adjacent die castings are gradually accumulated and become large is avoided, and the accuracy of data analysis is improved. FIG. 3 is a schematic diagram of error analysis.
S4, classifying when abnormal die castings appear according to analysis results, and reminding related technicians through display equipment and voice.
In step S4, when an abnormal die casting occurs according to the analysis result, the die casting is classified, and the abnormal die casting condition, such as the die casting number, the abnormality reason, the production requirement and the like, is displayed to the related technician through the display device, such as a computer or a mobile phone and the like, and simultaneously, the related technician is reminded through the voice, so that the related technician can conveniently and rapidly overhaul the abnormal die casting, and the working efficiency of the related technician is improved.
Embodiment one:
if the image contrast index of a certain die casting and the next die casting isSetting threshold +.>At this time->When the images are similar in comparison, the error accords with the standard, the die casting accords with the requirement, and reminding is not carried out; if the image contrast index of a certain die casting and the next die casting is +.>At this time->The fact that the image contrast error is large indicates that the die castings do not meet the requirements, the marks are abnormal die castings, and relevant technicians are reminded to carry out inspection and maintenance;
corresponding to the characteristic points of the die casting, ifAt each die casting, the characteristic point change vector is +.>,/>The included angle is->At the n-1 th die casting, the characteristic point variation vector is +.>,/>An included angle ofIn->At each die casting, the characteristic point change vector is +.>,/>Radius>The limiting distance of the n-1 th die casting is +.>,The n-1 th die casting meets the requirements, the +.>Limiting distance of individual die castingsDescription of the->And if the die castings do not meet the requirements, the die castings are abnormal at the characteristic points, and relevant technicians are reminded.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Finally, it should be noted that: the foregoing description is only a preferred embodiment of the present invention, and the present invention is not limited thereto, but it is to be understood that modifications and equivalents of some of the technical features described in the foregoing embodiments may be made by those skilled in the art, although the present invention has been described in detail with reference to the foregoing embodiments. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. A visual die-casting monitoring method based on big data is characterized in that: comprises the following steps:
s1, inputting basic data information of die casting equipment and die castings, and acquiring images of the die castings through camera equipment;
s2, preprocessing and analyzing the image according to the acquired image information, extracting the outline of the die casting in the image, and comparing adjacent die casting images;
s3, carrying out corresponding matching on the characteristic points according to the preprocessed image, and carrying out error analysis;
s4, classifying when abnormal die castings appear according to analysis results, and reminding related technicians through display equipment and voice;
in the step S2, extracting die casting contours in the images according to the acquired die casting image information, and comparing the adjacent images;
s201, performing binarization processing on the acquired image;
s202, performing two-dimensional Gaussian filtering on the image;
s203, eliminating noise interference of the image through convolution operation;
s204, extracting edge points of the die casting in the image according to the image gradient, and connecting to form the outline of the die casting;
the extracted contour image data forms a setWherein n is the number of die castings, the +.>Contour image data extracted from the individual die castings form a vector set +.>Contour image data extracted from the nth die casting form a vector set +.>Image contrast index +.>And (3) performing calculation:
wherein N is expressed as the number of die castings, and a threshold value is setWhen->If the ratio is equal, the die casting meets the requirement, if +.>When the die casting does not meet the requirement, the mark is abnormal die casting, and the die casting is directly carried outStep S4, reminding related technicians of checking and maintaining;
in the step S3, according to the image of the die casting extraction outline, when the image outline of the die casting meets the requirement, the characteristic points of the comparison image are correspondingly matched, and the error is analyzed;
according to the corresponding matching of the adjacent image feature points, the same feature point change vector is obtained to form a setWherein n is the number of die castings, and the included angle set is +.>In the first placeAt each die casting, the characteristic point change vector is +.>The included angle is->At the n-1 th die casting, the characteristic point variation vector is +.>The included angle is->In->At each die casting, the characteristic point change vector is +.>The limiting distance of the nth-1 die casting is +.>And (3) performing calculation:
then associate the included angleThe related included angle is the connection line of the n-1 th feature point and the first feature point and the n-1 th feature point and the +.>Included angles of the connecting lines of the characteristic points;
the following formula is used for the firstLimiting distance of individual die castings->And (3) performing calculation:
the feature point position of the first image is acquired asTaking the point as the center of a circle, the radius R sets the error range as follows:when limiting distance->When the characteristic points of the die castings meet the requirements, reminding is not carried out; when limiting distance +>And when the die casting is abnormal, reminding related technicians.
