CN113379728A - Method, system, equipment and readable storage medium for detecting defects on surface of rail - Google Patents
Method, system, equipment and readable storage medium for detecting defects on surface of rail Download PDFInfo
- Publication number
- CN113379728A CN113379728A CN202110749707.2A CN202110749707A CN113379728A CN 113379728 A CN113379728 A CN 113379728A CN 202110749707 A CN202110749707 A CN 202110749707A CN 113379728 A CN113379728 A CN 113379728A
- Authority
- CN
- China
- Prior art keywords
- rail
- picture
- defect
- real
- time
- 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 203
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000012549 training Methods 0.000 claims abstract description 35
- 238000003709 image segmentation Methods 0.000 claims abstract description 31
- 238000001514 detection method Methods 0.000 claims abstract description 27
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 10
- 238000004590 computer program Methods 0.000 claims description 10
- 238000005286 illumination Methods 0.000 claims description 10
- 238000013135 deep learning Methods 0.000 abstract description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000006243 chemical reaction Methods 0.000 description 5
- 230000006835 compression Effects 0.000 description 4
- 238000007906 compression Methods 0.000 description 4
- 238000007689 inspection Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- 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
- G06N3/04—Architecture, e.g. interconnection topology
-
- 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
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- 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
- G06T2207/10052—Images from lightfield camera
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/30236—Traffic on road, railway or crossing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Geometry (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a defect detection method, a system and equipment for a rail surface and a readable storage medium, wherein the defect detection method comprises the following steps: marking a defect area of the historical rail picture; inputting the marked historical rail picture as training data into a U-net network algorithm, and outputting a defect image segmentation model; the defect image segmentation model takes a rail picture as input and takes a defect picture in the rail picture as output; acquiring a real-time rail picture, and inputting a defect image segmentation model; outputting a current defect picture in the real-time rail picture; and calculating the position of the defect area in the real-time rail picture according to the current defect picture. The method comprises the steps of training by using a rail surface defect data set, adopting a deep learning image segmentation method and using a U-net-based model, recognizing the rail surface defects, detecting whether the rail surface has defects, defect positions and defect areas, and further timely judging whether the rail needs to be maintained.
Description
Technical Field
The invention belongs to the field of machine vision detection, and particularly relates to a method, a system and equipment for detecting defects on a rail surface and a readable storage medium.
Background
The safety problem of railway transportation is an important aspect, and rail surface safety is an important component in the safe operation of trains. The defect on the surface of the rail can influence the normal use of the rail, and is a hidden danger threatening the safe running of a train, moreover, the rail utilization rate is high, the rail needs to be regularly inspected, and the damaged part of the rail is maintained, so that the inspection on the surface of the rail is an essential work for maintaining the safety of railway transportation.
Traditional rail surface is patrolled and examined mainly by the manual work, mainly relies on the staff to follow the railway line and visualizes and hold the detection instrument, follows the mode inspection that the railway line walked. However, the manual inspection has many defects, and although the surface defect problem of the track can be effectively detected, the spot inspection rate is low, the accuracy rate is low, the efficiency is low, the labor intensity is high, and the working environment is severe.
Disclosure of Invention
The present invention provides a method, a system, a device and a readable storage medium for detecting defects on a rail surface, so as to overcome the above-mentioned defects in the prior art.
The invention solves the technical problems through the following technical scheme:
a method of defect detection of a rail surface, the method comprising:
marking a defect area of the historical rail picture;
inputting the marked historical rail picture as training data into a U-net network algorithm (an image segmentation algorithm), and outputting a defect image segmentation model; the defect image segmentation model takes a rail picture as input and takes a defect picture in the rail picture as output;
acquiring a real-time rail picture, and inputting the defect image segmentation model;
outputting a current defect picture in the real-time rail picture;
and calculating the position of the defect area in the real-time rail picture according to the current defect picture.
Preferably, before the step of using the marked historical rail picture as the training data, the defect detection method further includes:
and converting the marked historical rail picture from a single-channel image into a three-channel image.
Preferably, the step of calculating the position of the defect area in the real-time rail image according to the current defect image specifically includes:
calculating the mass center value of the defect area in the real-time rail picture by utilizing a first moment algorithm;
and determining the position of the defect region according to the centroid value.
