CN111091558A - Railway wagon swing bolster spring jumping fault image identification method - Google Patents
Railway wagon swing bolster spring jumping fault image identification method Download PDFInfo
- Publication number
- CN111091558A CN111091558A CN201911293704.1A CN201911293704A CN111091558A CN 111091558 A CN111091558 A CN 111091558A CN 201911293704 A CN201911293704 A CN 201911293704A CN 111091558 A CN111091558 A CN 111091558A
- Authority
- CN
- China
- Prior art keywords
- image
- bolster spring
- spring
- fault
- deep learning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
- G06T2207/10004—Still image; Photographic image
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- 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
- G06T2207/30108—Industrial image inspection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
A railway freight car bolster spring fleeing fault image identification method relates to the technical field of freight train detection, and aims at the problem of low detection accuracy rate of a manual detection mode in the prior art, and comprises the following steps: the method comprises the following steps: obtaining an approach truck image, and performing coarse positioning to obtain a coarse positioning gray level image set; step two: marking the rough positioning gray level image set to obtain a marked image set; step three: training a deep learning network model by using the marked image set and the original image; step four: and segmenting the swing bolster spring by using the trained deep learning network model, and judging whether the swing bolster spring has a fleeing fault according to a segmentation result. The invention utilizes the automatic image identification mode to replace manual detection, can automatically identify vehicle faults and give an alarm, has uniform operation standards, is not influenced by different experience, understanding and cognition degrees of detected drivers, and improves the detection efficiency and accuracy.
Description
Technical Field
The invention relates to the technical field of freight train detection, in particular to a railway wagon bolster spring fleeing fault image identification method.
Background
The bolster spring of the railway wagon is used for a running part of a railway vehicle, plays a role in buffering and shock absorption, and is located in the running part, if the bolster spring breaks, the running safety of the train is directly endangered, the car body is inclined due to the fact that the bolster spring breaks down, the running safety is endangered, and in the fault detection of the bolster spring, the fault detection is carried out in a mode of manually checking images. The conditions of fatigue, omission and the like are easily caused by vehicle inspection personnel in the working process, so that the appearance of missed inspection and wrong inspection is caused, and the driving safety is influenced. The detection efficiency and stability can be improved by adopting an automatic image identification mode. In recent years, deep learning and artificial intelligence are continuously developed, and the technology is continuously mature. Therefore, deep learning is adopted to identify the jumping fault of the swing bolster spring, and the detection accuracy can be effectively improved.
Disclosure of Invention
The purpose of the invention is: aiming at the problem of low detection accuracy of a manual detection mode in the prior art, the image identification method for the railway freight car bolster spring fleeing fault is provided.
The technical scheme of the invention is as follows:
a rail wagon bolster spring fleeing fault image identification method comprises the following steps:
the method comprises the following steps: obtaining an approach truck image, and performing coarse positioning to obtain a coarse positioning gray level image set;
step two: marking the rough positioning gray level image set to obtain a marked image set;
step three: training a deep learning network model by using the marked image set and the original image;
step four: the method comprises the following steps of utilizing a trained deep learning network model to divide the swing bolster spring, judging whether the swing bolster spring has a fleeing fault according to a division result, and performing the judging process:
firstly, predicting an image by using a deep learning network model, predicting a contour region of a bolster spring and an adjacent component by using a trained weight coefficient to obtain a predicted multi-value image, calculating the position relation between the inclination angle of each row of springs and the adjacent component by using an image processing mode, if the angle deviation is larger than a preset threshold value or the position relation between the inclination angle of each row of springs and the adjacent component is changed, determining that the bolster spring has a fault, and if the angle and the position relation are normal, processing the next bolster spring image.
Further, the method further comprises performing data amplification on the coarse positioning gray-scale image set before the second step.
Further, the data augmentation includes brightness enhancement, cropping, saturation adjustment, rotation, translation, scaling, and mirroring of the image.
Further, the activation function of the deep learning network model is as follows:
Mish＝x＊tanh(n(1+e^x))。
further, the loss function of the deep learning network model is as follows:
wherein, y represents the distribution of the real mark,
the distribution is predicted.
