CN116030030A - Integrated assessment method for internal and external defects of weld joint of prefabricated part - Google Patents
Integrated assessment method for internal and external defects of weld joint of prefabricated part Download PDFInfo
- Publication number
- CN116030030A CN116030030A CN202310103804.3A CN202310103804A CN116030030A CN 116030030 A CN116030030 A CN 116030030A CN 202310103804 A CN202310103804 A CN 202310103804A CN 116030030 A CN116030030 A CN 116030030A
- Authority
- CN
- China
- Prior art keywords
- defect
- determining
- horizontal
- block
- feature vector
- 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
-
- 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
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
- Investigating And Analyzing Materials By Characteristic Methods (AREA)
Abstract
The invention discloses an integrated assessment method for internal and external defects of a welding line of a prefabricated part. The method solves the problem that in the prior art, defects inside and outside the welding line are independently analyzed, and an evaluation result of the whole defect of the welding line is difficult to accurately obtain.
Description
Technical Field
The invention relates to the field of image processing, in particular to an integrated assessment method for internal and external defects of a welding line of a prefabricated part.
Background
The prefabricated parts are usually made of a plurality of metal parts by welding, welding joints are formed at welding positions, the quality of the welding joints directly influences the quality of the prefabricated parts, and therefore detection of the welding joints is often needed after the welding is finished. The existing weld defect detection method is generally to analyze defects inside and outside the weld separately to obtain defect detection results inside and outside the weld respectively. However, defects inside and outside the weld are associated, for example, gas may not escape during the welding process, pores may form inside the weld and voids may form outside the weld. Therefore, defects inside and outside the welding line are independently analyzed, and the defect evaluation result of the whole welding line is difficult to accurately obtain.
Accordingly, there is a need for improvement and development in the art.
Disclosure of Invention
The invention aims to solve the technical problems that aiming at the defects in the prior art, an integrated assessment method for internal and external defects of a welding line of a prefabricated part is provided, and aims to solve the problems that in the prior art, the defects inside and outside the welding line are independently analyzed, and the assessment result of the integral defects of the welding line is difficult to accurately obtain.
The technical scheme adopted by the invention for solving the problems is as follows:
in a first aspect, an embodiment of the present invention provides a method for integrally evaluating internal and external defects of a weld joint of a prefabricated member, where the method includes:
acquiring a plurality of defect areas inside and outside a welding line of a prefabricated part and defect types corresponding to the defect areas respectively, wherein the defect areas inside and outside the welding line are obtained based on different identification methods respectively;
determining a defect distribution map of the welding seam according to each defect area, wherein the defect distribution map comprises a plurality of defect blocks, and each defect block is used for reflecting the shape and the position of different defect areas;
obtaining neighborhood defect information corresponding to each defective block according to the defect distribution diagram, and determining a first weight value corresponding to each defective block according to the neighborhood defect information corresponding to each defective block;
determining second weight values corresponding to the defect blocks according to the defect types corresponding to the defect areas respectively;
determining a weight defect distribution map according to the first weight value and the second weight value which correspond to each defect block respectively;
and inputting the weight defect distribution map into a defect prediction model to obtain the defect grade corresponding to the welding line.
In one embodiment, the obtaining a plurality of defect areas inside and outside the weld joint of the prefabricated component and defect types corresponding to the defect areas respectively includes:
acquiring a point cloud image outside the welding line, and carrying out feature extraction on the point cloud image to obtain a feature vector corresponding to the point cloud image;
determining a plurality of defect areas outside the welding line and the defect types corresponding to the defect areas respectively according to the feature vectors;
and obtaining a ground penetrating radar scanning image of the inside of the welding seam, and determining a plurality of defect areas of the inside of the welding seam and the defect types corresponding to the defect areas respectively according to the ground penetrating radar scanning image.
In one embodiment, the feature extraction of the point cloud image to obtain a feature vector corresponding to the point cloud image includes:
image segmentation is carried out on the point cloud image to obtain a plurality of horizontal domains and a plurality of vertical domains, wherein each horizontal domain respectively comprises point clouds at different horizontal positions, and each vertical domain respectively comprises point clouds at different vertical positions;
acquiring horizontal feature vectors corresponding to the horizontal domains and vertical feature vectors corresponding to the vertical domains respectively;
determining a plurality of intersecting feature vectors according to each horizontal feature vector and each vertical feature vector, wherein each intersecting feature vector is determined based on the horizontal feature vector and the vertical feature vector in different combinations;
extracting features of the point cloud images to obtain integral feature vectors corresponding to the point cloud images;
and determining the feature vector according to each horizontal feature vector, each vertical feature vector, each crossed feature vector and the whole feature vector.
In one embodiment, the obtaining the horizontal feature vector corresponding to each horizontal domain and the vertical feature vector corresponding to each vertical domain includes:
determining a plurality of first fitting curves according to the horizontal domains, and respectively extracting features of the first fitting curves to obtain horizontal feature vectors, wherein the first fitting curves are respectively determined based on point clouds of different horizontal domains, and each first fitting curve is used for reflecting the corresponding relation between the horizontal coordinates and depth values of the point clouds in one horizontal domain;
and determining a plurality of second fitting curves according to the vertical domains, and respectively carrying out feature extraction on the second fitting curves to obtain vertical feature vectors, wherein the second fitting curves are respectively determined based on point clouds of different vertical domains, and each second fitting curve is used for reflecting the corresponding relation between the ordinate of the point cloud in one vertical domain and the depth value.