Preferably, after the step of calculating the position of the defect area in the real-time rail image according to the current defect image, the defect detection method further includes:
and storing the current defect picture according to the centroid value based on the run length code.
Preferably, after the step of calculating the position of the defect area in the real-time rail image according to the current defect image, the defect detection method further includes:
acquiring actual size data of the rail to which the real-time rail picture belongs;
and obtaining the actual area of the defect region according to the position of the defect region and the actual size data.
Preferably, the historical rail pictures include pictures under different illumination intensities, and the step of using the marked historical rail pictures as training data specifically includes:
and taking the marked historical rail pictures under different illumination intensities as the training data.
Preferably, before the step of outputting the current defect picture in the real-time rail pictures, the defect detection method further includes:
and detecting whether the real-time rail picture contains a defect area, if so, executing the step of outputting the current defect picture in the real-time rail picture.
A defect detection system for a rail surface, the defect detection system comprising:
the marking module is used for marking the defect area of the historical rail picture;
the training module is used for inputting the marked historical rail picture as training data into the U-net network model and outputting a defect image segmentation model; the defect image segmentation model takes a rail picture as input and takes a defect picture in the rail picture as output;
the real-time picture acquisition module is used for acquiring a real-time rail picture and inputting the defect image segmentation model;
the defect picture output module is used for outputting a current defect picture in the real-time rail pictures;
and the calculating module is used for calculating the position of the defect area in the real-time rail picture according to the current defect picture.
Preferably, the detection system further comprises:
and the channel conversion module is used for converting the marked historical rail picture into a three-channel image from a single-channel image.
Preferably, the calculation module comprises:
the mass center calculating unit is used for calculating the mass center value of the defect area in the real-time rail picture by utilizing a first moment algorithm;
and the position determining unit is used for determining the position of the defect area according to the centroid value.
Preferably, the defect detection system further comprises:
and the storage module is used for storing the current defect picture according to the centroid value based on the run length code.
Preferably, the defect detection system further comprises:
the rail size acquisition module is used for acquiring the actual size data of the rail of the real-time rail picture;
and the actual defect area determining module is used for obtaining the actual area of the defect area according to the position of the defect area and the actual size data.
Preferably, the historical rail pictures comprise pictures under different illumination intensities;
the training module is used for taking the marked historical rail pictures under different illumination intensities as the training data.
Preferably, the defect detection system further comprises:
and the detection module is used for detecting whether the real-time rail picture contains a defect area, and if so, the real-time picture acquisition module is called.
An electronic device comprises a memory, a processor and a computer program stored on the memory and operable on the processor, wherein the processor executes the computer program to implement the method for detecting the defect on the surface of the rail.
A computer-readable storage medium, on which a computer program is stored, wherein the program, when executed by a processor, implements the method for detecting defects on a rail surface as described above.
The positive progress effects of the invention are as follows: the method uses the rail surface defect data set, adopts a deep learning image segmentation method, and trains by using a U-net frame model, so that the rail surface defects are identified, and whether the rail surface has defects and defect positions are detected, so that the defect area is determined, and further, the method is favorable for timely judging whether the rail needs to be maintained in practical application.
Drawings
Fig. 1 is a flowchart of a method for detecting defects on a surface of a rail according to embodiment 1 of the present invention.
Fig. 2 is a flowchart of step 50 of the method for detecting defects on a rail surface according to embodiment 1 of the present invention.
Fig. 3 is a schematic diagram of processing an image by a defect image segmentation model in the method for detecting a defect on a rail surface according to embodiment 1 of the present invention.
Fig. 4 is a block diagram of a rail surface defect detection system according to embodiment 2 of the present invention.
Fig. 5 is a schematic structural diagram of an electronic device according to embodiment 3 of the present invention.
Detailed Description
The invention is further illustrated by the following examples, which are not intended to limit the scope of the invention.
