CN114878582B - Defect detection and analysis method and system for special steel - Google Patents
Defect detection and analysis method and system for special steel Download PDFInfo
- Publication number
- CN114878582B CN114878582B CN202210766498.7A CN202210766498A CN114878582B CN 114878582 B CN114878582 B CN 114878582B CN 202210766498 A CN202210766498 A CN 202210766498A CN 114878582 B CN114878582 B CN 114878582B
- Authority
- CN
- China
- Prior art keywords
- special steel
- ultrasonic
- defect detection
- image
- detected
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/04—Analysing solids
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10132—Ultrasound image
Abstract
The invention provides a defect detection and analysis method and a system for special steel, which are applied to the technical field of steel defect detection, and the method comprises the following steps: and (3) building a data acquisition coordinate set by obtaining the information of the special steel to be detected. And carrying out image acquisition and ultrasonic data acquisition on the basis of the data acquisition coordinate set through the image acquisition device and the ultrasonic data acquisition device, and outputting a special steel image set and an ultrasonic data set. And performing characteristic analysis on the special steel to be detected to obtain a sample data set. And generating a special steel defect detection model according to the sample data set. And inputting the special steel image set and the ultrasonic data set into a special steel defect detection model, and obtaining a special steel defect detection result according to the special steel defect detection model. The accuracy of the defect detection of the special steel is improved, and the influence of human factors on the detection result is reduced. The technical problem that due to the fact that a standard method for accurately analyzing the defects of the special steel is lacked, human influence factors on the detection result of the special steel are large is solved.
Description
Technical Field
The invention relates to the technical field of steel defect detection, in particular to a method and a system for detecting and analyzing defects of special steel.
Background
The special steel is made up by adding one or several alloy elements to carbon steel in a proper quantity to change the structure of steel, so that the steel possesses different properties, such as high strength and hardness, good plasticity and toughness, wear resistance, corrosion resistance and other excellent properties. The existing special steel defect detection is mostly carried out in a manual mode, then, the detection result is evaluated manually, and the detection result of special steel is influenced by human factors.
Therefore, the prior art has the technical problem that the obtained detection result of the special steel is large in artificial influence factor and inaccurate due to the lack of a standard method capable of accurately analyzing the defects of the special steel.
Disclosure of Invention
The application provides a defect detection and analysis method and system for special steel, which are used for solving the technical problems that the acquired detection result of the special steel is large in artificial influence factor and inaccurate due to the fact that a method which lacks a standard and can be used for accurately analyzing the defects of the special steel exists in the prior art.
In view of the above problems, the present application provides a method and a system for detecting and analyzing defects of special steel.
In a first aspect of the present application, a method for detecting and analyzing defects of special steel is provided, the system is in communication connection with a data acquisition device, wherein the data acquisition device includes an image acquisition device and an ultrasonic data acquisition device, and the method includes: obtaining information of the special steel to be detected; according to the information of the special steel to be detected, a data acquisition coordinate set is set up; respectively carrying out image acquisition and ultrasonic data acquisition on the basis of the data acquisition coordinate set according to the image acquisition device and the ultrasonic data acquisition device, and outputting a special steel image set and an ultrasonic data set; performing characteristic analysis on the special steel to be detected to obtain a sample data set; generating a special steel defect detection model according to the sample data set; embedding an excitation network layer into the special steel defect detection model, wherein the excitation network layer is an optimized network layer for defect detection; and inputting the special steel image set and the ultrasonic data set into the special steel defect detection model, and obtaining an output result according to the special steel defect detection model, wherein the output result is a special steel defect detection result.
In a second aspect of the present application, a system for detecting and analyzing defects of special steel is provided, the system is in communication connection with a data acquisition device, wherein the data acquisition device includes an image acquisition device and an ultrasonic data acquisition device, and the system includes: the steel information acquisition module to be detected is used for acquiring the information of the special steel to be detected; the data acquisition coordinate set construction module is used for constructing a data acquisition coordinate set according to the information of the special steel to be detected; the data acquisition module is used for respectively carrying out image acquisition and ultrasonic data acquisition on the basis of the data acquisition coordinate set according to the image acquisition device and the ultrasonic data acquisition device and outputting a special steel image set and an ultrasonic data set; the sample data acquisition module is used for carrying out characteristic analysis on the special steel to be detected to obtain a sample data set; the special steel defect detection model acquisition module is used for generating a special steel defect detection model according to the sample data set; the optimized network layer adding module is used for embedding an excitation network layer into the special steel defect detection model, wherein the excitation network layer is an optimized network layer for defect detection; and the detection result acquisition module is used for inputting the special steel image set and the ultrasonic data set into the special steel defect detection model and acquiring an output result according to the special steel defect detection model, wherein the output result is a special steel defect detection result.