In one embodiment, the obtaining, according to the defect distribution map, neighborhood defect information corresponding to each of the defect blocks includes:
determining a detection radius corresponding to each defective block according to the area of the defective block;
determining a circle center according to the center point of the defect block, and determining a detection circle corresponding to the defect block according to the circle center and the detection radius;
determining a defect area except for the defect block contained in the detection ring of the defect block according to the detection ring and the defect distribution diagram;
and determining the neighborhood defect information of the defect block according to the defect area.
In one embodiment, the determining the first weight value corresponding to each defective area according to the neighborhood defect information corresponding to each defective block includes:
sorting the defective blocks according to the neighborhood defect information corresponding to the defective blocks respectively;
and determining the first weight value corresponding to each defect area according to the sorting result.
In one embodiment, the determining a weighted defect distribution map according to the first weight value and the second weight value corresponding to each defect block includes:
determining target weight values corresponding to the defect blocks according to weighted average values of the first weight values and the second weight values corresponding to the defect blocks respectively;
and determining the weight defect distribution map according to the target weight values respectively corresponding to the defect blocks.
In a second aspect, an embodiment of the present invention further provides an integrated apparatus for evaluating defects inside and outside a weld joint of a prefabricated member, where the apparatus includes:
the image acquisition module is used for acquiring a plurality of defect areas inside and outside a welding line of the prefabricated component and defect types corresponding to the defect areas respectively, wherein the defect areas inside and outside the welding line are obtained based on different identification methods respectively;
determining a defect distribution map of the welding seam according to each defect area, wherein the defect distribution map comprises a plurality of defect blocks, and each defect block is used for reflecting the shape and the position of different defect areas;
the image processing module is used for acquiring neighborhood defect information corresponding to each defect block according to the defect distribution diagram, and determining a first weight value corresponding to each defect block according to the neighborhood defect information corresponding to each defect block;
determining second weight values corresponding to the defect blocks according to the defect types corresponding to the defect areas respectively;
determining a weight defect distribution map according to the first weight value and the second weight value which correspond to each defect block respectively;
and the image analysis module is used for inputting the weight defect distribution diagram into a defect prediction model to obtain the defect grade corresponding to the welding line.
In a third aspect, an embodiment of the present invention further provides a terminal, where the terminal includes a memory and one or more processors; the memory stores more than one program; the program includes instructions for performing the weld inside-outside defect integrated evaluation method of the prefabricated member as described in any one of the above; the processor is configured to execute the program.
In a fourth aspect, an embodiment of the present invention further provides a computer readable storage medium, on which a plurality of instructions are stored, where the instructions are adapted to be loaded and executed by a processor, so as to implement the steps of the method for evaluating the weld internal and external defects of any one of the prefabricated components.
The invention has the beneficial effects that: according to the embodiment of the invention, the weight defect distribution map for reflecting the overall defect condition of the welding seam is obtained by combining the defects inside and outside the welding seam, and the defect grade of the welding seam can be accurately estimated through the weight defect distribution map. The method solves the problem that in the prior art, defects inside and outside the welding line are independently analyzed, and an evaluation result of the whole defect of the welding line is difficult to accurately obtain.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings may be obtained according to the drawings without inventive effort to those skilled in the art.
FIG. 1 is a schematic flow chart of an integrated assessment method for internal and external defects of a weld joint of a prefabricated part, which is provided by the embodiment of the invention.
FIG. 2 is a schematic view of an internal module of an integrated assessment device for internal and external defects of a weld joint of a prefabricated member according to an embodiment of the present invention.
Fig. 3 is a schematic block diagram of a terminal according to an embodiment of the present invention.
Detailed Description
The invention discloses an integrated assessment method for internal and external defects of a welding line of a prefabricated part, which aims to make the purposes, the technical scheme and the effects of the invention clearer and more definite, and further details the invention by referring to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless expressly stated otherwise, as understood by those skilled in the art. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. The term "and/or" as used herein includes all or any element and all combination of one or more of the associated listed items.
It will be understood by those skilled in the art that all terms (including 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 unless defined otherwise. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
Aiming at the defects in the prior art, the invention provides an integrated assessment method for internal and external defects of a welding line of a prefabricated part, which comprises the steps of obtaining a plurality of defect areas of the internal and external parts of the welding line of the prefabricated part and defect types corresponding to the defect areas respectively, wherein the defect areas of the internal and external parts of the welding line are obtained based on different identification methods respectively; determining a defect distribution map of the welding seam according to each defect area, wherein the defect distribution map comprises a plurality of defect blocks, and each defect block is used for reflecting the shape and the position of different defect areas; obtaining neighborhood defect information corresponding to each defective block according to the defect distribution diagram, and determining a first weight value corresponding to each defective block according to the neighborhood defect information corresponding to each defective block; determining second weight values corresponding to the defect blocks according to the defect types corresponding to the defect areas respectively; determining a weight defect distribution map according to the first weight value and the second weight value which correspond to each defect block respectively; and inputting the weight defect distribution map into a defect prediction model to obtain the defect grade corresponding to the welding line. According to the invention, the weight defect distribution map for reflecting the overall defect condition of the welding seam is obtained by combining the defects inside and outside the welding seam, and the defect grade of the welding seam can be accurately estimated through the weight defect distribution map. The method solves the problem that in the prior art, defects inside and outside the welding line are independently analyzed, and an evaluation result of the whole defect of the welding line is difficult to accurately obtain.
