CN111080602B - Method for detecting foreign matters in water leakage hole of railway wagon - Google Patents
Method for detecting foreign matters in water leakage hole of railway wagon Download PDFInfo
- Publication number
- CN111080602B CN111080602B CN201911272274.5A CN201911272274A CN111080602B CN 111080602 B CN111080602 B CN 111080602B CN 201911272274 A CN201911272274 A CN 201911272274A CN 111080602 B CN111080602 B CN 111080602B
- Authority
- CN
- China
- Prior art keywords
- image
- water leakage
- layer
- leakage hole
- batch normalization
- 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
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N3/045—Combinations of networks
-
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
Abstract
A method for detecting foreign matters in a water leakage hole of a railway wagon relates to a method for detecting foreign matters in the water leakage hole, and belongs to the technical field of freight train detection. The invention aims to solve the problems of low efficiency and low accuracy of the existing detection of foreign matters in the water leakage holes. The method comprises the steps of collecting images, extracting the images containing the water leaking hole areas, constructing a sample data set, and training a target object segmentation network by using the sample data set; in the detection process, acquiring a real vehicle passing image and extracting a water leakage hole position image, and recording the water leakage hole position image as a water leakage hole position image to be detected; inputting the image of the water leakage hole part to be detected into a target object segmentation network to obtain a predicted binary image, finding a maximum foreign body communication domain if a foreign body is detected to exist in the binary image, calculating the area and height of the maximum foreign body communication domain, and performing fault alarm when the area of the foreign body is larger than an area threshold or the length is larger than a length threshold. The method is mainly used for detecting the foreign matters in the water leakage holes of the railway freight car.
Description
Technical Field
The invention relates to a method for detecting foreign matters in a water leakage hole. Belongs to the technical field of freight train detection.
Background
In order to ensure the running safety of the railway wagon, the railway wagon needs to be subjected to daily inspection operation. The method comprises the inspection of the water leakage holes of the railway freight car. The water leakage holes carried by the railway freight car are two holes at the lower part of the bogie of the railway freight car, and as shown in figure 2, two areas marked by white triangular frames in figure 2 are the water leakage holes of the bogie of the railway freight car. When the water leakage holes are in maintenance or the train runs, a plurality of foreign matters can enter, the water leakage holes are close to the train track below, and hard foreign matters can be partially clamped at the water leakage holes when the train runs, so that the track is damaged below.
For a long time, the inspection of the water leakage holes of the railway freight car adopts a manual image inspection mode to detect whether foreign matters are carried in the water leakage hole area, and car inspection personnel are extremely easy to fatigue in the working process, easily have the conditions of missing inspection and wrong inspection and are difficult to ensure the accuracy; and manual detection is very inefficient. Therefore, the fault detection of the truck has important significance in improving the detection efficiency and accuracy by adopting an automatic identification mode.
The image containing the water leakage hole part comprises the water leakage hole and a background area except the water leakage hole, and the background area can have a large amount of changes, so that detection is influenced, one is artificial interference of oil stains, chalks and the like, and the other is influence of foreign matters such as paper, plastics, snow and the like. The noise can be changed greatly in color, shape and position, such as white chalk writing, black plastic bags or gray grass branches, and the like, and is difficult to distinguish in the gray level image by using the traditional modes such as binarization operation, and the like, while the water leakage hole part is shot under the motion of different vehicle types and railway wagons to obtain an image, the shape and the position of the intercepted water leakage hole part in a large number of images are changed, and the traditional algorithm is difficult to adapt to the transformation in all gray level images, so that the robustness is poor.
Disclosure of Invention
The invention aims to solve the problems of low efficiency and low accuracy of the existing detection of foreign matters in the water leakage holes.