One or more technical solutions provided in the present application have at least the following technical effects or advantages:
according to the method provided by the embodiment of the application, the data acquisition coordinate set is constructed by acquiring the information of the steel to be detected. And carrying out an image acquisition device and an ultrasonic data acquisition device according to the data acquisition coordinate set. And then, performing characteristic analysis on the special steel to be detected to analyze the defect type of the special steel and corresponding acquired data to form a sample data set. Training the model through the sample data set to obtain a special steel defect detection model, and then embedding an excitation network layer in the model. And inputting the special steel image set and the ultrasonic data set into a steel defect detection model to obtain a final detection result. The method solves the technical problem that the detection result of the obtained special steel is large in artificial influence factor and inaccurate due to the fact that a standard-lacking method capable of accurately analyzing the defects of the special steel exists in the prior art. The accuracy of special steel defect detection is improved, and the technical effect of influence of human factors on the detection result is reduced.
The foregoing description is only an overview of the technical solutions of the present application, and the present application can be implemented according to the content of the description in order to make the technical means of the present application more clearly understood, and the following detailed description of the present application is given in order to make the above and other objects, features, and advantages of the present application more clearly understandable.
Drawings
FIG. 1 is a schematic flow chart of a defect detection and analysis method for special steel provided by the application;
fig. 2 is a schematic flow chart of a special steel image set obtained in the method for detecting and analyzing the defects of special steel provided by the present application;
fig. 3 is a schematic flow chart illustrating the process of acquiring an optimized special steel image set and an optimized ultrasonic data set in the method for detecting and analyzing the defects of the special steel provided by the present application;
FIG. 4 is a schematic structural diagram of a system for detecting and analyzing defects of special steel.
Description of reference numerals: the system comprises a steel information acquisition module to be detected 11, a data acquisition coordinate set construction module 12, a data acquisition module 13, a sample data acquisition module 14, a special steel defect detection model acquisition module 15, an optimized network layer adding module 16 and a detection result acquisition module 17.
Detailed Description
The application provides a defect detection and analysis method and system for special steel, which are used for solving the technical problems that the acquired detection result of the special steel is large in artificial influence factor and inaccurate due to the fact that a method which lacks a standard and can be used for accurately analyzing the defects of the special steel exists in the prior art.
The technical solution in the present application will be described clearly and completely with reference to the accompanying drawings. The described embodiments are only some of the implementations possible in the present application, and not all of the implementations possible in the present application.
Example one
As shown in fig. 1, the present application provides a method for detecting and analyzing defects of special steel, which is applied to a system for detecting and analyzing defects of special steel, the system being in communication connection with a data acquisition device, wherein the data acquisition device comprises an image acquisition device and an ultrasonic data acquisition device, and the method comprises:
step 100: obtaining information of the special steel to be detected;
the special steel is prepared by adding one or more alloy elements into carbon steel in a proper amount to change the tissue structure of the steel, so that the steel has different properties, such as high strength, high hardness, good plasticity and toughness, wear resistance, corrosion resistance and other excellent properties. Because the application environment of the special steel is harsh, the steel is often damaged when the special steel has defects, and safety accidents are caused. Therefore, it is necessary to perform flaw detection operation on the special steel, wherein the conventional flaw detection method uses a physical method to detect the steel, and in this embodiment, the special steel is subjected to flaw detection by performing image acquisition and ultrasonic data acquisition on the special steel. The special steel to be detected is the special steel to be detected in the embodiment of the application, and information of the special steel to be detected is obtained, wherein the information includes appearance information such as size, shape and the like of the special steel to be detected.
Step 200: according to the information of the special steel to be detected, a data acquisition coordinate set is set up;
specifically, the information of the special steel to be detected includes appearance information such as size and shape of the special steel to be detected through the acquired information of the special steel to be detected. And analyzing the appearance information of the special steel to be detected to obtain the size and the geometric shape of the special steel to be detected. When the ultrasonic probe detects the surface, the surface to be detected and the ultrasonic probe need to be placed in parallel. Therefore, the acquisition point of the ultrasonic probe needs to be determined according to the size and the geometric shape of the special steel to be detected. The specific acquisition coordinate of the ultrasonic probe, namely a data acquisition coordinate set, is determined according to the size and the geometric shape of the special steel, and the surface defects of the special steel are acquired according to the acquired data acquisition coordinate set.
Step 300: respectively carrying out image acquisition and ultrasonic data acquisition on the basis of the data acquisition coordinate set according to the image acquisition device and the ultrasonic data acquisition device, and outputting a special steel image set and an ultrasonic data set;
specifically, the surface of the special steel is collected based on a data collection coordinate set according to an image collection device and an ultrasonic data collection device, wherein the image collection device is used for collecting images of the positions of the data collection coordinate set, and the final collected images are obtained to form a special steel image set. The ultrasonic data acquisition device is used for acquiring ultrasonic data of the data acquisition coordinate set position and acquiring final ultrasonic data to form an ultrasonic data set.