Example 1
A method of detecting defects on a surface of a rail, as shown in fig. 1, the method comprising:
wherein a rail surface defect data set may be downloaded from an open source RSDDs dataset, each image of the data set having surface defects and containing a plurality of different defect shapes. The data volume in the training data set may be more or less, and there is no requirement for the size.
the rail surface defect data set is randomly divided into a training set and a verification set, and the training data is input into a U-net model for training. In addition, the method for detecting the rail surface defects by using the deep learning image segmentation has higher robustness and adaptability, and meanwhile, the U-net model occupies smaller memory resources and has lower requirements on computing resources.
In this embodiment, since the historical rail pictures include pictures at different illumination intensities, in step 20, it is actually implied that the marked historical rail pictures at different illumination intensities are used as training data, so that the model obtained by training can adapt to the interference of the on-site ambient light.
And 50, calculating to obtain the position of the defect area in the real-time rail picture according to the current defect picture.
In this embodiment, as shown in fig. 2, step 50 specifically includes:
and 502, determining the position of the defect area according to the centroid value.
The divided picture frames a defect area range according to pixel values, then a Hu first-order matrix value of the image in the area is calculated, and the position of the area is calculated as follows: the position of the defect area in the picture is represented by the sum of the pixels in the coordinate direction in the area. And calculating the centroid by using the Hu first moment as an expression of the position of the defect area, wherein the centroid has translation invariance.
In this embodiment, referring to fig. 1, after step 50, the defect detection method further includes:
and step 51, storing the current defect picture according to the centroid value based on the run length code.
Wherein, run-length coding is an image compression method, and if the image blocks with the same color in an image are larger, the number of the image blocks is smaller, and the obtained compression ratio is higher. The segmented defect image is stored by the method so as to reduce the storage space. A run length encoding step: the first byte is the start position and the second byte is the length, with pixels numbered from top to bottom and then from left to right. For example. '13' indicates that starting from pixel 1, the total number of pixels is 3, and after storing, a series of codes is formed, and '14105' indicates the 1 st, 2 nd, 3 rd, 4 th, 10 th, 11 th, 12 th, 13 th and 14 th pixels, and the pixel point of the defective picture is recorded in the mode of coding.
and step 53, obtaining the actual area of the defect area according to the position and the actual size data of the defect area.
The method comprises the steps of converting an image space into a physical space, calculating the number of pixels of the rail width in a picture after acquiring the actual size of a rail, and comparing the number of pixels with the number of the pixels with defects in the image to calculate the actual area of the defects. The prior knowledge of the physical size of the rail is used for converting the defect area in the picture into the actual area in the physical space, so that the coordinate system conversion error caused by introducing a camera coordinate system is avoided.
In this embodiment, referring to fig. 1, before step 20, the defect detection method further includes:
and 11, converting the marked historical rail picture from a single-channel image into a three-channel image.
Among them, it is preferable to use 3-channel images in consideration of the input of the model, and since the pictures in the data set are generally single-channel, the channels of the pictures need to be converted.
Before step 40, the defect detection method further includes:
and 31, detecting whether the real-time rail picture contains a defect area, and if so, executing a step 40.
The defect image segmentation model is used for detection, whether a defect area exists can be directly obtained, if the defect area does not exist, and if the defect area does exist, the subsequent determination of the defect area is further executed.
In this embodiment, a rail surface defect data set is used, a deep learning image segmentation method is adopted, and a U-net frame model is used for training, so that the rail surface defects are identified, and whether the rail surface has defects and defect positions are detected, so that the defect area is determined, and further, the rail surface defect data set is beneficial to timely judging whether maintenance is needed in practical application.
Example 2
A defect detection system for a rail surface is shown in FIG. 4, and comprises a marking module 1, a training module 2, a real-time picture acquisition module 3, a defect picture output module 4 and a calculation module 5;
the marking module 1 is used for marking the defect area of the historical rail picture;
wherein a rail surface defect data set may be downloaded from an open source RSDDs dataset, each image of the data set having surface defects and containing a plurality of different defect shapes. The data volume in the training data set may be more or less, and there is no requirement for the size.
The training module 2 is used for inputting the marked historical rail picture as training data into a U-net network model and outputting a defect image segmentation model; the defect image segmentation model takes a rail picture as input and takes a defect picture in the rail picture as output;
the rail surface defect data set is randomly divided into a training set and a verification set, and the training data is input into a U-net model for training. In addition, the method for detecting the rail surface defects by using the deep learning image segmentation has higher robustness and adaptability, and meanwhile, the U-net model occupies smaller memory resources and has lower requirements on computing resources.
