CN116433556A - Automatic labeling method for pipeline defects - Google Patents
Automatic labeling method for pipeline defects Download PDFInfo
- Publication number
- CN116433556A CN116433556A CN202111653019.2A CN202111653019A CN116433556A CN 116433556 A CN116433556 A CN 116433556A CN 202111653019 A CN202111653019 A CN 202111653019A CN 116433556 A CN116433556 A CN 116433556A
- Authority
- CN
- China
- Prior art keywords
- internal structure
- defect
- structure diagram
- sample
- target
- 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.)
- Pending
Links
- 230000007547 defect Effects 0.000 title claims abstract description 131
- 238000002372 labelling Methods 0.000 title claims description 18
- 238000010586 diagram Methods 0.000 claims abstract description 86
- 238000000034 method Methods 0.000 claims abstract description 20
- 238000001514 detection method Methods 0.000 claims description 28
- 238000013527 convolutional neural network Methods 0.000 claims description 12
- 238000001914 filtration Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 238000006243 chemical reaction Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 2
- 239000003208 petroleum Substances 0.000 description 2
- 230000002950 deficient Effects 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30204—Marker
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Quality & Reliability (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a method for automatically marking a pipeline defect, which comprises an acquisition unit, an execution unit, a generation unit and a determination unit, wherein the device comprises the following steps: acquiring electric signal data which are acquired by at least one target sensor and are aimed at a target pipeline; and executing preset operation on the electric signal data, generating a corresponding internal structure diagram based on the electric signal data after executing the preset operation, sending the acquired internal structure diagram to a server, acquiring the internal structure diagram and the design diagram of the pipeline from the server through third party equipment for comparison, and marking different places from the design diagram and in the design diagram. The invention can accurately display the position of the pipeline defect on the display screen, is beneficial to technicians or specialists to analyze the pipeline defect, eliminates the pipeline defect through staff, accelerates response time due to accurate position, and improves working efficiency.
Description
Technical Field
The invention belongs to the field of pipelines, and relates to a method for automatically marking defects of a pipeline.
Background
In some areas, piping is required during production or transportation. For example, the petroleum industry requires the transport of petroleum through pipelines. To avoid the occurrence of hazards and the resulting loss, it is often necessary to inspect the surface of the pipe and determine the type of defect present on the surface of the pipe. Currently, the surface of a pipe is generally inspected in two ways. Firstly, data about the surface of a pipeline is collected through a sensor, and then the collected data is analyzed by a technician, so that the type of suspected defects is judged. Second, data suspected to be defective is extracted from the data acquired by the sensor by means of an expert system, and then the extracted data is analyzed by a technician.
However, in any of the above modes, the detection result cannot be intuitively displayed, the defect condition of the pipeline cannot be mastered in time, and the defect of which section of pipeline cannot be accurately known.
Disclosure of Invention
1. The technical problems to be solved are as follows:
accurately grasp the position of the defect condition of the pipeline and intuitively embody the position.
2. The technical scheme is as follows:
in order to solve the above problems, the present invention provides a method for automatically labeling a pipe defect, by detecting a device for a pipe defect, the device including an acquisition unit configured to acquire electrical signal data for a target pipe acquired by at least one target sensor; an execution unit configured to perform a preset operation on the electrical signal data, wherein the preset operation includes filtering; a generation unit configured to generate a corresponding internal structure diagram based on the electrical signal data after performing the preset operation; a determining unit configured to determine a category of the defect and a location of the region in the internal structure map in response to determining that the internal structure map includes a region characterizing the target pipe as having the defect, comprising the steps of: acquiring electric signal data which are acquired by at least one target sensor and are aimed at a target pipeline; and executing preset operation on the electric signal data, generating a corresponding internal structure diagram based on the electric signal data after executing the preset operation, sending the obtained internal structure diagram to a server, obtaining the internal structure diagram and the design diagram of the pipeline from the server through third party equipment for comparison, and marking different places from the design diagram and in the design diagram to obtain an area of the internal structure diagram including the defect representing the target pipeline.