The method for detecting the foreign matters in the water leakage holes of the railway wagon comprises the following steps:
collecting an image, extracting the image containing a water leakage hole area, and constructing a sample data set, wherein the sample data set comprises two parts: a gray scale image set and a binary image set;
the gray level image set is a gray level image set containing a water leakage hole area; the binary image set is a binary segmentation image with foreign matters and without foreign matters, which is obtained by an artificial mark corresponding to the water leakage hole area image;
training a target object segmentation network by using a sample data set to obtain a trained target object segmentation network;
in the detection process, acquiring a real vehicle passing image and extracting a water leakage hole position image, and recording the water leakage hole position image as a water leakage hole position image to be detected;
inputting the image of the water leakage hole part to be detected into a target object segmentation network to obtain a predicted binary image, finding a maximum foreign body communication domain if a foreign body is detected to exist in the binary image, calculating the area and height of the maximum foreign body communication domain, and performing fault alarm when the area of the foreign body is larger than an area threshold or the length is larger than a length threshold; and in other cases, continuously processing the intercepted image of the next water leakage hole area.
Further, the target object segmentation network adopts an encoder-decoder network structure, and an encoder network model is as follows:
the encoder includes 6 coding units:
the encoding unit 1: 3 x 3 convolution layer → batch normalization layer → 2 x 2 pooling layer;
the encoding unit 2: 3 x 3 convolution layer → batch normalization layer → 2 x 2 pooling layer;
the encoding unit 3: [3 × 3 convolution layer → batch normalization layer ] × 3 → 2 × 2 pooling layer;
the encoding unit 4: [3 × 3 convolution layer → batch normalization layer ] × 3 → 2 × 2 pooling layer;
the encoding unit 5: [3 × 3 convolution layer → batch normalization layer ] × 3 → 2 × 2 pooling layer;
the encoding unit 6: 3 x 3 convolution layer → batch normalization layer → 3 x 3 convolution layer → batch normalization layer.
Further, the target object segmentation network adopts an encoder-decoder network structure, and a decoder network model is as follows:
the decoding unit 1: 2 × 2 upsampling layers → [1 × 1 convolutional layers → batch normalization layers ] × 1 → [3 × 3 convolutional layers → batch normalization layers ] × 2 of the coding units 5 and 6;
the decoding unit 2: 2 × 2 upsampling layers → [1 × 1 convolutional layers → batch normalization layers ] × 1 → [3 × 3 convolutional layers → batch normalization layers ] × 2 of the encoding unit 4 and the decoding unit 1;
the decoding unit 3: 2 × 2 upsampling layers of the decoding unit 2 → [1 × 1 convolutional layers → batch normalization layers ] × 1 → [3 × 3 convolutional layers → batch normalization layers ] × 2;
the decoding unit 4: the up-sampling layer of 4 × 4 of the decoding unit 3 → 1 × 1 convolution layer → batch normalization layer → 1 × 1 convolution layer.
Furthermore, the process of extracting the image containing the water leaking hole area is realized according to the axle distance information and the position prior knowledge of the water leaking hole area.
Further, the process of acquiring images and extracting images containing the water leaking hole areas is realized based on a line camera.
Further, data amplification operation is required in the process of constructing the sample data set; the data amplification form comprises random combination operations of vertical turning, scaling and translation on the image.
Has the advantages that:
1. the mode of utilizing image automatic identification replaces artifical the detection, can not only greatly reduce total amount of labour and every testing personnel's intensity of labour, can effectively improve the operating quality moreover, improves the stability and the precision that detect, and the rate of accuracy obtains very big improvement.
2. The invention adopts a semantic segmentation mode to detect the fault of the foreign matters carried on the water leakage holes, and is more favorable for solving the overall shape, especially the area information of the foreign matters compared with a detection algorithm. The overall scheme of the invention can improve the stability and precision of the overall algorithm, and has higher flexibility, accuracy and robustness compared with the traditional machine vision detection method of manual standard feature extraction.
Drawings
FIG. 1 is a schematic diagram of a foreign object detection process;
FIG. 2 is a schematic view of the position of the water leakage hole;
FIG. 3(a) is a schematic diagram of the gray scale of the rough cutting part of the water leakage hole, and FIG. 3(b) is a schematic diagram of a binary image;
FIG. 4 is a diagram of a deep learning network model;
fig. 5 is a flow chart of the determination of the foreign matter fault at the triangular hole.