As shown in fig. 1, the method includes:
step S100, obtaining a plurality of defect areas inside and outside a welding line of the prefabricated component and defect types corresponding to the defect areas respectively, wherein the defect areas inside and outside the welding line are obtained based on different identification methods respectively.
Specifically, since the outside of the weld is visible and the inside is invisible, different identification methods are required to be adopted for the outside and the inside of the weld to acquire a plurality of defect areas and defect types of the defect areas respectively corresponding to the outside and the inside of the weld.
In one implementation, step S100 specifically includes:
step S101, obtaining a point cloud image outside the welding seam, and carrying out feature extraction on the point cloud image to obtain a feature vector corresponding to the point cloud image;
step S102, determining a plurality of defect areas outside the welding line and the defect types corresponding to the defect areas respectively according to the feature vectors;
step S103, obtaining a ground penetrating radar scanning image of the inside of the welding seam, and determining a plurality of defect areas of the inside of the welding seam and the defect types corresponding to the defect areas respectively according to the ground penetrating radar scanning image.
Specifically, in the embodiment, a high-precision laser scanner (LiDAR) is installed on a portal frame vertically above a robot welding workstation in advance, and a point cloud image outside a welding line is acquired through a LiDAR technology, wherein the point cloud image comprises a large number of data points for reflecting the surface characteristics of the welding line. And then extracting the characteristics of the point cloud image to obtain characteristic vectors, and judging which areas outside the welding line are defect areas according to the characteristic vectors and respectively corresponding defect types of the defect areas. The types of defects outside the weld include: the excess height dimensions are undesirable, flash, undercut, crater, arc burn, surface porosity, surface cracking, weld distortion, warpage, and the like. In addition, the portal frame is also provided with a Ground Penetrating Radar (GPR), and the GPR is a method for detecting the characteristics and distribution rule of substances in a medium by utilizing an antenna to emit and receive high-frequency electromagnetic waves. Therefore, for the invisible inside of the welding seam, the embodiment obtains a ground penetrating radar scanning image of the inside of the welding seam through the GPR, analyzes which areas of the inside of the welding seam are defect areas according to the ground penetrating radar scanning image, and determines the defect types corresponding to the defect areas respectively. The types of defects inside the weld include: internal cracks, lack of penetration, lack of fusion, slag inclusion, air holes, and the like.
In one implementation, the step S101 specifically includes:
step S1011, performing image segmentation on the point cloud image to obtain a plurality of horizontal domains and a plurality of vertical domains, wherein each horizontal domain respectively comprises point clouds at different horizontal positions, and each vertical domain respectively comprises point clouds at different vertical positions;
step S1012, obtaining horizontal feature vectors corresponding to the horizontal domains and vertical feature vectors corresponding to the vertical domains respectively;
step S1013, determining a plurality of intersecting feature vectors according to each of the horizontal feature vectors and each of the vertical feature vectors, wherein each of the intersecting feature vectors is determined based on the horizontal feature vectors and the vertical feature vectors of different combinations, respectively;
step S1014, extracting features of the point cloud images to obtain integral feature vectors corresponding to the point cloud images;
step S1015, determining the feature vector according to each of the horizontal feature vector, each of the vertical feature vector, each of the intersecting feature vector, and the overall feature vector.
In order to accurately predict the defect area outside the weld, the embodiment needs to acquire the integral features and the multidimensional local features of the point cloud image. Specifically, for the whole, feature extraction is performed on the whole point cloud image to obtain a whole feature vector. For local, the point cloud image is divided into a plurality of horizontal domains according to different horizontal lines, and feature extraction is carried out on each horizontal domain to obtain a plurality of horizontal feature vectors. Meanwhile, the point cloud image is divided into a plurality of vertical domains according to different vertical lines, and feature extraction is carried out on each vertical domain respectively to obtain a plurality of vertical feature vectors. And then, carrying out feature intersection on each horizontal feature vector and each vertical feature vector to obtain a plurality of intersecting feature vectors. According to the embodiment, the data can be subjected to dimension reduction and calculation amount simplification by regional processing, and then the data characteristics with more dimensions are obtained by adopting characteristic crossing.
In one implementation, the step S1012 specifically includes:
step S10121, determining a plurality of first fitting curves according to each horizontal domain, and performing feature extraction on each first fitting curve to obtain each horizontal feature vector, wherein each first fitting curve is determined based on point clouds of different horizontal domains, and each first fitting curve is used for reflecting a corresponding relation between an abscissa of the point clouds in one horizontal domain and a depth value;
step S10122, determining a plurality of second fitting curves according to the vertical domains, and performing feature extraction on each second fitting curve to obtain each vertical feature vector, where each second fitting curve is determined based on point clouds of different vertical domains, and each second fitting curve is used to reflect a corresponding relationship between an ordinate of the point cloud and a depth value in one vertical domain.
In order to further reduce the calculation cost, the present embodiment adopts a curve fitting manner to obtain the characteristics of each horizontal domain and each vertical domain. Specifically, for each horizontal domain, the ordinate of each point cloud in the horizontal domain is equal, data points respectively corresponding to each point cloud are generated according to the abscissa and the depth value of each point cloud in the horizontal domain, the abscissa of each data point is the abscissa of each point cloud, the ordinate of each data point is the depth value of each point cloud, and then curve fitting is performed on all the data points corresponding to the horizontal domain to obtain a first fitting curve corresponding to the horizontal domain. And then extracting the characteristics of the first fitting curve to obtain a horizontal characteristic vector corresponding to the horizontal domain. For each vertical domain, the abscissa of each point cloud in the vertical domain is equal, data points corresponding to each point cloud are generated according to the ordinate and the depth value of each point cloud in the vertical domain, the abscissa of each data point is the ordinate of the point cloud, the ordinate of each data point is the depth value of the point cloud, and then curve fitting is carried out on all the data points corresponding to the vertical domain, so that a second fitting curve corresponding to the vertical domain is obtained. And then, carrying out feature extraction on the second fitting curve to obtain a vertical feature vector corresponding to the vertical domain. Since each horizontal domain/vertical domain includes a large amount of point cloud data, the calculation overhead in the feature extraction process can be greatly saved by the point-line conversion mode in the embodiment.