Based on the electric signal data after the preset operation is executed, the corresponding internal structure diagram is generated, and the method comprises the following steps: and performing image interpolation based on the electric signal data after the preset operation is performed, and generating a corresponding internal structure diagram.
The determining, in response to determining that the internal structure map includes a region characterizing a defect of the target pipe, a category of the defect and a location of the region in the internal structure map, comprising: inputting the internal structure diagram into a pre-trained defect detection model to obtain the type of the defect indicated by the region representing the defect of the target pipeline and the position information representing the position of the region representing the defect of the target pipeline in the internal structure diagram, wherein the defect detection model is used for indicating the correspondence between the type of the defect indicated by the internal structure diagram and the region representing the defect of the target pipeline and the position of the region representing the defect of the target pipeline in the internal structure diagram.
The defect detection model is obtained through training the following steps: obtaining a sample set, wherein the sample comprises a sample internal structure diagram and sample labeling information, the sample internal structure diagram comprises at least one region representing that a target pipeline has defects, and the sample labeling information is used for indicating the types of the defects indicated by the region representing that the target pipeline has the defects and the positions of the region indicating that the target pipeline has the defects in the sample internal structure diagram; and selecting samples from the sample set, respectively taking a sample internal structure diagram and sample labeling information of the selected samples as input and expected output of an initial model, and training to obtain a defect detection model.
The initial model comprises a deep convolutional neural network and a classifier, wherein the deep convolutional neural network is used for extracting the characteristics of an internal structure diagram.
The generating unit is further configured to: and performing image interpolation based on the electric signal data after the preset operation is performed, and generating a corresponding internal structure diagram.
The determining unit is further configured to: inputting the internal structure diagram into a pre-trained defect detection model to obtain the type of the defect indicated by the region representing the defect of the target pipeline and the position information representing the position of the region representing the defect of the target pipeline in the internal structure diagram, wherein the defect detection model is used for indicating the correspondence between the type of the defect indicated by the internal structure diagram and the region representing the defect of the target pipeline and the position of the region representing the defect of the target pipeline in the internal structure diagram.
The determination unit includes: an acquisition unit configured to acquire a sample set, the sample including a sample internal structure map including at least one region characterizing a defect of the target pipe, and sample annotation information indicating a type of defect indicated by the region characterizing the defect of the target pipe and a position of the region indicating the defect of the target pipe in the sample internal structure map; the training unit is configured to select samples from the sample set, respectively taking a sample internal structure diagram and sample labeling information of the selected samples as input and expected output of an initial model, and training to obtain a defect detection model.
The initial model comprises a deep convolutional neural network and a classifier, wherein the deep convolutional neural network is used for extracting the characteristics of an internal structure diagram.
3. The beneficial effects are that:
the invention can accurately display the position of the pipeline defect on the display screen, is beneficial to technicians or specialists to analyze the pipeline defect, eliminates the pipeline defect through staff, accelerates response time due to accurate position, and improves working efficiency.
Detailed Description
The present invention will be described in detail below.
The system architecture includes a plurality of sensors, a database server and a server. The network is used to provide a medium for communication links between the sensors, database servers, and servers. The network may include various connection categories such as wired, wireless communication links, or fiber optic cables, among others.
The sensor may interact with the database server and servers over a network to receive or transmit data, etc.
The sensor typically includes a sensing element, a conversion circuit, an auxiliary power supply, and the like. The sensing element is used for acquiring the physical quantity of the detected object. The conversion element is used for converting the physical quantity output by the sensitive element into an electric signal. The conversion circuit is configured to process (e.g., amplify) the electrical signal output from the conversion element, and output the processed electrical signal to other devices (e.g., database server and server). The auxiliary power supply can provide power for the sensitive element, the conversion circuit and the like. The sensor may include any number of sensing elements, conversion circuits, and auxiliary power sources.
The server can directly acquire the electric signal data acquired by the sensor and then provide various services. The server may also obtain electrical signal data pre-stored by the database server. The sensor and the collected electrical signal data may be directly stored in the local area of the server, and the server may directly extract and process the electrical signal data stored in the local area, and in this case, there may be no database server.