In this embodiment, since the historical rail pictures include pictures at different illumination intensities, the training module 2 is used to actually imply that the marked historical rail pictures at different illumination intensities are used as training data, so that the model obtained by training can adapt to the interference of the on-site ambient light.
The real-time picture acquisition module 3 is used for acquiring a real-time rail picture and inputting the defect image segmentation model;
the defect picture output module 4 is used for outputting a current defect picture in the real-time rail pictures;
and the calculating module 5 is used for calculating the position of the defect area in the real-time rail picture according to the current defect picture.
In this embodiment, referring to fig. 4, the calculating module 5 specifically includes:
the mass center calculating unit 5-1 is used for calculating the mass center value of the defect area in the real-time rail picture by utilizing a first moment algorithm;
a position determining unit 5-2 for determining the position of the defect region according to the centroid value.
The divided picture frames a defect area range according to pixel values, then a Hu first-order matrix value of the image in the area is calculated, and the position of the area is calculated as follows: the position of the defect area in the picture is represented by the sum of the pixels in the coordinate direction in the area. And calculating the centroid by using the Hu first moment as an expression of the position of the defect area, wherein the centroid has translation invariance.
In this embodiment, the defect detection system further includes a storage module 6, a rail size obtaining module 7, and a defect actual area determining module 8;
and the storage module 6 is used for storing the current defect picture according to the centroid value based on the run length code.
Wherein, run-length coding is an image compression method, and if the image blocks with the same color in an image are larger, the number of the image blocks is smaller, and the obtained compression ratio is higher. The segmented defect image is stored by the method so as to reduce the storage space. A run length encoding step: the first byte is the start position and the second byte is the length, with pixels numbered from top to bottom and then from left to right. For example. '13' indicates that starting from pixel 1, the total number of pixels is 3, and after storing, a series of codes is formed, and '14105' indicates the 1 st, 2 nd, 3 rd, 4 th, 10 th, 11 th, 12 th, 13 th and 14 th pixels, and the pixel point of the defective picture is recorded in the mode of coding. The rail size obtaining module 7 is used for obtaining the actual size data of the rail of the real-time rail picture;
and the actual defect area determining module 8 is configured to obtain an actual area of the defect area according to the position of the defect area and the actual size data.
The method comprises the steps of converting an image space into a physical space, calculating the number of pixels of the rail width in a picture after acquiring the actual size of a rail, and comparing the number of pixels with the number of the pixels with defects in the image to calculate the actual area of the defects. The prior knowledge of the physical size of the rail is used for converting the defect area in the picture into the actual area in the physical space, so that the coordinate system conversion error caused by introducing a camera coordinate system is avoided.
In this embodiment, referring to fig. 4, the detection system further includes a channel conversion module 9 and a detection module 10-1:
the channel conversion module 9 is used for converting the marked historical rail picture from a single-channel image into a three-channel image.
The detection module 10-1 is configured to detect whether the real-time rail picture includes a defect area, and if so, invoke the real-time picture acquisition module 3.
In this embodiment, a rail surface defect data set is used, a deep learning image segmentation method is adopted, and a U-net frame model is used for training, so that the rail surface defects are identified, and whether the rail surface has defects and defect positions are detected, so that the defect area is determined, and further, the rail surface defect data set is beneficial to timely judging whether maintenance is needed in practical application.
Example 3
An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the method of detecting defects on a rail surface of embodiment 1 when executing the computer program.
Fig. 5 is a schematic structural diagram of an electronic device provided in this embodiment. Fig. 5 illustrates a block diagram of an exemplary electronic device 90 suitable for use in implementing embodiments of the present invention. The electronic device 90 shown in fig. 5 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiment of the present invention.
As shown in fig. 5, the electronic device 90 may take the form of a general purpose computing device, which may be a server device, for example. The components of the electronic device 90 may include, but are not limited to: at least one processor 91, at least one memory 92, and a bus 93 that connects the various system components (including the memory 92 and the processor 91).
The bus 93 includes a data bus, an address bus, and a control bus.
The processor 91 executes various functional applications and data processing by running a computer program stored in the memory 92.
The electronic device 90 may also communicate with one or more external devices 94 (e.g., keyboard, pointing device, etc.). Such communication may be through an input/output (I/O) interface 95. Also, the electronic device 90 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet) via a network adapter 96. The network adapter 96 communicates with the other modules of the electronic device 90 via the bus 93. It should be appreciated that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the electronic device 90, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems, etc.
It should be noted that although in the above detailed description several units/modules or sub-units/modules of the electronic device are mentioned, such a division is merely exemplary and not mandatory. Indeed, the features and functionality of two or more of the units/modules described above may be embodied in one unit/module, according to embodiments of the application. Conversely, the features and functions of one unit/module described above may be further divided into embodiments by a plurality of units/modules.
Example 4
A computer-readable storage medium on which a computer program is stored, which when executed by a processor, implements the method of detecting defects on a rail surface of embodiment 1.
More specific examples, among others, that the readable storage medium may employ may include, but are not limited to: a portable disk, a hard disk, random access memory, read only memory, erasable programmable read only memory, optical storage device, magnetic storage device, or any suitable combination of the foregoing.
In a possible implementation, the invention may also be implemented in the form of a program product comprising program code for causing a terminal device to perform a method for detecting defects on a rail surface as described in example 1, when the program product is run on the terminal device.
Where program code for carrying out the invention is written in any combination of one or more programming languages, the program code may be executed entirely on the user device, partly on the user device, as a stand-alone software package, partly on the user device and partly on a remote device or entirely on the remote device.
While specific embodiments of the invention have been described above, it will be appreciated by those skilled in the art that this is by way of example only, and that the scope of the invention is defined by the appended claims. Various changes and modifications to these embodiments may be made by those skilled in the art without departing from the spirit and scope of the invention, and these changes and modifications are within the scope of the invention.
Claims (10)
1. A method of detecting defects on a surface of a rail, the method comprising:
marking a defect area of the historical rail picture;
inputting the marked historical rail picture as training data into a U-net network algorithm, and outputting a defect image segmentation model; the defect image segmentation model takes a rail picture as input and takes a defect picture in the rail picture as output;
acquiring a real-time rail picture, and inputting the defect image segmentation model;
outputting a current defect picture in the real-time rail picture;
and calculating the position of the defect area in the real-time rail picture according to the current defect picture.
2. The method of claim 1, wherein prior to the step of using the marked historical rail image as training data, the method further comprises:
and converting the marked historical rail picture from a single-channel image into a three-channel image.
3. The method according to claim 1, wherein the step of calculating the position of the defect area in the real-time rail image according to the current defect image comprises:
calculating the mass center value of the defect area in the real-time rail picture by utilizing a first moment algorithm;
and determining the position of the defect region according to the centroid value.
4. A method of detecting defects on a rail surface according to claim 3, wherein after the step of calculating the location of the defect area in the real-time rail picture from the current defect picture, the method further comprises:
and storing the current defect picture according to the centroid value based on the run length code.
5. The method of claim 1, wherein after the step of calculating the location of the defect area in the real-time rail image based on the current defect image, the method further comprises:
acquiring actual size data of the rail to which the real-time rail picture belongs;
and obtaining the actual area of the defect region according to the position of the defect region and the actual size data.
6. The method according to claim 1, wherein the historical rail images include images under different illumination intensities, and the step of using the marked historical rail images as training data includes:
and taking the marked historical rail pictures under different illumination intensities as the training data.
7. The method of claim 1, wherein prior to the step of outputting the current defect picture of the real-time rail pictures, the method further comprises:
and detecting whether the real-time rail picture contains a defect area, if so, executing the step of outputting the current defect picture in the real-time rail picture.
8. A rail surface defect detection system, comprising:
the marking module is used for marking the defect area of the historical rail picture;
the training module is used for inputting the marked historical rail picture as training data into the U-net network model and outputting a defect image segmentation model; the defect image segmentation model takes a rail picture as input and takes a defect picture in the rail picture as output;
the real-time picture acquisition module is used for acquiring a real-time rail picture and inputting the defect image segmentation model;
the defect picture output module is used for outputting a current defect picture in the real-time rail pictures;
and the calculating module is used for calculating the position of the defect area in the real-time rail picture according to the current defect picture.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of detecting defects on a rail surface according to any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium on which a computer program is stored, the program, when executed by a processor, implementing a method of detecting defects on a rail surface according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110749707.2A CN113379728A (en) | 2021-07-02 | 2021-07-02 | Method, system, equipment and readable storage medium for detecting defects on surface of rail |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110749707.2A CN113379728A (en) | 2021-07-02 | 2021-07-02 | Method, system, equipment and readable storage medium for detecting defects on surface of rail |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113379728A true CN113379728A (en) | 2021-09-10 |
Family
ID=77580651
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110749707.2A Pending CN113379728A (en) | 2021-07-02 | 2021-07-02 | Method, system, equipment and readable storage medium for detecting defects on surface of rail |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113379728A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114166855A (en) * | 2022-02-11 | 2022-03-11 | 南京邮电大学 | Real-time rail defect detection method |
| CN114529507A (en) * | 2021-12-30 | 2022-05-24 | 广西慧云信息技术有限公司 | Shaving board surface defect detection method based on visual transducer |
| CN114677342A (en) * | 2022-03-14 | 2022-06-28 | 西南交通大学 | Unsupervised defect detection method |
| CN114721369A (en) * | 2022-02-28 | 2022-07-08 | 上海电气集团股份有限公司 | Intelligent inspection system applied to rigid landfill |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101750422A (en) * | 2010-01-07 | 2010-06-23 | 秦皇岛凯维科技有限公司 | On-line automatic detection device for glass defect |
| CN102854191A (en) * | 2012-07-18 | 2013-01-02 | 湖南大学 | Real-time visual detection and identification method for high speed rail surface defect |
| CN104931585A (en) * | 2015-05-29 | 2015-09-23 | 湖北三江航天江北机械工程有限公司 | Composite material debonding defect ultrasonic C-scan detection area assessment method |
| CN107564002A (en) * | 2017-09-14 | 2018-01-09 | 广东工业大学 | Plastic tube detection method of surface flaw, system and computer-readable recording medium |
| CN109886950A (en) * | 2019-02-22 | 2019-06-14 | 北京百度网讯科技有限公司 | The defect inspection method and device of circuit board |
| CN110211101A (en) * | 2019-05-22 | 2019-09-06 | 武汉理工大学 | A kind of rail surface defect rapid detection system and method |
| CN111402209A (en) * | 2020-03-03 | 2020-07-10 | 广州中国科学院先进技术研究所 | U-Net-based high-speed railway steel rail damage detection method |
| CN111583223A (en) * | 2020-05-07 | 2020-08-25 | 上海闻泰信息技术有限公司 | Defect detection method, defect detection device, computer equipment 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 |
| CN112070135A (en) * | 2020-08-28 | 2020-12-11 | 广东电网有限责任公司 | Power equipment image detection method and device, power equipment and storage medium |
| CN112767369A (en) * | 2021-01-25 | 2021-05-07 | 佛山科学技术学院 | Defect identification and detection method and device for small hardware and computer readable storage medium |
| CN112967272A (en) * | 2021-03-25 | 2021-06-15 | 郑州大学 | Welding defect detection method and device based on improved U-net and terminal equipment |
-
2021
- 2021-07-02 CN CN202110749707.2A patent/CN113379728A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101750422A (en) * | 2010-01-07 | 2010-06-23 | 秦皇岛凯维科技有限公司 | On-line automatic detection device for glass defect |
| CN102854191A (en) * | 2012-07-18 | 2013-01-02 | 湖南大学 | Real-time visual detection and identification method for high speed rail surface defect |
| CN104931585A (en) * | 2015-05-29 | 2015-09-23 | 湖北三江航天江北机械工程有限公司 | Composite material debonding defect ultrasonic C-scan detection area assessment method |
| CN107564002A (en) * | 2017-09-14 | 2018-01-09 | 广东工业大学 | Plastic tube detection method of surface flaw, system and computer-readable recording medium |
| CN109886950A (en) * | 2019-02-22 | 2019-06-14 | 北京百度网讯科技有限公司 | The defect inspection method and device of circuit board |
| CN110211101A (en) * | 2019-05-22 | 2019-09-06 | 武汉理工大学 | A kind of rail surface defect rapid detection system and method |
| CN111402209A (en) * | 2020-03-03 | 2020-07-10 | 广州中国科学院先进技术研究所 | U-Net-based high-speed railway steel rail damage detection method |
| CN111583223A (en) * | 2020-05-07 | 2020-08-25 | 上海闻泰信息技术有限公司 | Defect detection method, defect detection device, computer equipment 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 |
| CN112070135A (en) * | 2020-08-28 | 2020-12-11 | 广东电网有限责任公司 | Power equipment image detection method and device, power equipment and storage medium |
| CN112767369A (en) * | 2021-01-25 | 2021-05-07 | 佛山科学技术学院 | Defect identification and detection method and device for small hardware and computer readable storage medium |
| CN112967272A (en) * | 2021-03-25 | 2021-06-15 | 郑州大学 | Welding defect detection method and device based on improved U-net and terminal equipment |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114529507A (en) * | 2021-12-30 | 2022-05-24 | 广西慧云信息技术有限公司 | Shaving board surface defect detection method based on visual transducer |
| CN114529507B (en) * | 2021-12-30 | 2024-05-17 | 广西慧云信息技术有限公司 | Visual transducer-based particle board surface defect detection method |
| CN114166855A (en) * | 2022-02-11 | 2022-03-11 | 南京邮电大学 | Real-time rail defect detection method |
| CN114721369A (en) * | 2022-02-28 | 2022-07-08 | 上海电气集团股份有限公司 | Intelligent inspection system applied to rigid landfill |
| CN114677342A (en) * | 2022-03-14 | 2022-06-28 | 西南交通大学 | Unsupervised defect detection method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113379728A (en) | Method, system, equipment and readable storage medium for detecting defects on surface of rail | |
| CN111402209B (en) | U-Net-based high-speed railway steel rail damage detection method | |
| CN111047609B (en) | Pneumonia focus segmentation method and device | |
| CN112418216B (en) | Text detection method in complex natural scene image | |
| CN112070135A (en) | Power equipment image detection method and device, power equipment and storage medium | |
| US20210383526A1 (en) | Method for training defect detector | |
| CN109934873B (en) | Method, device and equipment for acquiring marked image | |
| CN116721104B (en) | Live three-dimensional model defect detection method and device, electronic equipment and storage medium | |
| CN110689000A (en) | Vehicle license plate identification method based on vehicle license plate sample in complex environment | |
| CN110796078A (en) | Vehicle light detection method and device, electronic equipment and readable storage medium | |
| CN115331086B (en) | Brake shoe breakage and rivet loss fault detection method | |
| CN110807416A (en) | Digital instrument intelligent recognition device and method suitable for mobile detection device | |
| CN112686887A (en) | Method, system, equipment and medium for detecting concrete surface cracks | |
| CN114529821A (en) | Offshore wind power safety monitoring and early warning method based on machine vision | |
| CN110322442A (en) | A kind of building surface crack detecting method based on SegNet | |
| Han et al. | SSGD: A smartphone screen glass dataset for defect detection | |
| CN118037678A (en) | Industrial surface defect detection method and device based on improved variation self-encoder | |
| JP7410323B2 (en) | Abnormality detection device, abnormality detection method and abnormality detection system | |
| CN112686888A (en) | Method, system, equipment and medium for detecting cracks of concrete sleeper | |
| CN115527207B (en) | Train brake adjuster control rod nut fault detection method based on deep neural network | |
| CN117351505A (en) | Information code identification method, device, equipment and storage medium | |
| KR20210115121A (en) | Deeep learning-based disaster safety building custom 3D modeling dataset construction method | |
| CN112541436B (en) | Concentration analysis method and device, electronic equipment and computer storage medium | |
| CN114821806A (en) | Method and device for determining behavior of operator, electronic equipment and storage medium | |
| CN111696154B (en) | Coordinate positioning method, device, 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 |