In one implementation, the horizontal domain of the present embodiment may be defined as a horizontal line, or may be defined as a horizontal region, such as y= [0,5].
In another implementation manner, the method for obtaining the overall feature vector includes:
mapping each first fitting curve and each second fitting curve into the same coordinate graph to obtain a comprehensive graph;
and extracting the characteristics of the comprehensive graph to obtain the integral characteristic vector.
In one implementation, the step S102 specifically includes:
inputting the feature vector into a pre-trained prediction model to obtain a plurality of defect areas outside the welding line and the defect types corresponding to the defect areas respectively.
In one implementation, the step S103 specifically includes:
inputting the ground penetrating radar scanning image into a 2D convolution network (2D-CNN) which is trained in advance to obtain image features corresponding to the ground penetrating radar scanning image;
inputting the image characteristics into a pre-trained residual neural network (ResNet) to obtain a plurality of defect areas in the weld joint and the defect types corresponding to the defect areas respectively.
In another implementation manner, the method for obtaining the defect types of the defect areas outside the welding seam and corresponding to the defect areas respectively includes:
inputting the point cloud image into a 3D convolution network (3D-CNN) to obtain the integral feature vector;
inputting the point cloud image into a KNN clustering network, carrying out point cloud clustering through the KNN clustering network to obtain point cloud clusters respectively corresponding to the point clouds, and determining local feature vectors respectively corresponding to the point clouds according to the point cloud clusters respectively corresponding to the point clouds;
inputting the whole feature vector and each local feature vector into a pre-trained target model, and generating semantic information corresponding to each point cloud respectively through the target model, wherein the semantic information of each point cloud is used for reflecting whether the point cloud has defects and corresponding defect types;
and determining a plurality of defect areas outside the welding line and the defect types corresponding to the defect areas according to the semantic information corresponding to each point cloud.
As shown in fig. 1, the method further includes:
step 200, determining a defect distribution diagram of the welding seam according to each defect area, wherein the defect distribution diagram comprises a plurality of defect blocks, and each defect block is used for reflecting the shape and the position of different defect areas.
Specifically, after obtaining defect areas inside and outside the weld joint, drawing defect distribution diagrams according to the positions and shapes of the defect areas, wherein each defect block in the defect distribution diagrams is used for referring to one defect area. It should be noted that, the bottom map of the defect distribution map may be a perspective view, including image information of the inside and the outside of the weld, and there may be an overlapping area between the defect blocks. In the embodiment, all defect areas inside and outside the weld joint are summarized through the defect distribution diagram, so that the follow-up integrated evaluation of the defects inside and outside the weld joint is facilitated.
As shown in fig. 1, the method further includes:
step S300, neighborhood defect information corresponding to each defective block is obtained according to the defect distribution diagram, and a first weight value corresponding to each defective block is determined according to the neighborhood defect information corresponding to each defective block.
Specifically, since the defect distribution map summarizes all the defect blocks inside and outside the welding line, the distance relation between the defect blocks can be obtained through the defect distribution map, and then the defect quantity contained in the neighborhood of each defect block is judged, so that the neighborhood defect information of each defect block is obtained. For each defective block, if the defects included in the neighborhood of the defective block are more, which means that the defective block is located in a defect concentration area, the defective block should be emphasized and analyzed in the subsequent defect level evaluation, and then a higher first weight value is allocated to the defective block.
In one implementation manner, the obtaining, according to the defect distribution map, the neighborhood defect information corresponding to each defect block includes:
step S301, determining a detection radius corresponding to each defective block according to the area of the defective block;
step S302, determining a circle center according to the center point of the defect block, and determining a detection circle corresponding to the defect block according to the circle center and the detection radius;
step S303, determining the defect area except the defect block contained in the detection ring of the defect block according to the detection ring and the defect distribution diagram;
step S304, determining the neighborhood defect information of the defect block according to the defect area.
Specifically, since the areas of the defect blocks may be different, and even a large difference may occur, the detection circle (i.e., the neighborhood) of each defect block in this embodiment is determined based on the area size of each defect block, and the smaller the area, the smaller the detection circle corresponding to the defect block, and the larger the area, the larger the detection circle corresponding to the defect block, so as to obtain more accurate neighborhood defect information of each defect block.
In one implementation manner, the determining, according to the neighborhood defect information corresponding to each defective block, a first weight value corresponding to each defective area, includes:
step S305, sorting the defective blocks according to the neighborhood defect information corresponding to the defective blocks;
step S306, determining the first weight value corresponding to each defect area according to the sorting result.
Specifically, each defective block is ordered according to neighborhood defect information of each defective block. If the sorting principle is that the neighborhood defect area is from large to small, distributing a first weight value of each defect area according to the sorting result to be sequentially decreased; if the sequencing result is that the neighborhood defect area is from small to large, the first weight value of each defect area is distributed according to the sequencing result and sequentially increased.
As shown in fig. 1, the method further includes:
step 400, determining second weight values corresponding to the defect blocks according to the defect types corresponding to the defect areas.
Since the different types of defects have different degrees of influence on the weld performance, the defect type of each defective region needs to be considered when performing defect evaluation of the weld. Specifically, a second weight value of each defective block is determined according to the defect type of each defective area. The higher the second weight value, the greater the impact of the defect type of the defect block on the weld performance; the lower the second weight value, the less the effect of the defect type of the defect block on the weld performance.
As shown in fig. 1, the method further includes:
step S500, determining a weight defect distribution diagram according to the first weight value and the second weight value respectively corresponding to each defect block.
Specifically, the present embodiment comprehensively determines the weight of each defective block based on the first weight value and the second weight value of each defective block. And then attaching a weight label to each defect block in the defect distribution map to obtain a weight defect distribution map. In the analysis of the weighted defect distribution map, the defect blocks with different weight values are paid attention to different degrees, and the higher the weight value, the higher the attention degree of the defect block, and conversely, the lower the attention degree.
In one implementation, the step S500 specifically includes:
step S501, determining target weight values corresponding to the defect blocks according to weighted average values of the first weight values and the second weight values corresponding to the defect blocks respectively;
step S502, determining the weighted defect distribution diagram according to the target weight values respectively corresponding to the defect blocks.
Specifically, the importance degrees corresponding to the defect type and the neighborhood defect information respectively may be preset, so as to obtain weights corresponding to the first weight value and the second weight value respectively. Then, for each defect block, multiplying the first weight value and the second weight value of the defect block by corresponding weights respectively, and then summing and averaging to obtain the target weight value of the defect block. And attaching a weight label to the defect distribution map based on the target weight value of each defect block, thus obtaining a weight defect distribution map.
As shown in fig. 1, the method further includes:
and S600, inputting the weight defect distribution diagram into a defect prediction model to obtain the defect grade corresponding to the welding line.
Specifically, this embodiment trains a defect prediction model in advance, the defect prediction model is built based on the attention mechanism, and the defect prediction model has learned the complex mapping relationship between input and output based on a large amount of training data in advance. The weight defect distribution map is input into the defect prediction model, parts with different weight values on the weight defect distribution map are stressed by different degrees, and finally the defect prediction model outputs the defect grade of the welding seam of the prefabricated part based on the weight defect distribution map. The system or the staff can judge whether the welding seam of the prefabricated part meets the standard according to the defect grade, so that unqualified prefabricated parts can be screened out in time.
In one implementation, the loss function of the defect prediction model is a cross entropy loss function, and the defect prediction model is back-propagated based on the cross entropy loss function value during training, and model parameters are updated in a gradient descent mode.
Based on the above embodiment, the present invention further provides an integrated assessment device for internal and external defects of a weld joint of a prefabricated member, as shown in fig. 2, the device includes:
the image acquisition module 01 is used for acquiring a plurality of defect areas inside and outside a welding line of the prefabricated component and defect types corresponding to the defect areas respectively, wherein the defect areas inside and outside the welding line are obtained based on different identification methods respectively;
determining a defect distribution map of the welding seam according to each defect area, wherein the defect distribution map comprises a plurality of defect blocks, and each defect block is used for reflecting the shape and the position of different defect areas;
the image processing module 02 is configured to obtain, according to the defect distribution map, neighborhood defect information corresponding to each of the defect blocks, and determine, according to the neighborhood defect information corresponding to each of the defect blocks, a first weight value corresponding to each of the defect blocks;
determining second weight values corresponding to the defect blocks according to the defect types corresponding to the defect areas respectively;
determining a weight defect distribution map according to the first weight value and the second weight value which correspond to each defect block respectively;
and the image analysis module 03 is used for inputting the weight defect distribution map into a defect prediction model to obtain the defect grade corresponding to the welding line.
Based on the above embodiment, the present invention also provides a terminal, and a functional block diagram thereof may be shown in fig. 3. The terminal comprises a processor, a memory, a network interface and a display screen which are connected through a system bus. Wherein the processor of the terminal is adapted to provide computing and control capabilities. The memory of the terminal includes a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface of the terminal is used for communicating with an external terminal through a network connection. The computer program, when executed by a processor, is configured to implement a method for integrated assessment of internal and external defects of a weld joint of a prefabricated component. The display screen of the terminal may be a liquid crystal display screen or an electronic ink display screen.
It will be appreciated by those skilled in the art that the functional block diagram shown in fig. 3 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the terminal to which the present inventive arrangements may be applied, and that a particular terminal may include more or less components than those shown, or may combine some of the components, or have a different arrangement of components.
In one implementation, the memory of the terminal has one or more programs stored therein and configured to be executed by one or more processors, the one or more programs including instructions for performing a weld inside and outside defect integrated assessment method of a prefabricated component.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
In summary, the invention discloses an integrated assessment method for internal and external defects of a weld joint of a prefabricated part, which comprises the steps of obtaining a plurality of defect areas of the internal and external parts of the weld joint of the prefabricated part and defect types corresponding to the defect areas respectively, wherein the defect areas of the internal and external parts of the weld joint are obtained based on different identification methods respectively; determining a defect distribution map of the welding seam according to each defect area, wherein the defect distribution map comprises a plurality of defect blocks, and each defect block is used for reflecting the shape and the position of different defect areas; obtaining neighborhood defect information corresponding to each defective block according to the defect distribution diagram, and determining a first weight value corresponding to each defective block according to the neighborhood defect information corresponding to each defective block; determining second weight values corresponding to the defect blocks according to the defect types corresponding to the defect areas respectively; determining a weight defect distribution map according to the first weight value and the second weight value which correspond to each defect block respectively; and inputting the weight defect distribution map into a defect prediction model to obtain the defect grade corresponding to the welding line. According to the invention, the weight defect distribution map for reflecting the overall defect condition of the welding seam is obtained by combining the defects inside and outside the welding seam, and the defect grade of the welding seam can be accurately estimated through the weight defect distribution map. The method solves the problem that in the prior art, defects inside and outside the welding line are independently analyzed, and an evaluation result of the whole defect of the welding line is difficult to accurately obtain.