The servers (e.g., servers and database servers) may be hardware or software. When the server is hardware, the server may be implemented as a distributed server cluster formed by a plurality of servers, or may be implemented as a single server. When the server is software, it may be implemented as a plurality of software or software modules (e.g., to provide distributed services), or as a single software or software module.
A method for automatically marking a pipe defect, by means of a device for detecting a pipe defect, the device comprising an acquisition unit configured to acquire electrical signal data for a target pipe acquired by at least one target sensor; an execution unit configured to perform a preset operation on the electrical signal data, wherein the preset operation includes filtering; a generation unit configured to generate a corresponding internal structure diagram based on the electrical signal data after performing the preset operation; a determining unit configured to determine a category of the defect and a location of the region in the internal structure map in response to determining that the internal structure map includes a region characterizing the target pipe as having the defect, comprising the steps of: acquiring electric signal data which are acquired by at least one target sensor and are aimed at a target pipeline; and executing preset operation on the electric signal data, generating a corresponding internal structure diagram based on the electric signal data after executing the preset operation, sending the obtained internal structure diagram to a server, obtaining the internal structure diagram and the design diagram of the pipeline from the server through third party equipment for comparison, and marking different places from the design diagram and in the design diagram to obtain an area of the internal structure diagram including the defect representing the target pipeline.
According to the method provided by the embodiment of the application, firstly, voltage signal data which are acquired by the target sensor and are aimed at the target pipeline are subjected to preset operation. And then, performing image reconstruction on the voltage signal data after the preset operation is performed, and generating an internal structure diagram. Further, the location in the internal structure map of the region of the internal structure map that characterizes the defect in the target pipe and the type of the defect may be determined. The method further determines the type of the defect existing in the target pipeline by generating an internal structure diagram aiming at the target pipeline.
The defect detection model can be obtained through training as follows. The execution body of the training defect detection model may be the same as or different from the execution body of the method for detecting a pipe defect.
First, a sample set is obtained. The sample comprises a sample internal structure diagram and sample labeling information. The sample internal structure map includes at least one region that characterizes a defect in the target pipe. The sample labeling information is used for indicating the type of defect indicated by the region representing the defect of the target pipeline and the position of the region indicating the defect of the target pipeline in the internal structure diagram of the sample.
In these implementations, the execution body that trains the defect detection model may have stored in advance the sample internal structure map and the sample annotation information. Thus, the execution subject that trains the defect detection model can directly obtain a sample from the local as a sample set. In addition, the execution subject that trains the defect detection model may also obtain samples from a communicatively connected database server as a sample set.
Secondly, selecting samples from the sample set, taking a sample internal structure diagram and sample labeling information of the selected samples as input and expected output of an initial model respectively, and training to obtain a defect detection model.
Based on the electric signal data after the preset operation is executed, the corresponding internal structure diagram is generated, and the method comprises the following steps: and performing image interpolation based on the electric signal data after the preset operation is performed, and generating a corresponding internal structure diagram.
The determining, in response to determining that the internal structure map includes a region characterizing a defect of the target pipe, a category of the defect and a location of the region in the internal structure map, comprising: inputting the internal structure diagram into a pre-trained defect detection model to obtain the type of the defect indicated by the region representing the defect of the target pipeline and the position information representing the position of the region representing the defect of the target pipeline in the internal structure diagram, wherein the defect detection model is used for indicating the correspondence between the type of the defect indicated by the internal structure diagram and the region representing the defect of the target pipeline and the position of the region representing the defect of the target pipeline in the internal structure diagram.
The defect detection model is obtained through training the following steps: obtaining a sample set, wherein the sample comprises a sample internal structure diagram and sample labeling information, the sample internal structure diagram comprises at least one region representing that a target pipeline has defects, and the sample labeling information is used for indicating the types of the defects indicated by the region representing that the target pipeline has the defects and the positions of the region indicating that the target pipeline has the defects in the sample internal structure diagram; and selecting samples from the sample set, respectively taking a sample internal structure diagram and sample labeling information of the selected samples as input and expected output of an initial model, and training to obtain a defect detection model.