Further, the method further comprises a secondary verification step, wherein the secondary verification step specifically comprises the following steps: and according to the fault identified in the fourth step, according to a height map provided by the 3D image information, judging whether the height of the spring with the spring running out is changed or not through image processing and mode identification, if so, giving a fault alarm, and if the detection is normal, processing the next image of the swing bolster spring.
Further, the labeled image set is generated by training a model according to a small amount of data of a fixed part near the spring and then utilizing a deep learning model.
Further, the marked image set marks the images according to the positions of the springs in different columns in the images as follows: three different categories of front, middle and back.
The invention has the beneficial effects that:
1. the mode of utilizing image automatic identification replaces artifical detection, can automatic identification vehicle trouble and report to the police, and the operation standard is unified, no longer receives the influence that inspection car personnel experience, understanding and cognitive degree are different, improves detection efficiency and rate of accuracy.
2. The deep learning algorithm is applied to automatic identification of the jumping-out fault of the swing bolster spring, and the stability and the precision of the whole algorithm are improved.
3. After a small amount of data of a fixed part near the spring is trained, the model is automatically generated by deep learning, and the efficiency of generating a data set can be improved and the robustness of the model can be enhanced by only adding or modifying part of the marks of the spring.
4. The present invention addresses confusion between different entities of springs by labeling different columns of springs in a coarse positioning subgraph as different categories.
5. The invention can not only identify whether the spring has the play fault, but also select the output fault type: left fleeing and right fleeing; channeling up and down; inside and outside fleeing, etc.
6. The alarm of deep learning identification can be further verified for the second time through a 3D image processing program, and the number of false alarms is reduced.
Drawings
Fig. 1 is a flow chart of fault identification.
Fig. 2 is a flowchart of weight coefficient calculation.
Fig. 3 is a U-shaped deep learning network model.
Detailed Description
The first embodiment is as follows: specifically describing the embodiment with reference to fig. 1 to 3, the method for identifying the railway freight car bolster spring fleeing fault image in the embodiment comprises the following steps:
the method comprises the following steps: obtaining an approach truck image, and performing coarse positioning to obtain a coarse positioning gray level image set;
step two: marking the rough positioning gray level image set to obtain a marked image set;
step three: training a deep learning network model by using the marked image set and the original image;
step four: the method comprises the following steps of utilizing a trained deep learning network model to divide the swing bolster spring, judging whether the swing bolster spring has a fleeing fault according to a division result, and performing the judging process:
firstly, predicting an image by using a deep learning network model, predicting a contour region of a bolster spring and an adjacent component by using a trained weight coefficient to obtain a predicted multi-value image, calculating the relation between the inclination angle of each row of springs and the adjacent component by using an image processing mode, if the angle deviation is larger than a preset threshold value or the position relation with the adjacent component is changed, determining that the bolster spring is in a fault state, and if the angle and position relation is normal, processing the next bolster spring image.
1. Establishing a sample data set
High-definition equipment is respectively built around the rail of the truck, and after the truck passes through the equipment, high-definition images of the two sides, the bottom and the upper part of the truck in all directions are obtained. The truck parts can be influenced by natural conditions such as light weight of goods, rain, mud stain, oil stain, white paint, black paint, foreign matters, ice and snow, chalk characters and the like or artificial conditions. Also, there may be differences in the images taken at different sites. Thus, there are many differences between the bolster spring images. Therefore, in the process of collecting the image data of the bolster springs, the images of the bolster springs under various conditions are collected as completely as possible to ensure the diversity.
The sample data set includes: roughly positioning a gray-scale image set and a mark image set. The coarsely positioned gray level image set is a high-definition gray level image shot by the equipment. The marked image set is a segmentation image of the bolster spring component and the adjacent component, and the partial image is a multi-value image and is obtained in a manual marking mode. There is a one-to-one correspondence between the coarse positioning grayscale image data set and the marker image data set, i.e., each grayscale image corresponds to one marker image.
Although the creation of the sample data set includes images under various conditions, data amplification of the sample data set is still required to improve the stability of the algorithm. The amplification form comprises operations of brightness enhancement, cropping, saturation adjustment, rotation, translation, scaling, mirror image and the like of the image, and each operation is performed under random conditions, so that the diversity and applicability of the sample can be ensured to the greatest extent.
The method has the problems that the spring part has more outlines, the marking is complex, and the sample data set is difficult to establish. The invention firstly marks a small amount of data and then trains out deep learning. And predicting new mark data by using the model of the initial training, and finely adjusting the condition that part of mark data has deviation, so that the efficiency of establishing the sample data set can be improved.
In order to avoid the vehicle speed, the angular deviation of a camera, the weight of goods and the like, the size of the samples in the data set is uniformly scaled to N x N, and uniform image preprocessing is carried out, so that the robustness of the system can be improved.
2. Computing sample data set weights
First, the weight coefficient is initialized and initialized in a random manner.
Secondly, the sample data is subjected to gray level normalization processing, namely, the gray level value of the sample data is normalized to be in the range of 0 to 1. And inputting the normalized data serving as input data into a U-Net deep learning network for data transformation.
The U-Net deep learning network mainly comprises operations such as convolution (convolution), Pooling (Pooling) and activation function (Mish) functions. Convolution is the operation of element-by-element multiplication and summation of a two-dimensional filter matrix (convolution kernel) with a two-dimensional image to be processed. Pooling is to reduce the dimension of the input image, reduce pixel information and only retain important information. Maximum pooling (max-pooling) preserves the maximum value within each block. Mish is a new smooth non-monotonic neural activation function, whose formula is as follows:
Mish＝x＊tanh(ln(1+e^x))(1)
mish has a strong theoretical source, a model training curve is stable and stable, and the performance of Mish is superior to ReLU in terms of training stability and accuracy in a test.
The U-Net deep learning network comprises the following steps:
the first step is a feature extraction section. And 4 times of downsampling, and realizing multi-scale feature identification of the image features by the network by adopting 5 posing layers with convolution kernel sizes of 3x 3.
The second step is that: an upsampling section. Up-sampling was performed 4 times, with 5 convolution kernels of size 3x 3. The up-sampling part fuses the output of the feature extraction part, so that the multi-scale features are actually fused together, and four times of fusion processes exist in the network.
The third step: and outputting the transformed data.
Judging the category of each pixel point by initializing the weight and the function of transforming data to obtain a predicted image, carrying out differential comparison on the predicted image and a real label image, and taking the calculation result of the classified cross entropy as a loss function, namely a formula (2):
wherein, y represents the distribution of the real mark,
the distribution is predicted.
The loss values are calculated by a loss function, and the weights are optimized by an optimizer Adam. The Adam optimizer has the advantages of high efficiency, small occupied memory, suitability for large-scale data and the like.
As shown in the following equation (3):
w is the weight, WiThe learning rate is set to 0.0002, high learning rate means that more steps are taken in weight update, so the model may take less time to converge to the optimal set of weights.
After passing through the loss function and the optimizer, calculating a new weight coefficient, updating the weight coefficient, and completing one training iteration. The program will repeat this process, completing all images for a fixed number of iterations, but not updating the weights for each iteration, only the lower weights of the loss function will be updated until the optimal weight coefficients are found.
3. Swing bolster spring break-out fault discrimination
And after the real vehicle passing image is subjected to data transformation by using a U-Net deep learning network, predicting the outline area of the bolster spring and the adjacent component by using the trained weight coefficient to obtain a predicted multi-value image, wherein the value 0 in the multi-value image is an irrelevant area, and the value other than 0 is the area of the bolster spring and the adjacent component. Normally, each row of swing bolster springs are parallel to each other, and the position relationship between the head ring and the tail ring of the springs and nearby components is fixed. If the swing of the swing bolster spring occurs due to faults, in a multi-value swing bolster spring image, the relation between the inclination angle of each row of springs and an adjacent part is calculated by using an image processing mode, and if the angle deviation is larger than a preset threshold value or the position relation between the angle deviation and the adjacent part is changed, the fault alarm is carried out after the coordinates of the swing bolster springs are converted. And if the angle and the position relation is normal, processing the next image of the swing bolster spring.
4. Fault detail type output and fault secondary verification
The identification result before the step 3 has high identification rate, and the step is optional. And for the fault identified in the step, judging whether the height of the spring with the spring running out is changed or not according to a height map provided by the 3D image information through image processing and mode identification, if so, performing fault alarm in the step 3, and if the detection is normal, processing the next swing bolster spring image. The present invention can select whether to output a detailed failure type of spring pop-out.
It should be noted that the detailed description is only for explaining and explaining the technical solution of the present invention, and the scope of protection of the claims is not limited thereby. It is intended that all such modifications and variations be included within the scope of the invention as defined in the following claims and the description.
Claims (8)
1. A rail wagon bolster spring fleeing fault image identification method is characterized by comprising the following steps:
the method comprises the following steps: obtaining an approach truck image, and performing coarse positioning to obtain a coarse positioning gray level image set;
step two: marking the rough positioning gray level image set to obtain a marked image set;
step three: training a deep learning network model by using the marked image set and the original image;
step four: the method comprises the following steps of utilizing a trained deep learning network model to divide the swing bolster spring, judging whether the swing bolster spring has a fleeing fault according to a division result, and performing the judging process:
firstly, predicting an image by using a deep learning network model, predicting a contour region of a bolster spring and an adjacent component by using a trained weight coefficient to obtain a predicted multi-value image, calculating the position relation between the inclination angle of each row of springs and the adjacent component by using an image processing mode, if the angle deviation is larger than a preset threshold value or the position relation between the inclination angle of each row of springs and the adjacent component is changed, determining that the bolster spring has a fault, and if the angle and the position relation are normal, processing the next bolster spring image.
2. The method for identifying railway wagon bolster spring-out fault image as claimed in claim 1, wherein the method further comprises performing data amplification on the rough positioning gray scale image set before the second step.
3. The method for identifying the fault image of the spring play of the bolster of the railway wagon of claim 2, wherein the data amplification comprises brightness enhancement, clipping, saturation adjustment, rotation, translation, scaling and mirror image of the image.
4. The method for identifying the railway wagon bolster spring fleeing fault image according to claim 1, wherein the activation function of the deep learning network model is as follows:
Mish＝x＊tanh(n(1+e^x))。
5. the method for identifying the railway wagon bolster spring fleeing fault image according to claim 1, wherein the loss function of the deep learning network model is as follows:
wherein, y represents the distribution of the real mark,
the distribution is predicted.
6. The method for identifying the railway wagon bolster spring fleeing fault image as claimed in claim 1, wherein the method further comprises a secondary verification step, and the secondary verification step specifically comprises the following steps: and according to the fault identified in the fourth step, according to a height map provided by the 3D image information, judging whether the height of the spring with the spring running out is changed or not through image processing and mode identification, if so, giving a fault alarm, and if the detection is normal, processing the next image of the swing bolster spring.
7. The method for identifying the railway wagon bolster spring fleeing fault image as claimed in claim 1, wherein the marker image set is generated by utilizing a deep learning model after training the model according to a small amount of data of a fixed part near a spring.