Detailed Description
The first embodiment is as follows: the present embodiment is described in detail with reference to figure 1,
the embodiment provides a method for detecting foreign matters in a water leakage hole of a railway wagon, which comprises the following steps:
step1, line image collection
Shooting the running truck to obtain a high-definition gray-scale whole truck image.
Due to the influence of interference factors, the images of the water leaking hole areas in the obtained whole vehicle image have large difference. In order to enhance the robustness of the recognition algorithm, images under various conditions need to be obtained during the process of collecting image data.
According to the axle distance information of hardware, the position of the water leakage hole area and other prior knowledge, the image of the water leakage hole area is cut out from the high-definition gray-scale whole-vehicle image, so that the calculated amount can be reduced to a great extent, and the recognition speed of the scheme is improved.
As shown in fig. 3(a), the cut-out image of the water leakage hole portion is a rectangle including the water leakage hole and the background area excluding the water leakage hole. The traditional algorithm is difficult to adapt to the transformation in all gray level images, the robustness is poor, the foreign matter in the water leakage hole part in the image is detected by adopting the neural network, and the robustness and the accuracy are very good.
It can also be seen from fig. 3(a) that the environment in the water leakage hole image is complex and not easy to be detected and recognized;
the sample data set includes two parts: a grayscale image set and a binary image set (group try image), as shown in fig. 3(a) and 3 (b);
the gray level image set is a set of images of the water leakage hole area cut out;
the binary image set is a binary segmentation image with foreign matters and without foreign matters, which is obtained by an artificial mark corresponding to the water leakage hole area image; for example, a pixel value corresponding to a foreign object is non-zero, and a pixel value corresponding to no foreign object is 0;
the gray-scale image set and the binary image set are in one-to-one correspondence.
Data amplification of sample dataset: to further improve the robustness of the algorithm, data amplification of the sample data set is still required. The amplification form is mainly by performing random combination operations such as vertical turning, scaling, translation and the like on the image.
The target object segmentation network constructed by the invention adopts an encoder-decoder network structure, as shown in fig. 2.
The real-time performance requirement on detection in automatic truck identification detection is higher, so the method realizes the real-time performance and high precision of detection by reducing characteristic network parameters. Aiming at the robustness requirement of fault detection, the invention introduces a batch normalization layer (batch normalization) and a convolution layer with a convolution kernel of 1x1, thereby improving the learning convergence speed and accelerating the training speed while ensuring the detection precision.
The encoder-decoder network is as follows:
the encoder network model:
the encoder includes 6 coding units:
the encoding unit 1: 3 x 3 convolution layer → batch normalization layer → 2 x 2 pooling layer
The encoding unit 2: 3 x 3 convolution layer → batch normalization layer → 2 x 2 pooling layer
The encoding unit 3: [ 3X 3 convolution layer → batch normalization layer ]. X3 → 2X 2 pooling layer
The encoding unit 4: [ 3X 3 convolution layer → batch normalization layer ]. X3 → 2X 2 pooling layer
The encoding unit 5: [ 3X 3 convolution layer → batch normalization layer ]. X3 → 2X 2 pooling layer
The encoding unit 6: 3 x 3 convolution layer → batch normalization layer → 3 x 3 convolution layer → batch normalization layer
Decoder network model:
the decoder comprises 4 coding units, the convolution kernel of the decoder part is 1x1, and the characteristic network parameters can be effectively reduced.
The decoding unit 1: 2 x 2 upsampling layers → [ 1x1 convolutional layers → batch normalization layers ] → [3 x 3 convolutional layers → batch normalization layers ] × 2 of coding units 5 and 6
The decoding unit 2: 2 x 2 upsampling layers → [ 1x1 convolutional layers → batch normalization layers ] → [3 x 3 convolutional layers → batch normalization layers ] × 2 of the encoding unit 4 and the decoding unit 1
The decoding unit 3: 2 × 2 upsampling layers of the decoding unit 2 → [1 × 1 convolutional layers → batch normalization layers ] × 1 → [3 × 3 convolutional layers → batch normalization layers ] × 2
The decoding unit 4: the up-sampling layer of 4 × 4 of the decoding unit 3 → 1 × 1 convolution layer → batch normalization layer → 1 × 1 convolution layer (softmax activation function).