As shown in fig. 2, the method steps 300 provided in the embodiment of the present application further include:
step 310: the relative distance between the special steel to be detected and the image acquisition device is configured, and the relative distance in the stage is output;
step 320: acquiring images according to the relative distances of the stages, and outputting a multi-stage special steel image set;
step 330: determining a stage image acquisition coordinate by performing image characteristic analysis on the multi-stage special steel image set;
step 340: performing defect image recognition according to the stage image acquisition coordinates, and outputting an identification special steel image set, wherein the identification special steel image set is an image with defects;
step 350: and outputting the special steel image set according to the identification special steel image set.
Specifically, the relative distance between the special steel to be detected and the image acquisition device is configured to obtain the relative acquisition distance between the special steel to be detected and the image acquisition device, that is, the stage relative distance, wherein the relative acquisition distance between the special steel to be detected and the data acquisition device can be set according to the sequence of the relative distances, for example, the relative acquisition distance between the special steel to be detected and the image acquisition device is acquired once by 20 centimeters, once by 15 centimeters and once by 5 centimeters to obtain the stage acquisition result. And acquiring images according to the relative acquisition distance between the special steel to be detected and the image acquisition device, and outputting the acquired special steel image to be detected, namely a multi-stage special steel image set. And performing characteristic analysis on the acquired multi-stage special steel image set to determine stage image acquisition coordinates, wherein the image acquisition coordinates comprise relative distance coordinates and position coordinates, and the position coordinates correspond to the positions of the data acquisition coordinate set. And identifying the defect images according to the phase image acquisition coordinates, and outputting a special steel identification image set of the phase images, wherein the special steel identification image set comprises the images of the special steel to be detected with defects and the position identification of the images. And outputting a special steel image set according to the identification special steel image set, wherein the special steel image set is an image and position set of the special steel to be detected with defects.
The method steps 300 provided by the embodiment of the present application further include:
step 360: obtaining a preset sensitivity threshold;
step 370: performing sensitivity rising according to the preset sensitivity threshold value to obtain ultrasonic wave distribution data of the special steel to be detected under different sensitivities;
step 380: analyzing the defects of the special steel to be detected according to the ultrasonic distribution data, and outputting ultrasonic characteristics;
step 390: and identifying and classifying the defects according to the ultrasonic characteristics to obtain a defect detection result corresponding to the ultrasonic distribution data.
Specifically, the sensitivity is the ability of the instrument and the probe to detect the smallest defect in the ultrasonic flaw detection, and the smaller the detected defect, the higher the sensitivity, and the larger the detected defect, the lower the sensitivity. Since it is necessary to detect defects of different sizes in ultrasonic flaw detection, it is necessary to adjust the sensitivity of ultrasonic flaw detection. And obtaining a preset sensitivity threshold, wherein the preset sensitivity threshold is the optimal sensitivity threshold when the defects with different sizes are detected. And (4) performing sensitivity increase according to a preset sensitivity threshold, namely improving the sensitivity of ultrasonic flaw detection and acquiring ultrasonic distribution data of the special steel to be detected under different sensitivities. And analyzing the defect of the special steel to be detected according to the ultrasonic distribution data, analyzing the ultrasonic distribution data, acquiring the ultrasonic waveform characteristics which can reflect the defect of the special steel, and acquiring the ultrasonic waveform characteristics which can reflect the defect of the special steel. The defects of different types of special steel can show different types of ultrasonic characteristics during ultrasonic flaw detection, for example, when cracks appear in steel parts, the waveform characteristics of the defects have envelope lines, the waveforms are relatively independent, and the defects can be found when the probe is slid from different directions. And analyzing the types of the defects according to the ultrasonic waveform characteristics reflecting the defects of the special steel, acquiring the defect types of the special steel to be detected, and finally acquiring the defect detection result corresponding to the ultrasonic distribution data. The ultrasonic wave distribution data are analyzed to obtain the final defect detection result, so that the defect of the special steel is accurately detected.
The method steps 390 provided by the embodiment of the present application further include:
step 391: acquiring a waveform sample feature set of ultrasonic detection according to the sample data set;
step 392: obtaining a defect detection category based on the waveform sample feature set, wherein the waveform sample feature set corresponds to the defect detection category one to one;
step 393: training a defect detection classifier according to the mapping relation between the waveform sample feature set and the defect detection category to generate a defect detection classification model;
step 394: and outputting a defect detection result corresponding to the ultrasonic distribution data based on the defect detection classification model.
Specifically, different types of ultrasonic characteristics can be reflected when the defects of different types of special steel are subjected to ultrasonic flaw detection, for example, envelope lines exist in waveform characteristics of the defects when cracks appear in steel, waveforms are relatively independent, and the defects can be found when the probe is slid from different directions. And acquiring a data set of the steel sample to be detected, wherein the data set comprises ultrasonic waveform characteristics and corresponding special steel defects, and the waveform sample characteristic sets correspond to the defect detection categories one by one. And training a defect detection classifier according to the mapping relation between the waveform sample feature set and the defect detection category, taking the sample data set as training data of the classifier, training the classification model, and obtaining a final defect detection classification model. The classification model used for training may be selected according to actual needs, for example, a decision tree model may be selected as the classification model, which is not specifically limited herein. By obtaining the defect detection classification model, the defects of the special steel can be obtained through the distribution data of the ultrasonic waves, and the final defect detection result is output.