8. The method for identifying the railway freight car bolster spring fleeing fault image as claimed in claim 1, wherein the marking image set marks the springs according to different columns according to the positions of the springs in the image as: three different categories of front, middle and back.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911293704.1A CN111091558B (en) | 2019-12-12 | 2019-12-12 | Railway wagon swing bolster spring jumping fault image identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911293704.1A CN111091558B (en) | 2019-12-12 | 2019-12-12 | Railway wagon swing bolster spring jumping fault image identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111091558A true CN111091558A (en) | 2020-05-01 |
| CN111091558B CN111091558B (en) | 2020-08-21 |
Family
ID=70395094
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911293704.1A Active CN111091558B (en) | 2019-12-12 | 2019-12-12 | Railway wagon swing bolster spring jumping fault image identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111091558B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111862408A (en) * | 2020-06-16 | 2020-10-30 | 北京华电天仁电力控制技术有限公司 | Intelligent access control method |
| CN112037207A (en) * | 2020-09-01 | 2020-12-04 | 哈尔滨市科佳通用机电股份有限公司 | Method for detecting closing fault of automatic brake valve plug handle during railway wagon derailment |
| CN112085723A (en) * | 2020-09-09 | 2020-12-15 | 哈尔滨市科佳通用机电股份有限公司 | Automatic detection method for spring jumping fault of truck bolster |
| CN112132824A (en) * | 2020-09-30 | 2020-12-25 | 哈尔滨市科佳通用机电股份有限公司 | Automatic detection method for failure of freight car axle box spring |
| CN112183323A (en) * | 2020-09-27 | 2021-01-05 | 哈尔滨市科佳通用机电股份有限公司 | Motor car stone sweeper loss fault image recognition method based on deep learning |
| CN112329859A (en) * | 2020-11-06 | 2021-02-05 | 哈尔滨市科佳通用机电股份有限公司 | Method for identifying lost fault image of sand spraying pipe nozzle of railway motor car |
| CN113177965A (en) * | 2021-04-09 | 2021-07-27 | 上海工程技术大学 | Coal rock full-component extraction method based on improved U-net network and application thereof |
| CN115170890A (en) * | 2022-07-28 | 2022-10-11 | 哈尔滨市科佳通用机电股份有限公司 | Method for identifying breakage fault of connecting pull rod chain of railway wagon |
| CN115272287A (en) * | 2022-08-19 | 2022-11-01 | 哈尔滨市科佳通用机电股份有限公司 | Fault detection method, medium and system for rail wagon buffer and slave plate |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102332089A (en) * | 2011-06-23 | 2012-01-25 | 北京康拓红外技术股份有限公司 | Railway wagon brake shoe key going-out fault recognition method based on artificial neural network |
| CN106226050A (en) * | 2016-07-15 | 2016-12-14 | 北京航空航天大学 | A kind of TFDS fault automatic identifying method |
| CN106778740A (en) * | 2016-12-06 | 2017-05-31 | 北京航空航天大学 | A kind of TFDS non-faulting image detecting methods based on deep learning |
| CN109241986A (en) * | 2018-05-30 | 2019-01-18 | 北京飞鸿云际科技有限公司 | For rail traffic vehicles part diagram as the sample production method of identification model |
| CN109800788A (en) * | 2018-12-18 | 2019-05-24 | 中国海洋大学 | A method of reducing image segmentation data set mark |
| CN110490195A (en) * | 2019-08-07 | 2019-11-22 | 桂林电子科技大学 | A kind of water meter dial plate Recognition of Reading method |
-
2019
- 2019-12-12 CN CN201911293704.1A patent/CN111091558B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102332089A (en) * | 2011-06-23 | 2012-01-25 | 北京康拓红外技术股份有限公司 | Railway wagon brake shoe key going-out fault recognition method based on artificial neural network |
| CN106226050A (en) * | 2016-07-15 | 2016-12-14 | 北京航空航天大学 | A kind of TFDS fault automatic identifying method |
| CN106778740A (en) * | 2016-12-06 | 2017-05-31 | 北京航空航天大学 | A kind of TFDS non-faulting image detecting methods based on deep learning |
| CN109241986A (en) * | 2018-05-30 | 2019-01-18 | 北京飞鸿云际科技有限公司 | For rail traffic vehicles part diagram as the sample production method of identification model |
| CN109800788A (en) * | 2018-12-18 | 2019-05-24 | 中国海洋大学 | A method of reducing image segmentation data set mark |
| CN110490195A (en) * | 2019-08-07 | 2019-11-22 | 桂林电子科技大学 | A kind of water meter dial plate Recognition of Reading method |
Non-Patent Citations (3)
| Title |
|---|
| SUN JUNHUA等: "Automatic multi-fault recogniton in TFDS based on convolutional neural network", 《NEUROCOMPUTING》 * |
| 姜媛: "货车枕簧丢失故障动态图像识别方法", 《光学技术》 * |