It is to be understood that the invention is not limited in its application to the examples described above, but is capable of modification and variation in light of the above teachings by those skilled in the art, and that all such modifications and variations are intended to be included within the scope of the appended claims.
Claims (10)
1. An integrated assessment method for internal and external defects of a weld joint of a prefabricated component, which is characterized by comprising the following steps:
acquiring a plurality of defect areas inside and outside a welding line of a prefabricated part and defect types corresponding to the defect areas respectively, wherein the defect areas inside and outside the welding line are obtained based on different identification methods respectively;
determining a defect distribution map of the welding seam according to each defect area, wherein the defect distribution map comprises a plurality of defect blocks, and each defect block is used for reflecting the shape and the position of different defect areas;
obtaining neighborhood defect information corresponding to each defective block according to the defect distribution diagram, and determining a first weight value corresponding to each defective block according to the neighborhood defect information corresponding to each defective block;
determining second weight values corresponding to the defect blocks according to the defect types corresponding to the defect areas respectively;
determining a weight defect distribution map according to the first weight value and the second weight value which correspond to each defect block respectively;
and inputting the weight defect distribution map into a defect prediction model to obtain the defect grade corresponding to the welding line.
2. The method for integrally evaluating internal and external defects of a weld joint of a prefabricated member according to claim 1, wherein the acquiring a plurality of defect areas inside and outside the weld joint of the prefabricated member and defect types respectively corresponding to the defect areas comprises:
acquiring a point cloud image outside the welding line, and carrying out feature extraction on the point cloud image to obtain a feature vector corresponding to the point cloud image;
determining a plurality of defect areas outside the welding line and the defect types corresponding to the defect areas respectively according to the feature vectors;
and obtaining a ground penetrating radar scanning image of the inside of the welding seam, and determining a plurality of defect areas of the inside of the welding seam and the defect types corresponding to the defect areas respectively according to the ground penetrating radar scanning image.
3. The integrated assessment method for internal and external defects of a weld joint of a prefabricated member according to claim 2, wherein the feature extraction of the point cloud image to obtain a feature vector corresponding to the point cloud image comprises the following steps:
image segmentation is carried out on the point cloud image to obtain a plurality of horizontal domains and a plurality of vertical domains, wherein each horizontal domain respectively comprises point clouds at different horizontal positions, and each vertical domain respectively comprises point clouds at different vertical positions;
acquiring horizontal feature vectors corresponding to the horizontal domains and vertical feature vectors corresponding to the vertical domains respectively;
determining a plurality of intersecting feature vectors according to each horizontal feature vector and each vertical feature vector, wherein each intersecting feature vector is determined based on the horizontal feature vector and the vertical feature vector in different combinations;
extracting features of the point cloud images to obtain integral feature vectors corresponding to the point cloud images;
and determining the feature vector according to each horizontal feature vector, each vertical feature vector, each crossed feature vector and the whole feature vector.
4. The method for integrated assessment of internal and external defects of a weld joint of a prefabricated member according to claim 3, wherein the obtaining of the horizontal feature vector respectively corresponding to each horizontal domain and the vertical feature vector respectively corresponding to each vertical domain comprises:
determining a plurality of first fitting curves according to the horizontal domains, and respectively extracting features of the first fitting curves to obtain horizontal feature vectors, wherein the first fitting curves are respectively determined based on point clouds of different horizontal domains, and each first fitting curve is used for reflecting the corresponding relation between the horizontal coordinates and depth values of the point clouds in one horizontal domain;
and determining a plurality of second fitting curves according to the vertical domains, and respectively carrying out feature extraction on the second fitting curves to obtain vertical feature vectors, wherein the second fitting curves are respectively determined based on point clouds of different vertical domains, and each second fitting curve is used for reflecting the corresponding relation between the ordinate of the point cloud in one vertical domain and the depth value.
5. The method for integrated assessment of internal and external defects of a weld joint of a prefabricated member according to claim 1, wherein the obtaining neighborhood defect information corresponding to each defective block according to the defect distribution map comprises:
determining a detection radius corresponding to each defective block according to the area of the defective block;
determining a circle center according to the center point of the defect block, and determining a detection circle corresponding to the defect block according to the circle center and the detection radius;
determining a defect area except for the defect block contained in the detection ring of the defect block according to the detection ring and the defect distribution diagram;
and determining the neighborhood defect information of the defect block according to the defect area.
6. The method for integrated assessment of internal and external defects of a weld joint of a prefabricated member according to claim 5, wherein said determining a first weight value respectively corresponding to each defective region according to the neighborhood defect information respectively corresponding to each defective block comprises:
sorting the defective blocks according to the neighborhood defect information corresponding to the defective blocks respectively;
and determining the first weight value corresponding to each defect area according to the sorting result.