The method step 380 provided by the embodiment of the present application further includes:
step 381: obtaining an ultrasonic probe according to the ultrasonic data acquisition device, wherein an inertial sensor is embedded in the ultrasonic data acquisition device;
step 382: carrying out angle parallel sensing on the relative position of the ultrasonic probe and the special steel to be detected based on the inertial sensor to obtain a real-time sensing angle of the ultrasonic probe;
step 383: and judging whether the real-time sensing angle of the ultrasonic probe is a parallel sensing angle or not, and if not, acquiring reminding information.
Specifically, an ultrasonic probe of an ultrasonic data acquisition device is obtained, and an inertial sensor is embedded in the ultrasonic data acquisition device. Because need keep the contact point of probe and wait to detect the parallelism of steel in real time when carrying out the detection of detecting a flaw, when the contact point of probe and wait to detect the angle of steel and change by the ultrasonic wave of probe transmission by the bottom reflection return to the probe, the part is gathered by ultrasonic transducer, and the part is refracted back to the bottom again, and the ultrasonic data that obtains this moment is then inaccurate, consequently need go on through embedded inertial sensor among the ultrasonic data collection system, acceleration and gradient measurement. And then, carrying out angle parallel sensing on the relative position of the ultrasonic probe and the special steel to be detected based on the inertial sensor, wherein the angle parallel sensing is to detect whether the contact angle of the ultrasonic probe and the special steel to be detected is parallel or not, and acquiring a real-time sensing angle of the ultrasonic probe, wherein the real-time sensing angle is the contact angle of ultrasonic waves and the special steel to be detected. And judging whether the real-time sensing angle of the ultrasonic probe is a parallel sensing angle, namely judging whether the contact between the ultrasonic probe and the special steel to be detected is parallel contact, and if the contact between the ultrasonic probe and the special steel to be detected is not parallel contact, namely the real-time sensing angle of the ultrasonic probe is not the parallel sensing angle, acquiring reminding information. Whether the real-time sensing angle is the parallel sensing angle or not is judged by acquiring the real-time sensing angle, so that the accuracy of ultrasonic measurement data is further ensured, and the final output detection result is more accurate.
Step 400: performing characteristic analysis on the special steel to be detected to obtain a sample data set;
specifically, the special steel to be detected is subjected to characteristic analysis, the defects of the special steel to be detected and corresponding ultrasonic data, namely ultrasonic waveform characteristics and image characteristics, are obtained, different types of ultrasonic characteristics can be embodied when the defects of the special steel of different types are subjected to ultrasonic flaw detection, the ultrasonic waveform characteristics and the defects of the special steel have corresponding relations, and if cracks appear in a steel piece, the waveform characteristics have envelope lines, the waveforms are relatively independent, and the defects of notches can be found when a probe is slid from different directions. Therefore, the corresponding special steel defect type can be obtained according to the analysis of the detected ultrasonic waveform characteristics. Meanwhile, when the special steel has defects, corresponding image data can be reflected on the surface of the steel, and a sample data set is generated according to the corresponding relation between the image data and the ultrasonic data and the defects of the special steel to be detected.
The method steps 400 provided by the embodiments of the present application include:
step 410: obtaining the surface roughness according to the information of the special steel to be detected;
step 420: generating self-checking conditions based on the surface roughness, wherein the self-checking conditions comprise image self-checking conditions and ultrasonic self-checking conditions;
step 430: and self-checking the special steel image set and the ultrasonic data set based on the self-checking condition, and correcting the special steel image set and the ultrasonic data set according to a self-checking result.
Specifically, the surface roughness of the steel to be detected is obtained according to the information of the steel to be detected. The method includes the steps that a self-checking condition is generated according to the surface roughness of steel to be detected, when the surface of steel to be detected is rough, the obtained special steel image data set and ultrasonic data set can be affected, for example, the defect characteristics of data in the special steel image data set and the ultrasonic data set obtained when the surface of the steel to be detected is rough are not obvious, large interference exists, and the detection result obtained by taking the data as input data is possibly inaccurate. Therefore, a self-inspection condition needs to be generated according to the surface roughness, and when the roughness reaches a certain degree, the image self-inspection condition and the ultrasonic self-inspection condition are respectively set for the image data set and the ultrasonic data set correspondingly. When the self-checking condition is that a specific characteristic caused by roughness appears in the image and the ultrasonic data, the characteristic in the image and the ultrasonic data is subjected to self-checking. The image and the ultrasonic data are self-checked through the self-checking condition, and the situation that the acquired image and the ultrasonic data are judged to be a defect due to the influence of roughness is prevented. And then, correcting the special steel image set and the ultrasonic data set according to the self-checking result, and filtering out specific characteristics generated by the image and the ultrasonic data due to the influence of the roughness during correction. The situation that the roughness affects the acquired image and ultrasonic data so that the defect is judged is avoided.