| 黄焰 等: "机车枕簧故障光电动态检测研究", 《软件》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111862408A (en) * | 2020-06-16 | 2020-10-30 | 北京华电天仁电力控制技术有限公司 | Intelligent access control method |
| CN112037207A (en) * | 2020-09-01 | 2020-12-04 | 哈尔滨市科佳通用机电股份有限公司 | Method for detecting closing fault of automatic brake valve plug handle during railway wagon derailment |
| CN112085723A (en) * | 2020-09-09 | 2020-12-15 | 哈尔滨市科佳通用机电股份有限公司 | Automatic detection method for spring jumping fault of truck bolster |
| CN112085723B (en) * | 2020-09-09 | 2021-04-02 | 哈尔滨市科佳通用机电股份有限公司 | Automatic detection method for spring jumping fault of truck bolster |
| CN112183323A (en) * | 2020-09-27 | 2021-01-05 | 哈尔滨市科佳通用机电股份有限公司 | Motor car stone sweeper loss fault image recognition method based on deep learning |
| CN112132824A (en) * | 2020-09-30 | 2020-12-25 | 哈尔滨市科佳通用机电股份有限公司 | Automatic detection method for failure of freight car axle box spring |
| CN112329859A (en) * | 2020-11-06 | 2021-02-05 | 哈尔滨市科佳通用机电股份有限公司 | Method for identifying lost fault image of sand spraying pipe nozzle of railway motor car |
| CN113177965A (en) * | 2021-04-09 | 2021-07-27 | 上海工程技术大学 | Coal rock full-component extraction method based on improved U-net network and application thereof |
| CN115170890A (en) * | 2022-07-28 | 2022-10-11 | 哈尔滨市科佳通用机电股份有限公司 | Method for identifying breakage fault of connecting pull rod chain of railway wagon |
| CN115272287A (en) * | 2022-08-19 | 2022-11-01 | 哈尔滨市科佳通用机电股份有限公司 | Fault detection method, medium and system for rail wagon buffer and slave plate |
| CN115272287B (en) * | 2022-08-19 | 2023-04-07 | 哈尔滨市科佳通用机电股份有限公司 | Fault detection method, medium and system for rail wagon buffer and slave plate |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111091558B (en) | 2020-08-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111091558B (en) | Railway wagon swing bolster spring jumping fault image identification method | |
| CN111080611A (en) | Railway wagon bolster spring fracture fault image identification method | |
| CN111091542B (en) | Image identification method for breakage fault of spring supporting plate of railway wagon bogie | |
| CN103049763B (en) | Context-constraint-based target identification method | |
| CN111080608A (en) | Method for recognizing closing fault image of automatic brake valve plug handle of railway wagon in derailment | |
| CN111091547B (en) | Railway wagon brake beam strut fracture fault image identification method | |
| CN111091541B (en) | Method for identifying fault of missing nut in cross beam assembly of railway wagon | |
| CN111079817B (en) | Method for identifying fault image of cross beam of railway wagon | |
| CN111079746B (en) | Railway wagon axle box spring fault image identification method | |
| CN111091548B (en) | Railway wagon adapter dislocation fault image identification method and system based on deep learning | |
| CN111080621B (en) | Method for identifying railway wagon floor damage fault image | |
| CN111080600A (en) | Fault identification method for split pin on spring supporting plate of railway wagon | |
| CN111091544A (en) | Method for detecting breakage fault of side integrated framework of railway wagon bogie | |
| CN111079822A (en) | Method for identifying dislocation fault image of middle rubber and upper and lower plates of axle box rubber pad | |
| CN111079630B (en) | Fault identification method for railway wagon brake beam with incorrect installation position | |
| CN108932471A (en) | A kind of vehicle checking method | |
| CN112329858B (en) | Image recognition method for breakage fault of anti-loosening iron wire of railway motor car | |
| CN111652228B (en) | Railway wagon sleeper beam hole foreign matter detection method | |
| CN113516652A (en) | Battery surface defect and adhesive detection method, device, medium and electronic equipment | |
| CN113158954A (en) | Automatic traffic off-site zebra crossing area detection method based on AI technology | |
| CN111091554B (en) | Railway wagon swing bolster fracture fault image identification method | |
| CN116188449B (en) | Rail wagon relief valve pull rod split pin loss fault identification method and equipment | |
| CN116958639A (en) | Method for detecting and enhancing false behavior of traffic light recognition model | |
| CN114155518A (en) | Expressway visor inclination identification method based on deep semantic segmentation network and image correction | |
| CN113989255A (en) | Subway tunnel lining shedding recognition model training method and recognition method based on Mask-RCNN |
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 |