7. The method for integrated evaluation of internal and external defects of a weld joint of a prefabricated member according to claim 1, wherein said determining a weighted defect distribution map based on said first weight value and said second weight value respectively corresponding to each of said defective blocks comprises:
determining target weight values corresponding to the defect blocks according to weighted average values of the first weight values and the second weight values corresponding to the defect blocks respectively;
and determining the weight defect distribution map according to the target weight values respectively corresponding to the defect blocks.
8. An integrated assessment device for internal and external defects of a weld joint of a prefabricated component, which is characterized by comprising:
the image acquisition module is used for acquiring a plurality of defect areas inside and outside a welding line of the prefabricated component and defect types corresponding to the defect areas respectively, wherein the defect areas inside and outside the welding line are obtained based on different identification methods respectively;
determining a defect distribution map of the welding seam according to each defect area, wherein the defect distribution map comprises a plurality of defect blocks, and each defect block is used for reflecting the shape and the position of different defect areas;
the image processing module is used for acquiring neighborhood defect information corresponding to each defect block according to the defect distribution diagram, and determining a first weight value corresponding to each defect block according to the neighborhood defect information corresponding to each defect block;
determining second weight values corresponding to the defect blocks according to the defect types corresponding to the defect areas respectively;
determining a weight defect distribution map according to the first weight value and the second weight value which correspond to each defect block respectively;
and the image analysis module is used for inputting the weight defect distribution diagram into a defect prediction model to obtain the defect grade corresponding to the welding line.
9. A terminal comprising a memory and one or more processors; the memory stores more than one program; the program comprising instructions for performing the weld inside-outside defect integrated evaluation method of a prefabricated member according to any one of claims 1 to 7; the processor is configured to execute the program.
10. A computer readable storage medium having stored thereon a plurality of instructions adapted to be loaded and executed by a processor to carry out the steps of the method for integrated assessment of defects inside and outside a weld of a prefabricated component according to any one of the preceding claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310103804.3A CN116030030B (en) | 2023-02-13 | 2023-02-13 | Integrated assessment method for internal and external defects of weld joint of prefabricated part |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310103804.3A CN116030030B (en) | 2023-02-13 | 2023-02-13 | Integrated assessment method for internal and external defects of weld joint of prefabricated part |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116030030A true CN116030030A (en) | 2023-04-28 |
| CN116030030B CN116030030B (en) | 2023-08-29 |
Family
ID=86078070
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310103804.3A Active CN116030030B (en) | 2023-02-13 | 2023-02-13 | Integrated assessment method for internal and external defects of weld joint of prefabricated part |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116030030B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH1051693A (en) * | 1996-07-30 | 1998-02-20 | Toshiba Medical Eng Co Ltd | Defective picture element correction processing unit for image pickup element |
| US20060018534A1 (en) * | 2004-07-26 | 2006-01-26 | Dainippon Screen Mfg. Co., Ltd. | Technique for detecting a defect of an object by area segmentation of a color image of the object |
| JP2006308473A (en) * | 2005-04-28 | 2006-11-09 | Nippon Steel Corp | Method for inspecting periodic defect, and device |
| US20140175071A1 (en) * | 2011-06-29 | 2014-06-26 | Trumpf Laser - Und Systemtechnik Gmbh | Method for Detecting Defects During a Laser-Machining Process and Laser-Machining Device |
| US20140348415A1 (en) * | 2013-05-27 | 2014-11-27 | ThinkSmart IT Solutions Private Limited | System and method for identifying defects in welds by processing x-ray images |
| CN109342456A (en) * | 2018-09-14 | 2019-02-15 | 广东工业大学 | A kind of welding point defect detection method, device, equipment and readable storage medium storing program for executing |
| CN110866910A (en) * | 2019-11-13 | 2020-03-06 | 上海电气集团股份有限公司 | Weld quality prediction method, system, device and computer-readable storage medium |
| US20200117561A1 (en) * | 2018-10-11 | 2020-04-16 | International Business Machines Corporation | Using system errors and manufacturer defects in system components causing the system errors to determine a quality assessment value for the components |
| CN111932489A (en) * | 2020-06-03 | 2020-11-13 | 西安电子科技大学 | Weld defect detection method, system, storage medium, computer device and terminal |
| CN111929314A (en) * | 2020-08-26 | 2020-11-13 | 湖北汽车工业学院 | Wheel hub weld visual detection method and detection system |
| CN113421263A (en) * | 2021-08-24 | 2021-09-21 | 深圳市信润富联数字科技有限公司 | Part defect detection method, device, medium and computer program product |
| CN113702403A (en) * | 2021-08-02 | 2021-11-26 | 岭澳核电有限公司 | Method, equipment and medium for detecting welding seam of BOSS head of nuclear power station |
| WO2022156873A1 (en) * | 2021-01-22 | 2022-07-28 | MTU Aero Engines AG | A computer-implemented method for training an artificial neural network, method for inspecting a component, test system for examining components, and computer program |
| CN114862745A (en) * | 2021-02-04 | 2022-08-05 | 中国石油天然气股份有限公司 | Weld defect identification method, and training method and device of weld defect identification model |
| CN115082667A (en) * | 2021-03-16 | 2022-09-20 | 腾讯云计算（北京）有限责任公司 | Image processing method, device, equipment and storage medium |
-
2023
- 2023-02-13 CN CN202310103804.3A patent/CN116030030B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH1051693A (en) * | 1996-07-30 | 1998-02-20 | Toshiba Medical Eng Co Ltd | Defective picture element correction processing unit for image pickup element |
| US20060018534A1 (en) * | 2004-07-26 | 2006-01-26 | Dainippon Screen Mfg. Co., Ltd. | Technique for detecting a defect of an object by area segmentation of a color image of the object |
| JP2006308473A (en) * | 2005-04-28 | 2006-11-09 | Nippon Steel Corp | Method for inspecting periodic defect, and device |
| US20140175071A1 (en) * | 2011-06-29 | 2014-06-26 | Trumpf Laser - Und Systemtechnik Gmbh | Method for Detecting Defects During a Laser-Machining Process and Laser-Machining Device |
| US20140348415A1 (en) * | 2013-05-27 | 2014-11-27 | ThinkSmart IT Solutions Private Limited | System and method for identifying defects in welds by processing x-ray images |
| CN109342456A (en) * | 2018-09-14 | 2019-02-15 | 广东工业大学 | A kind of welding point defect detection method, device, equipment and readable storage medium storing program for executing |
| US20200117561A1 (en) * | 2018-10-11 | 2020-04-16 | International Business Machines Corporation | Using system errors and manufacturer defects in system components causing the system errors to determine a quality assessment value for the components |
| CN110866910A (en) * | 2019-11-13 | 2020-03-06 | 上海电气集团股份有限公司 | Weld quality prediction method, system, device and computer-readable storage medium |
| CN111932489A (en) * | 2020-06-03 | 2020-11-13 | 西安电子科技大学 | Weld defect detection method, system, storage medium, computer device and terminal |
| CN111929314A (en) * | 2020-08-26 | 2020-11-13 | 湖北汽车工业学院 | Wheel hub weld visual detection method and detection system |
| WO2022156873A1 (en) * | 2021-01-22 | 2022-07-28 | MTU Aero Engines AG | A computer-implemented method for training an artificial neural network, method for inspecting a component, test system for examining components, and computer program |
| CN114862745A (en) * | 2021-02-04 | 2022-08-05 | 中国石油天然气股份有限公司 | Weld defect identification method, and training method and device of weld defect identification model |
| CN115082667A (en) * | 2021-03-16 | 2022-09-20 | 腾讯云计算（北京）有限责任公司 | Image processing method, device, equipment and storage medium |
| CN113702403A (en) * | 2021-08-02 | 2021-11-26 | 岭澳核电有限公司 | Method, equipment and medium for detecting welding seam of BOSS head of nuclear power station |
| CN113421263A (en) * | 2021-08-24 | 2021-09-21 | 深圳市信润富联数字科技有限公司 | Part defect detection method, device, medium and computer program product |
Non-Patent Citations (2)
| Title |
|---|
| 刘涵;郭润元;: "基于X射线图像和卷积神经网络的石油钢管焊缝缺陷检测与识别", 仪器仪表学报, no. 04, pages 250 - 259 * |
| 董德威;颜云辉;陈立武;丛家慧;: "带钢表面缺陷严重程度的模糊综合评判研究", 计算机工程与设计, no. 04, pages 1484 - 1487 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116030030B (en) | 2023-08-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Radopoulou et al. | Automated detection of multiple pavement defects | |
| Kim et al. | Automated bridge component recognition from point clouds using deep learning | |
| CN111322985B (en) | Tunnel clearance analysis method, device and system based on laser point cloud | |
| Tan et al. | Automatic detection of sewer defects based on improved you only look once algorithm | |
| Duque et al. | Bridge deterioration quantification protocol using UAV | |
| WO2020124247A1 (en) | Automated inspection system and associated method for assessing the condition of shipping containers | |
| CN110969611A (en) | Pipeline weld defect detection method, device and system and storage medium | |
| CN111292275A (en) | Point cloud data filtering method and device based on complex ground and computer equipment | |
| Turkan et al. | Terrestrial laser scanning-based bridge structural condition assessment | |
| Kalasapudi et al. | Adaptive 3D imaging and tolerance analysis of prefabricated components for accelerated construction | |
| Kong et al. | Vision‐aided three‐dimensional damage quantification and finite element model geometric updating for reinforced concrete structures | |
| CN115713476A (en) | Visual detection method and device based on laser welding and readable storage medium | |
| Yamane et al. | Detecting and localising damage based on image recognition and structure from motion, and reflecting it in a 3D bridge model | |
| Dong et al. | Intelligent segmentation and measurement model for asphalt road cracks based on modified mask R-CNN algorithm | |
| CN116030030B (en) | Integrated assessment method for internal and external defects of weld joint of prefabricated part | |
| Catbas et al. | A review of latest trends in bridge health monitoring | |
| Zhou et al. | Corrosion inspection and evaluation of crane metal structure based on UAV vision | |
| CN116958099B (en) | Cable abrasion detection method, system, device and computer equipment | |
| Ranyal et al. | Enhancing pavement health assessment: An attention-based approach for accurate crack detection, measurement, and mapping | |
| US20220163325A1 (en) | Method for determining the geometry of a defect and for determining a load limit | |
| US20230297843A1 (en) | Deep learning method for defect characterization | |
| Yuan et al. | Automated pixel-level crack detection and quantification using deep convolutional neural networks for structural condition assessment | |
| Zhang et al. | High‐resolution 3‐D geometry updating of digital functional models using point cloud processing and surface cut | |
| Koch et al. | Machine vision techniques for condition assessment of civil infrastructure | |
| CN113159089A (en) | Pavement damage identification method, system, computer equipment and 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 |