Optimizer and loss function selection:
the image data calculates a cross entropy loss function (loss value) loss value of a prediction (predict) image and an original marked true value (GT) image output by the segmentation network, and the optimizer Adam performs optimization weight to gradually reduce the loss value, so that the segmentation network model learns continuous real data characteristics. The Adam optimizer has the advantages of high efficiency, small occupied memory, suitability for large-scale data and the like.
Training of the model:
based on a loss function and an optimizer, the loss value of the cross entropy loss function calculated by a predicted image output by the target object segmentation network in a verification set (an image not participating in training) and a true value image of an original mark is reduced to be a standard through training, and the weight coefficient is updated and iterated until the optimal weight coefficient is found, so that the target object segmentation network is determined.
step 5, judging foreign matter fault at water leakage hole part
In the detection process, acquiring a real vehicle passing image and extracting a water leakage hole position image, and recording the water leakage hole position image as a water leakage hole position image to be detected;
as shown in fig. 5, inputting the image of the water leakage hole to be detected into the target object segmentation network to obtain a predicted binary image, firstly, judging that a non-zero region exists in the detected binary image, if the non-zero region does not exist in the detected binary image, ending, and continuing to process the next water leakage hole region intercepted image; if the maximum non-zero connected domain exists, the maximum non-zero connected domain is found, the area and the height of the maximum non-zero connected domain are calculated, if the area of the maximum non-zero connected domain is larger than a threshold value a or the height of the maximum non-zero connected domain is larger than a threshold value b, an alarm is given, and if the area of the maximum non-zero connected domain is not larger than the threshold.
Claims (4)
1. The method for detecting the foreign matters in the water leakage holes of the railway wagon is characterized by comprising the following steps of:
collecting an image, extracting the image containing a water leakage hole area, and constructing a sample data set, wherein the sample data set comprises two parts: a gray scale image set and a binary image set;
the gray level image set is a gray level image set containing a water leakage hole area; the binary image set is a binary segmentation image with foreign matters and without foreign matters, which is obtained by an artificial mark corresponding to the water leakage hole area image;
training a target object segmentation network by using a sample data set to obtain a trained target object segmentation network; the target object segmentation network employs an encoder-decoder network structure,
the encoder network model is as follows:
the encoder includes 6 coding units:
the encoding unit 1: 3 x 3 convolution layer → batch normalization layer → 2 x 2 pooling layer;
the encoding unit 2: 3 x 3 convolution layer → batch normalization layer → 2 x 2 pooling layer;
the encoding unit 3: [3 × 3 convolution layer → batch normalization layer ] × 3 → 2 × 2 pooling layer;
the encoding unit 4: [3 × 3 convolution layer → batch normalization layer ] × 3 → 2 × 2 pooling layer;
the encoding unit 5: [3 × 3 convolution layer → batch normalization layer ] × 3 → 2 × 2 pooling layer;
the encoding unit 6: 3 x 3 convolutional layer → batch normalization layer → 3 x 3 convolutional layer → batch normalization layer;
the decoder network model is as follows:
the decoding unit 1: 2 × 2 upsampling layers → [1 × 1 convolutional layers → batch normalization layers ] × 1 → [3 × 3 convolutional layers → batch normalization layers ] × 2 of the coding units 5 and 6;
the decoding unit 2: 2 × 2 upsampling layers → [1 × 1 convolutional layers → batch normalization layers ] × 1 → [3 × 3 convolutional layers → batch normalization layers ] × 2 of the encoding unit 4 and the decoding unit 1;
the decoding unit 3: 2 × 2 upsampling layers of the decoding unit 2 → [1 × 1 convolutional layers → batch normalization layers ] × 1 → [3 × 3 convolutional layers → batch normalization layers ] × 2;
the decoding unit 4: the up-sampling layer of 4 × 4 → 1 × 1 convolution layer → batch normalization layer → 1 × 1 convolution layer of the decoding unit 3;
in the detection process, acquiring a real vehicle passing image and extracting a water leakage hole position image, and recording the water leakage hole position image as a water leakage hole position image to be detected;
inputting the image of the water leakage hole part to be detected into a target object segmentation network to obtain a predicted binary image, finding a maximum foreign body communication domain if a foreign body is detected to exist in the binary image, calculating the area and height of the maximum foreign body communication domain, and performing fault alarm when the area of the foreign body is larger than an area threshold or the length is larger than a length threshold; and in other cases, continuously processing the intercepted image of the next water leakage hole area.
2. The method for detecting foreign matters in the drain hole of the railway wagon as claimed in claim 1, wherein the process of collecting the image and extracting the image containing the drain hole area is realized according to the axle distance information and the position prior knowledge of the drain hole area.
3. The method for detecting foreign matters in the drain holes of the railway wagon as claimed in claim 2, wherein the process of acquiring images and extracting the images containing the areas of the drain holes is realized based on a line camera.
4. The method for detecting foreign matters in the drain holes of the railway wagon as claimed in claim 1, wherein data amplification operation is required in the process of constructing the sample data set; the data amplification form comprises random combination operations of vertical turning, scaling and translation on the image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911272274.5A CN111080602B (en) | 2019-12-12 | 2019-12-12 | Method for detecting foreign matters in water leakage hole of railway wagon |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911272274.5A CN111080602B (en) | 2019-12-12 | 2019-12-12 | Method for detecting foreign matters in water leakage hole of railway wagon |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111080602A CN111080602A (en) | 2020-04-28 |
| CN111080602B true CN111080602B (en) | 2020-10-09 |
Family
ID=70314018
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911272274.5A Active CN111080602B (en) | 2019-12-12 | 2019-12-12 | Method for detecting foreign matters in water leakage hole of railway wagon |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111080602B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111652228B (en) * | 2020-05-21 | 2020-12-04 | 哈尔滨市科佳通用机电股份有限公司 | Railway wagon sleeper beam hole foreign matter detection method |
| CN112001908B (en) * | 2020-08-25 | 2021-03-09 | 哈尔滨市科佳通用机电股份有限公司 | Railway freight car sleeper beam hole carried foreign matter detection method |
| CN112102293B (en) * | 2020-09-16 | 2021-03-02 | 哈尔滨市科佳通用机电股份有限公司 | Rapid detection method for foreign matters in triangular holes of railway wagon |
| CN112508906B (en) * | 2020-12-02 | 2021-07-20 | 哈尔滨市科佳通用机电股份有限公司 | Method, system and device for rapidly detecting foreign matters in water leakage hole of railway wagon |
| CN112614097B (en) * | 2020-12-16 | 2022-02-01 | 哈尔滨市科佳通用机电股份有限公司 | Method for detecting foreign matter on axle box rotating arm of railway train |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101936915A (en) * | 2010-07-30 | 2011-01-05 | 哈尔滨工业大学（威海） | Method for detecting dirt on central region of bottom of beer bottle |
| CN103984961A (en) * | 2014-05-30 | 2014-08-13 | 成都西物信安智能系统有限公司 | Image detection method for detecting foreign matter at bottom of vehicle |
| CN107451999A (en) * | 2017-08-16 | 2017-12-08 | 中惠创智无线供电技术有限公司 | foreign matter detecting method and device based on image recognition |
| US9990561B2 (en) * | 2015-11-23 | 2018-06-05 | Lexmark International, Inc. | Identifying consumer products in images |
| CN109871829A (en) * | 2019-03-15 | 2019-06-11 | 北京行易道科技有限公司 | A kind of detection model training method and device based on deep learning |
| CN110020652A (en) * | 2019-01-07 | 2019-07-16 | 新而锐电子科技(上海)有限公司 | The dividing method of Tunnel Lining Cracks image |
| CN110059758A (en) * | 2019-04-24 | 2019-07-26 | 海南长光卫星信息技术有限公司 | A kind of remote sensing image culture pond detection method based on semantic segmentation |
| CN110175982A (en) * | 2019-04-16 | 2019-08-27 | 浙江大学城市学院 | A kind of defect inspection method based on target detection |
| 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 |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10510146B2 (en) * | 2016-10-06 | 2019-12-17 | Qualcomm Incorporated | Neural network for image processing |
| CN109727270B (en) * | 2018-12-10 | 2021-03-26 | 杭州帝视科技有限公司 | Motion mechanism and texture feature analysis method and system of cardiac nuclear magnetic resonance image |
| CN110189334B (en) * | 2019-05-28 | 2022-08-09 | 南京邮电大学 | Medical image segmentation method of residual error type full convolution neural network based on attention mechanism |
-
2019
- 2019-12-12 CN CN201911272274.5A patent/CN111080602B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101936915A (en) * | 2010-07-30 | 2011-01-05 | 哈尔滨工业大学（威海） | Method for detecting dirt on central region of bottom of beer bottle |
| CN103984961A (en) * | 2014-05-30 | 2014-08-13 | 成都西物信安智能系统有限公司 | Image detection method for detecting foreign matter at bottom of vehicle |
| US9990561B2 (en) * | 2015-11-23 | 2018-06-05 | Lexmark International, Inc. | Identifying consumer products in images |
| CN107451999A (en) * | 2017-08-16 | 2017-12-08 | 中惠创智无线供电技术有限公司 | foreign matter detecting method and device based on image recognition |
| CN110020652A (en) * | 2019-01-07 | 2019-07-16 | 新而锐电子科技(上海)有限公司 | The dividing method of Tunnel Lining Cracks image |
| CN109871829A (en) * | 2019-03-15 | 2019-06-11 | 北京行易道科技有限公司 | A kind of detection model training method and device based on deep learning |
| CN110175982A (en) * | 2019-04-16 | 2019-08-27 | 浙江大学城市学院 | A kind of defect inspection method based on target detection |
| CN110059758A (en) * | 2019-04-24 | 2019-07-26 | 海南长光卫星信息技术有限公司 | A kind of remote sensing image culture pond detection method based on semantic segmentation |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111080602A (en) | 2020-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111080602B (en) | Method for detecting foreign matters in water leakage hole of railway wagon | |
| CN111079734B (en) | Method for detecting foreign matters in triangular holes of railway wagon | |
| CN111652227B (en) | Method for detecting damage fault of bottom floor of railway wagon | |
| Liang et al. | Defect detection of rail surface with deep convolutional neural networks | |
| CN111080650B (en) | Method for detecting looseness and loss faults of small part bearing blocking key nut of railway wagon | |
| CN112434695B (en) | Upper pull rod fault detection method based on deep learning | |
| CN110211101A (en) | A kind of rail surface defect rapid detection system and method | |
| CN111079747A (en) | Railway wagon bogie side frame fracture fault image identification method | |
| CN104021574A (en) | Method for automatically identifying pavement diseases | |
| CN109840483B (en) | Landslide crack detection and identification method and device | |
| CN103903268A (en) | Road surface crack detection method based on multiple lower-layer marked features | |
| CN111652295B (en) | Railway wagon coupler yoke key joist falling fault identification method | |
| CN111080621B (en) | Method for identifying railway wagon floor damage fault image | |
| CN111091548B (en) | Railway wagon adapter dislocation fault image identification method and system based on deep learning | |
| CN104794721A (en) | Quick optic disc positioning method based on multi-scale macula detection | |
| CN111079821A (en) | Derailment automatic braking pull ring falling fault image identification method | |
| CN116757990A (en) | Railway fastener defect online detection and identification method based on machine vision | |
| CN112330646A (en) | Motor car bottom abnormity detection method based on two-dimensional image | |
| CN116630813B (en) | Highway road surface construction quality intelligent detection system | |
| CN111652228B (en) | Railway wagon sleeper beam hole foreign matter detection method | |
| CN113657305B (en) | Video-based intelligent detection method for black smoke vehicle and ringeman blackness level | |
| CN112288717A (en) | Method for detecting foreign matters on side part of motor train unit train | |
| CN105389579A (en) | Method for automatically identifying loss fault of cutout cock handle of train | |
| KR102585146B1 (en) | Pothole detection device and method based on deep learning | |
| CN115424128A (en) | Fault image detection method and system for lower link of freight car bogie |
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 |