Step 500: generating a special steel defect detection model according to the sample data set;
specifically, the sample data set includes image features, ultrasonic data and corresponding special steel defect types, and the data correspond to defect detection types one to one, that is, one special steel defect type corresponds to one image data and one ultrasonic data. And carrying out supervision training on the model according to the sample data set, wherein the model is specifically a neural network model. And taking the seed image data and the ultrasonic data in the sample data set as input data, taking the defect detection types corresponding to the image data and the ultrasonic data as supervision data to supervise and train the neural network model, and obtaining a final training result, namely the special steel defect detection model. The special steel defect detection model is used for detecting the corresponding special steel defects according to the ultrasonic data and the image data.
Step 600: embedding an excitation network layer into the special steel defect detection model, wherein the excitation network layer is an optimized network layer for defect detection;
step 700: and inputting the special steel image set and the ultrasonic data set into the special steel defect detection model, and obtaining an output result according to the special steel defect detection model, wherein the output result is a special steel defect detection result.
Specifically, the excitation network layer is embedded into the special steel defect detection model, wherein the excitation network layer is an excitation function for optimizing the neural network model and is used for introducing nonlinear factors into the neural network model, so that the obtained output result is more accurate. For example, according to historical detection data and detection results, it is known that in a detection time period from 8 am to 8 pm, a detection result at noon is closest to a real result, and other time periods generate errors with different sizes to obtain a nonlinear relation between the detection time period and the detection errors, and the nonlinear relation is used as an excitation network layer to perform parameter optimization on input parameters influencing the detection time period. Excitation functions such as detection distance, detection time and detection sensitivity are introduced into the neural network model to excite the special steel defect detection model, so that the detection distance, the detection time and the detection sensitivity all affect the output result of the special steel defect detection model, and the output result is closer to a real detection result. And inputting the special steel image set and the ultrasonic data set into the special steel defect detection model to obtain a final output result, wherein the output result is a special steel defect detection result.
As shown in fig. 3, a method step 600 provided by the embodiment of the present application includes:
step 610: inputting the special steel image set and the ultrasonic data set into the special steel defect detection model, and outputting a parameter optimization result according to the excitation network layer in the special steel defect detection model, wherein the parameter optimization result comprises detection distance, detection time and detection sensitivity;
step 620: processing the special steel image set and the ultrasonic data set according to the parameter optimization result, and outputting the optimized special steel image set and the optimized ultrasonic data set;
step 630: and obtaining the output result based on the optimized special steel image set and the optimized ultrasonic data set.
Specifically, the special steel image set and the ultrasonic data set are input into a special steel defect detection model, and a parameter optimization result is output according to an excitation network layer in the special steel defect detection model, wherein the parameter optimization result comprises detection distance, detection time and detection sensitivity. And optimizing the special steel image set and the ultrasonic data set according to the obtained parameter optimization result, optimizing elements in the sets, and outputting the optimized special steel image set and the optimized ultrasonic data set. And then inputting the optimized special steel image set and the optimized ultrasonic data set into a special steel defect detection model to obtain a final output result of the model. The final output detection result is more accurate.
In summary, the method provided by the embodiment of the application constructs the data acquisition coordinate set by acquiring the information of the steel to be detected. And carrying out an image acquisition device and an ultrasonic data acquisition device according to the data acquisition coordinate set. And then, performing characteristic analysis on the special steel to be detected, and analyzing the defect type of the special steel and the corresponding acquired data to form a sample data set. Training the model through a sample data set to obtain a special steel defect detection model, and then embedding an excitation network layer in the model. And inputting the special steel image set and the ultrasonic data set into a steel defect detection model to obtain a final detection result. And the input parameters are optimized by adding an excitation network layer, so that the input parameters are more accurate, and the accuracy of the detection result is improved. The method solves the technical problem that the detection result of the obtained special steel is large in artificial influence factor and inaccurate due to the fact that a standard-lacking method capable of accurately analyzing the defects of the special steel exists in the prior art. The accuracy of the defect detection of the special steel is improved, and the technical effect of influence of human factors on the detection result is reduced.
Example two
Based on the same inventive concept as the defect detection and analysis method of special steel in the previous embodiment, as shown in fig. 4, the present application provides a defect detection and analysis system of special steel, the system is connected with a data acquisition device in a communication manner, wherein the data acquisition device comprises an image acquisition device and an ultrasonic data acquisition device, and the method comprises:
the steel information acquisition module 11 is used for acquiring information of special steel to be detected;
the data acquisition coordinate set construction module 12 is used for constructing a data acquisition coordinate set according to the information of the special steel to be detected;
the data acquisition module 13 is used for respectively acquiring images and ultrasonic data based on the data acquisition coordinate set according to the image acquisition device and the ultrasonic data acquisition device, and outputting a special steel image set and an ultrasonic data set;
the sample data acquisition module 14 is used for performing characteristic analysis on the special steel to be detected to obtain a sample data set;
the special steel defect detection model acquisition module 15 is used for generating a special steel defect detection model according to the sample data set;
an optimized network layer adding module 16, configured to embed an excitation network layer into the special steel defect detection model, where the excitation network layer is an optimized network layer for performing defect detection;
and the detection result acquisition module 17 is configured to input the special steel image set and the ultrasonic data set into the special steel defect detection model, and acquire an output result according to the special steel defect detection model, where the output result is a special steel defect detection result.
Further, the data obtaining module 13 is further configured to:
the relative distance between the special steel to be detected and the image acquisition device is configured, and the relative distance in the stage is output;
acquiring images according to the relative distances of the stages, and outputting a multi-stage special steel image set;
determining a stage image acquisition coordinate by carrying out image characteristic analysis on the multi-stage special steel image set;
performing defect image recognition according to the stage image acquisition coordinates, and outputting an identification special steel image set, wherein the identification special steel image set is an image with defects;
and outputting the special steel image set according to the identification special steel image set.
Further, the data obtaining module 13 is further configured to:
obtaining a preset sensitivity threshold;
performing sensitivity rising according to the preset sensitivity threshold value to obtain ultrasonic wave distribution data of the special steel to be detected under different sensitivities;
analyzing the defects of the special steel to be detected according to the ultrasonic distribution data, and outputting ultrasonic characteristics;
and identifying and classifying the defects according to the ultrasonic characteristics to obtain a defect detection result corresponding to the ultrasonic distribution data.
Further, the sample data obtaining module 14 is further configured to:
acquiring a waveform sample feature set of ultrasonic detection according to the sample data set;
obtaining a defect detection category based on the waveform sample feature set, wherein the waveform sample feature set corresponds to the defect detection category one to one;
training a defect detection classifier according to the mapping relation between the waveform sample feature set and the defect detection category to generate a defect detection classification model;
and outputting a defect detection result corresponding to the ultrasonic distribution data based on the defect detection classification model.
Further, the detection result obtaining module 17 is further configured to:
inputting the special steel image set and the ultrasonic data set into the special steel defect detection model, and outputting a parameter optimization result according to the excitation network layer in the special steel defect detection model, wherein the parameter optimization result comprises detection distance, detection time and detection sensitivity;
processing the special steel image set and the ultrasonic data set according to the parameter optimization result, and outputting the optimized special steel image set and the optimized ultrasonic data set;
and obtaining the output result based on the optimized special steel image set and the optimized ultrasonic data set.
Further, the data obtaining module 13 is further configured to:
obtaining the surface roughness according to the information of the special steel to be detected;
generating self-checking conditions based on the surface roughness, wherein the self-checking conditions comprise image self-checking conditions and ultrasonic self-checking conditions;
and self-checking the special steel image set and the ultrasonic data set based on the self-checking condition, and correcting the special steel image set and the ultrasonic data set according to a self-checking result.
Further, the data obtaining module 13 is further configured to:
obtaining an ultrasonic probe according to the ultrasonic data acquisition device, wherein an inertial sensor is embedded in the ultrasonic data acquisition device;
carrying out angle parallel sensing on the relative position of the ultrasonic probe and the special steel to be detected based on the inertial sensor to obtain a real-time sensing angle of the ultrasonic probe;
and judging whether the real-time sensing angle of the ultrasonic probe is a parallel sensing angle or not, and if not, acquiring reminding information.
The second embodiment is used for executing the method as in the first embodiment, and both the execution principle and the execution basis can be obtained through the content recorded in the first embodiment, which is not described herein again. Although the present application has been described in connection with particular features and embodiments thereof, the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application, and the content thus obtained also falls within the scope of protection of the present application.
Claims (7)
1. The method for detecting and analyzing the defects of the special steel is applied to a defect detecting and analyzing system of the special steel, the system is in communication connection with a data acquisition device, wherein the data acquisition device comprises an image acquisition device and an ultrasonic data acquisition device, and the method comprises the following steps:
obtaining information of the special steel to be detected;
according to the information of the special steel to be detected, a data acquisition coordinate set is set up;
respectively carrying out image acquisition and ultrasonic data acquisition on the basis of the data acquisition coordinate set according to the image acquisition device and the ultrasonic data acquisition device, and outputting a special steel image set and an ultrasonic data set;
performing characteristic analysis on the special steel to be detected to obtain a sample data set;
generating a special steel defect detection model according to the sample data set;
embedding an excitation network layer into the special steel defect detection model, wherein the excitation network layer is an optimized network layer for defect detection;
inputting the special steel image set and the ultrasonic data set into the special steel defect detection model, and obtaining an output result according to the special steel defect detection model, wherein the output result is a special steel defect detection result;
inputting the special steel image set and the ultrasonic data set into the special steel defect detection model, and obtaining an output result according to the special steel defect detection model, wherein the method further comprises the following steps:
inputting the special steel image set and the ultrasonic data set into the special steel defect detection model, and outputting a parameter optimization result according to the excitation network layer in the special steel defect detection model, wherein the parameter optimization result comprises detection distance, detection time and detection sensitivity;
processing the special steel image set and the ultrasonic data set according to the parameter optimization result, and outputting the optimized special steel image set and the optimized ultrasonic data set;
and obtaining the output result based on the optimized special steel image set and the optimized ultrasonic data set.
2. The method of claim 1, wherein the image acquisition and the ultrasound data acquisition are performed based on the set of data acquisition coordinates according to the image acquisition device and the ultrasound data acquisition device, respectively, the method further comprising:
the relative distance between the special steel to be detected and the image acquisition device is configured, and the relative distance of stages is output;
acquiring images according to the relative distances of the stages, and outputting a multi-stage special steel image set;
determining a stage image acquisition coordinate by carrying out image characteristic analysis on the multi-stage special steel image set;
performing defect image recognition according to the stage image acquisition coordinates, and outputting an identification special steel image set, wherein the identification special steel image set is an image with defects;
and outputting the special steel image set according to the identification special steel image set.
3. The method of claim 2, wherein the method further comprises:
obtaining a preset sensitivity threshold;
performing sensitivity rising according to the preset sensitivity threshold value to obtain ultrasonic wave distribution data of the special steel to be detected under different sensitivities;
analyzing the defects of the special steel to be detected according to the ultrasonic distribution data, and outputting ultrasonic characteristics;
and identifying and classifying the defects according to the ultrasonic characteristics to obtain a defect detection result corresponding to the ultrasonic distribution data.
4. The method according to claim 3, wherein the defect identification and classification is performed according to the ultrasonic features to obtain a defect detection result corresponding to the ultrasonic distribution data, and the method further comprises:
obtaining a waveform sample feature set of ultrasonic detection according to the sample data set;
obtaining a defect detection category based on the waveform sample feature set, wherein the waveform sample feature set corresponds to the defect detection category one to one;
training a defect detection classifier according to the mapping relation between the waveform sample feature set and the defect detection category to generate a defect detection classification model;
and outputting a defect detection result corresponding to the ultrasonic distribution data based on the defect detection classification model.
5. The method of claim 1, wherein after outputting the set of special steel images and the set of ultrasonic data, the method further comprises:
obtaining the surface roughness according to the information of the special steel to be detected;
generating self-checking conditions based on the surface roughness, wherein the self-checking conditions comprise image self-checking conditions and ultrasonic self-checking conditions;
and self-checking the special steel image set and the ultrasonic data set based on the self-checking condition, and correcting the special steel image set and the ultrasonic data set according to a self-checking result.
6. The method of claim 1, wherein the method further comprises:
obtaining an ultrasonic probe according to the ultrasonic data acquisition device, wherein an inertial sensor is embedded in the ultrasonic data acquisition device;
carrying out angle parallel sensing on the relative position of the ultrasonic probe and the special steel to be detected based on the inertial sensor to obtain a real-time sensing angle of the ultrasonic probe;
and judging whether the real-time sensing angle of the ultrasonic probe is a parallel sensing angle or not, and if not, acquiring reminding information.
7. The system for detecting and analyzing the defects of the special steel is characterized by being in communication connection with a data acquisition device, wherein the data acquisition device comprises an image acquisition device and an ultrasonic data acquisition device, and the system comprises:
the system comprises a to-be-detected steel information acquisition module, a to-be-detected special steel information acquisition module and a to-be-detected special steel information acquisition module, wherein the to-be-detected steel information acquisition module is used for acquiring information of the to-be-detected special steel;
the data acquisition coordinate set construction module is used for constructing a data acquisition coordinate set according to the information of the special steel to be detected;
the data acquisition module is used for respectively carrying out image acquisition and ultrasonic data acquisition on the basis of the data acquisition coordinate set according to the image acquisition device and the ultrasonic data acquisition device and outputting a special steel image set and an ultrasonic data set;
the sample data acquisition module is used for carrying out characteristic analysis on the special steel to be detected to obtain a sample data set;
the special steel defect detection model acquisition module is used for generating a special steel defect detection model according to the sample data set;
the optimized network layer adding module is used for embedding an excitation network layer into the special steel defect detection model, wherein the excitation network layer is an optimized network layer for defect detection;
the detection result acquisition module is used for inputting the special steel image set and the ultrasonic data set into the special steel defect detection model and acquiring an output result according to the special steel defect detection model, wherein the output result is a special steel defect detection result;
the detection result acquisition module is further configured to:
inputting the special steel image set and the ultrasonic data set into the special steel defect detection model, and outputting a parameter optimization result according to the excitation network layer in the special steel defect detection model, wherein the parameter optimization result comprises detection distance, detection time and detection sensitivity;
processing the special steel image set and the ultrasonic data set according to the parameter optimization result, and outputting the optimized special steel image set and the optimized ultrasonic data set;
and obtaining the output result based on the optimized special steel image set and the optimized ultrasonic data set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210766498.7A CN114878582B (en) | 2022-07-01 | 2022-07-01 | Defect detection and analysis method and system for special steel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210766498.7A CN114878582B (en) | 2022-07-01 | 2022-07-01 | Defect detection and analysis method and system for special steel |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114878582A CN114878582A (en) | 2022-08-09 |
| CN114878582B true CN114878582B (en) | 2022-10-18 |
Family
ID=82683562
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210766498.7A Active CN114878582B (en) | 2022-07-01 | 2022-07-01 | Defect detection and analysis method and system for special steel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114878582B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115983687B (en) * | 2022-12-22 | 2023-09-29 | 北京弥天科技有限公司 | Intelligent detection management system and method for quality of cold-rolled strip steel |
| CN116205922B (en) * | 2023-05-05 | 2023-10-13 | 张家港广大特材股份有限公司 | Smelting control method and system based on steel internal defect data mining |
| CN116511998B (en) * | 2023-06-14 | 2023-10-20 | 冈田智能(江苏)股份有限公司 | Defect detection method and system for circular cutterhead |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106706759B (en) * | 2016-12-15 | 2021-02-26 | 国网山东省电力公司电力科学研究院 | Method for evaluating defects of welding joint of P92 steel main steam pipeline of ultra-supercritical generator set |
| CN111024728B (en) * | 2019-12-26 | 2022-04-26 | 江西交通职业技术学院 | Railway detection method and system based on computer vision and ultrasonic flaw detection |
| CN112232400B (en) * | 2020-10-12 | 2023-06-20 | 太原科技大学 | Stainless steel weld ultrasonic defect detection method based on depth feature fusion |
| CN113239504B (en) * | 2021-06-30 | 2022-01-28 | 西南石油大学 | Pipeline corrosion defect prediction method based on optimized neural network |
| CN114264723A (en) * | 2021-12-22 | 2022-04-01 | 上海骏然网络科技有限公司 | Textile quality detection system and method based on ultrasonic detection |
| CN114463282A (en) * | 2022-01-13 | 2022-05-10 | 江苏六道建材有限公司 | Defect detection method, defect detection device, electronic equipment and computer readable medium |
-
2022
- 2022-07-01 CN CN202210766498.7A patent/CN114878582B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN114878582A (en) | 2022-08-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114878582B (en) | Defect detection and analysis method and system for special steel | |
| CN109253921B (en) | Evaluation method for detecting strength of concrete test block | |
| CN110082429B (en) | Tunnel lining nondestructive testing auxiliary judgment method combining machine learning | |
| CN102928435A (en) | Aircraft skin damage identification method and device based on image and ultrasound information fusion | |
| CN109190272B (en) | Concrete structure defect detection method based on elastic waves and machine learning | |
| CN116818914B (en) | Glass and nondestructive testing method for processed product thereof | |
| CN114926462A (en) | Intelligent detection method and system for metal material surface defects | |
| CN108844856B (en) | Sleeve grouting defect nondestructive testing method based on impact elastic wave and machine learning | |
| CN101074927A (en) | Method for diagnosing fruit diseases based on visible and near-infrared spectral characteristic band | |
| CN110568082A (en) | cable wire breakage distinguishing method based on acoustic emission signals | |
| CN110455923B (en) | Rapid evaluation method for anchor rod anchoring quality grade | |
| CN113720914B (en) | Ultrasonic flaw detection system and ultrasonic flaw detection method | |
| CN110231401B (en) | Inter-measuring point information correlation method and civil structure nondestructive testing method | |
| CN111879709A (en) | Method and device for detecting spectral reflectivity of lake water body | |
| CN116297867A (en) | Acoustic emission multi-parameter coupling-based wood crack depth detection method | |
| CN114881938A (en) | Grain size detection method and system based on wavelet analysis and neural network | |
| CN113624759A (en) | Apple nondestructive testing method based on machine learning | |
| CN110749650B (en) | Anchor rod anchoring compactness grade evaluation method based on support vector machine | |
| CN110455922B (en) | K-nearest neighbor method-based anchor rod anchoring quality grade evaluation method | |
| CN112730634B (en) | Concrete defect detection method and system | |
| CN110826598B (en) | Anchor rod anchoring compactness grade evaluation method based on convolutional neural network | |
| CN113837163B (en) | Tunnel monitoring method and system based on three-dimensional ground penetrating radar and storage medium | |
| CN117805853B (en) | Laser reflection type edge detection device and method | |
| CN108955564A (en) | Laser data method for resampling and system | |
| CN115144259B (en) | Method and system for detecting deformation resistance of steel |
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 |