2. The visual die casting monitoring method based on big data according to claim 1, wherein the method comprises the following steps: placing an image in a coordinate system, wherein the image isWherein x and y represent pixel values of the image; first->The variance of the profile image extracted from the individual die castings is +.>First->The variance of the profile image extracted from the individual die castings is +.>K is expressed as the scale ratio between two adjacent images, +.>And->Expressed as a corresponding two-dimensional Gaussian filter function, then +.>Convolution of the image of the individual die-cast parts with the Gaussian kernel L and +.>Convolution of the image of the individual die cast parts with the Gaussian kernel>The method comprises the following steps:
the scale index E is calculated by the following formula:
taylor expansion is carried out on the two-dimensional Gaussian filter function f through the following formula to obtain an expanded function:
Wherein,,values denoted as extreme points>For scale space coordinates, T is denoted transpose, when located at the extreme point +.>At the time of Taylor expansion function->Setting the threshold value as +.>When->When the point contrast is low, deleting from the candidate points;
converting the image into a real symmetric matrix Z, and screening the characteristic points through the following formula:
wherein,,expressed as the sum of matrix eigenvalues, +.>Represented as determinant of matrix>Expressed as the ratio of the maximum eigenvalue to the minimum eigenvalue, and when the eigenvalue satisfies the inequality, the eigenvalue is the satisfactory one, combined with the gradient amplitudeAnd gradient line direction->And realizing the corresponding matching of the feature points of the adjacent contrast images.
3. The visual die casting monitoring method based on big data according to claim 2, wherein: in step S4, when an abnormal die casting occurs according to the analysis result, the die casting is classified, and the abnormal die casting condition is displayed to the related technician through the display device, and simultaneously, the abnormal die casting is reminded through voice.
4. A big data based die casting visual monitoring system for implementing the big data based die casting visual monitoring method of any of claims 1-3, characterized in that: the visual monitoring system comprises: the system comprises a data acquisition module, a database, a data analysis module and a user reminding module;
the data acquisition module is connected with the database, the database is connected with the data analysis module, and the data analysis module is connected with the user reminding module; the data acquisition module is used for acquiring basic data information, acquiring die casting image information through the camera equipment, the database is used for carrying out encryption storage on acquired data and analysis results, the data analysis module is used for carrying out analysis processing on the acquired data, and the user reminding module is used for classifying when abnormal die castings appear according to the analysis results and reminding a user.
5. The visual die casting monitoring system based on big data according to claim 4, wherein: the data acquisition module comprises a basic data input unit and an image acquisition unit, wherein the basic data input unit is used for inputting basic data information of die casting equipment and die castings, and the image acquisition unit is used for carrying out image acquisition on the die castings through the camera equipment.
6. The visual die casting monitoring system based on big data according to claim 5, wherein: the database comprises a data storage unit, a data encryption unit and a data cleaning unit, wherein the data storage unit is used for storing collected data and analysis results through a SAN technology, the data encryption unit is used for encrypting data in the whole process through a homomorphic encryption algorithm, and the data cleaning unit is used for automatically cleaning die casting data information meeting requirements after die casting production is completed.
7. The visual die casting monitoring system based on big data according to claim 6, wherein: the data analysis module comprises an image preprocessing unit and an error analysis unit, wherein the image preprocessing unit is used for preprocessing and analyzing images according to collected image data, extracting outlines of die castings in the images, comparing adjacent die casting images, and the error analysis unit is used for analyzing and processing errors among the identified images and monitoring the quality of the die castings according to error continuity changes.
8. The visual die casting monitoring system based on big data according to claim 7, wherein: the user reminding module comprises a die casting classification unit, a screen display unit and a voice reminding unit, wherein the die casting classification unit is used for separating abnormal die castings according to analysis results, the screen display unit is used for displaying abnormal die casting conditions for related technicians through screen display equipment, and the voice reminding unit reminds the related technicians through voice.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310451150.3A CN116168349B (en) | 2023-04-25 | 2023-04-25 | Big data-based die-casting visual monitoring system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310451150.3A CN116168349B (en) | 2023-04-25 | 2023-04-25 | Big data-based die-casting visual monitoring system and method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116168349A CN116168349A (en) | 2023-05-26 |
CN116168349B true CN116168349B (en) | 2023-06-30 |
Family
ID=86420445
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310451150.3A Active CN116168349B (en) | 2023-04-25 | 2023-04-25 | Big data-based die-casting visual monitoring system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116168349B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116735592A (en) * | 2023-06-14 | 2023-09-12 | 大连铸云智能制造有限公司 | Die casting detection method and system |
CN117131530B (en) * | 2023-10-20 | 2024-01-30 | 合肥亚明汽车部件有限公司 | Intelligent factory sensitive data encryption protection method |
CN117753942B (en) * | 2023-12-06 | 2024-07-02 | 广州市型腔模具制造有限公司 | Temperature control system and method for integrated die-casting die |
CN118129670A (en) * | 2024-05-07 | 2024-06-04 | 江苏中科云控智能工业装备有限公司 | Automatic calibration system and method for integrated die-casting frame based on artificial intelligence |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6776212B1 (en) * | 2002-06-21 | 2004-08-17 | L&P Property Management Company | Die casting process incorporating computerized pattern recognition techniques |
CN103624234B (en) * | 2013-10-25 | 2016-09-28 | 北京航科发动机控制系统科技有限公司 | Centrifugal sensing die casting machine work process automatic monitoring method based on image procossing |
CN113310992A (en) * | 2021-05-28 | 2021-08-27 | 中山市满鑫科技有限公司 | Die-casting finished product detection system based on image recognition technology |
-
2023
- 2023-04-25 CN CN202310451150.3A patent/CN116168349B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN116168349A (en) | 2023-05-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116168349B (en) | Big data-based die-casting visual monitoring system and method | |
US11093790B2 (en) | Distance statistics based method for 3D sonar point cloud image enhancement | |
WO2022088620A1 (en) | State detection method and apparatus for camera lens, device and storage medium | |
CN108154149B (en) | License plate recognition method based on deep learning network sharing | |
CN110956092B (en) | Intelligent metallographic detection rating method and system based on deep learning | |
CN116777907A (en) | Sheet metal part quality detection method | |
CN108257125B (en) | Depth image quality non-reference evaluation method based on natural scene statistics | |
CN113554649B (en) | Defect detection method and device, computer equipment and storage medium | |
CN110956180B (en) | Detection method and system of counterweight weight, acquisition method and system and crane | |
CN112215079B (en) | Global multistage target tracking method | |
CN109101976A (en) | A kind of detection method of arc extinguishing grid pieces surface defect | |
Li et al. | A two-stage multiscale residual attention network for light guide plate defect detection | |
CN108288274A (en) | Mold detection method, device and electronic equipment | |
CN116050678A (en) | Die-casting product processing test system and method based on cloud computing | |
CN117392133B (en) | Die casting burr detection system and method based on machine vision | |
CN117372966A (en) | Turntable state monitoring method based on video monitoring | |
CN117495891A (en) | Point cloud edge detection method and device and electronic equipment | |
CN110738625B (en) | Image resampling method, device, terminal and computer readable storage medium | |
CN113240579A (en) | Intelligent industrial product defect detection method and device and computer storage medium thereof | |
CN111524097B (en) | Method and device for detecting horn mouth of two devices, storage medium and equipment | |
CN117315723B (en) | Digital management method and system for mold workshop based on artificial intelligence | |
CN117911792B (en) | Pin detecting system for voltage reference source chip production | |
CN117152447B (en) | Intelligent management method and system for punching die | |
CN116665138B (en) | Visual detection method and system for stamping processing of automobile parts | |
CN111950493B (en) | Image recognition method, device, terminal equipment and readable storage medium |
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 |