The initial model comprises a deep convolutional neural network and a classifier, wherein the deep convolutional neural network is used for extracting the characteristics of an internal structure diagram.
The generating unit is further configured to: and performing image interpolation based on the electric signal data after the preset operation is performed, and generating a corresponding internal structure diagram.
The determining unit is further configured to: inputting the internal structure diagram into a pre-trained defect detection model to obtain the type of the defect indicated by the region representing the defect of the target pipeline and the position information representing the position of the region representing the defect of the target pipeline in the internal structure diagram, wherein the defect detection model is used for indicating the correspondence between the type of the defect indicated by the internal structure diagram and the region representing the defect of the target pipeline and the position of the region representing the defect of the target pipeline in the internal structure diagram.
The determination unit includes: an acquisition unit configured to acquire a sample set, the sample including a sample internal structure map including at least one region characterizing a defect of the target pipe, and sample annotation information indicating a type of defect indicated by the region characterizing the defect of the target pipe and a position of the region indicating the defect of the target pipe in the sample internal structure map; the training unit is configured to select samples from the sample set, respectively taking a sample internal structure diagram and sample labeling information of the selected samples as input and expected output of an initial model, and training to obtain a defect detection model.
The initial model comprises a deep convolutional neural network and a classifier, wherein the deep convolutional neural network is used for extracting the characteristics of an internal structure diagram.
Claims (9)
1. A method for automatically marking a pipe defect, by means of a device for detecting a pipe defect, the device comprising an acquisition unit configured to acquire electrical signal data for a target pipe acquired by at least one target sensor; an execution unit configured to perform a preset operation on the electrical signal data, wherein the preset operation includes filtering; a generation unit configured to generate a corresponding internal structure diagram based on the electrical signal data after performing the preset operation; a determining unit configured to determine a category of the defect and a location of the region in the internal structure map in response to determining that the internal structure map includes a region characterizing the target pipe as having the defect, comprising the steps of: acquiring electric signal data which are acquired by at least one target sensor and are aimed at a target pipeline; and executing preset operation on the electric signal data, generating a corresponding internal structure diagram based on the electric signal data after executing the preset operation, sending the obtained internal structure diagram to a server, obtaining the internal structure diagram and the design diagram of the pipeline from the server through third party equipment for comparison, and marking different places from the design diagram and in the design diagram to obtain an area of the internal structure diagram including the defect representing the target pipeline.
2. The method of claim 1, wherein the generating a corresponding internal structure map based on the electrical signal data after performing the preset operation comprises: and performing image interpolation based on the electric signal data after the preset operation is performed, and generating a corresponding internal structure diagram.
3. The method of claim 1 or 2, wherein the determining, in response to determining that the internal structure map includes characterizing an area of the target pipe in which a defect exists, a category of the defect and a location of the area in the internal structure map includes: inputting the internal structure diagram into a pre-trained defect detection model to obtain the type of the defect indicated by the region representing the defect of the target pipeline and the position information representing the position of the region representing the defect of the target pipeline in the internal structure diagram, wherein the defect detection model is used for indicating the correspondence between the type of the defect indicated by the internal structure diagram and the region representing the defect of the target pipeline and the position of the region representing the defect of the target pipeline in the internal structure diagram.
4. A method according to claim 3, wherein the defect detection model is trained by: obtaining a sample set, wherein the sample comprises a sample internal structure diagram and sample labeling information, the sample internal structure diagram comprises at least one region representing that a target pipeline has defects, and the sample labeling information is used for indicating the types of the defects indicated by the region representing that the target pipeline has the defects and the positions of the region indicating that the target pipeline has the defects in the sample internal structure diagram; and selecting samples from the sample set, respectively taking a sample internal structure diagram and sample labeling information of the selected samples as input and expected output of an initial model, and training to obtain a defect detection model.
5. The method of claim 4, wherein the initial model includes a deep convolutional neural network and a classifier, the deep convolutional neural network to extract features of an internal structure map.
6. The method of any of claims 1-5, wherein the generating unit is further configured to: and performing image interpolation based on the electric signal data after the preset operation is performed, and generating a corresponding internal structure diagram.
7. The method of claim 6, wherein the determination unit is further configured to: inputting the internal structure diagram into a pre-trained defect detection model to obtain the type of the defect indicated by the region representing the defect of the target pipeline and the position information representing the position of the region representing the defect of the target pipeline in the internal structure diagram, wherein the defect detection model is used for indicating the correspondence between the type of the defect indicated by the internal structure diagram and the region representing the defect of the target pipeline and the position of the region representing the defect of the target pipeline in the internal structure diagram.
8. The method of claim 7, wherein the determining unit comprises: an acquisition unit configured to acquire a sample set, the sample including a sample internal structure map including at least one region characterizing a defect of the target pipe, and sample annotation information indicating a type of defect indicated by the region characterizing the defect of the target pipe and a position of the region indicating the defect of the target pipe in the sample internal structure map; the training unit is configured to select samples from the sample set, respectively taking a sample internal structure diagram and sample labeling information of the selected samples as input and expected output of an initial model, and training to obtain a defect detection model.
9. The method of claim 8, wherein the initial model includes a deep convolutional neural network and a classifier, the deep convolutional neural network to extract features of an internal structure map.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111653019.2A CN116433556A (en) | 2021-12-31 | 2021-12-31 | Automatic labeling method for pipeline defects |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111653019.2A CN116433556A (en) | 2021-12-31 | 2021-12-31 | Automatic labeling method for pipeline defects |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116433556A true CN116433556A (en) | 2023-07-14 |
Family
ID=87089514
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111653019.2A Pending CN116433556A (en) | 2021-12-31 | 2021-12-31 | Automatic labeling method for pipeline defects |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116433556A (en) |
-
2021
- 2021-12-31 CN CN202111653019.2A patent/CN116433556A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111242123B (en) | Power equipment fault diagnosis method based on infrared image | |
| CN110411580B (en) | Diagnosis method and diagnosis system for heating defects of electric equipment | |
| CN105158678A (en) | Printed circuit board short-circuit fault rapid detection device | |
| CN110910383A (en) | Phase modulator fault monitoring method and system based on infrared image recognition | |
| CN101676699A (en) | Multipoint temperature acquisition method | |
| CN110648001A (en) | Inspection method and system for rail transit signal system | |
| CN114460439A (en) | Digital integrated circuit test system | |
| CN114255784A (en) | Substation equipment fault diagnosis method based on voiceprint recognition and related device | |
| CN115932530A (en) | Method for calibrating semiconductor detection equipment | |
| US11755007B2 (en) | System and method for determining a health condition and an anomaly of an equipment using one or more sensors | |
| CN111309791A (en) | Automatic data acquisition method for detection instrument | |
| CN112183555B (en) | Method and system for detecting welding quality, electronic device and storage medium | |
| US11520793B2 (en) | Servicing, diagnosing, logging, and surveying building equipment assets | |
| CN112215517A (en) | Intelligent auxiliary system and method for regular inspection of pressure pipeline | |
| CN116433556A (en) | Automatic labeling method for pipeline defects | |
| CN116662800A (en) | Rolling bearing fault diagnosis method based on self-adaptive attention mechanism | |
| CN114895163A (en) | Cable inspection positioning device and method based on cable insulation performance | |
| JP2008064643A (en) | Pressure transmitter inspecting system for plant | |
| JPH06121474A (en) | Patrol and inspection support device for electric station | |
| CN109975686B (en) | Circuit board short circuit automatic identification method based on infrared image processing | |
| CN113063843A (en) | Pipeline defect identification method and device and storage medium | |
| CN106501360A (en) | Based on the porcelain insulator defect detecting device from comparative trend analysis and vibroacoustics, system and method | |
| CN108763063B (en) | Software defect detection method without defect labeling data | |
| CN110489604A (en) | A kind of analytic method and system for gas turbin test measurement data | |
| JP6793598B2 (en) | Inspection support equipment, inspection support methods, and inspection support programs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |