CN116452858A - Rail wagon connecting pull rod round pin breaking fault identification method and system - Google Patents
Rail wagon connecting pull rod round pin breaking fault identification method and system Download PDFInfo
- Publication number
- CN116452858A CN116452858A CN202310302963.6A CN202310302963A CN116452858A CN 116452858 A CN116452858 A CN 116452858A CN 202310302963 A CN202310302963 A CN 202310302963A CN 116452858 A CN116452858 A CN 116452858A
- Authority
- CN
- China
- Prior art keywords
- round pin
- pull rod
- image
- connecting pull
- detected
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000001514 detection method Methods 0.000 claims abstract description 54
- 238000001914 filtration Methods 0.000 claims description 9
- 238000012952 Resampling Methods 0.000 claims description 3
- 230000008878 coupling Effects 0.000 claims 1
- 238000010168 coupling process Methods 0.000 claims 1
- 238000005859 coupling reaction Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 9
- 238000013135 deep learning Methods 0.000 abstract description 6
- 230000006870 function Effects 0.000 description 8
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The invention discloses a method and a system for identifying breaking faults of round pins of a connecting pull rod of a railway wagon, and relates to an image detection method and system. The method aims at solving the problem that when the detection effect of the connecting pull rod with a larger target is detected by the existing fault identification mode using deep learning, the better detection effect of the connecting pull rod round pin with a smaller target is difficult to reach, wherein the method comprises the following specific steps: step one, collecting an image to be detected, which comprises a part to be detected; detecting an image to be detected through a pre-trained target detection model, and detecting the positions of the connecting pull rod and the broken round pin; and thirdly, calculating a foreground cross ratio IoF of the connecting pull rod and the broken round pin according to the positions of the connecting pull rod and the broken round pin, and carrying out round pin breaking fault alarming and showing the position of the corresponding broken round pin when IoF is larger than a set IoF threshold value.
Description
Technical Field
The invention relates to an image detection method and system.
Background
The failure of the connecting pull rod round pin to break is a failure endangering driving safety, and serious consequences can occur if the failure is not found in time. At present, human eyes are mainly used for carrying out fault finding on the whole vehicle, and the whole vehicle has large finding range, more parts, more vehicles and more fault forms, so that the work is a mechanical work with strong repeatability, high strength and easy fatigue. When the workers are tired, the occurrence of missed detection and false detection is caused, and the driving safety is affected.
Furthermore, in order to replace the existing manual detection, the fault identification can be effectively identified by using a deep learning mode, the detection speed and the accuracy are improved, the manual working state is improved, the workload is reduced, and the automatic identification alarm is only needed to be confirmed manually.
However, when the fault detection is performed on the round pins of the connecting pull rod, the image contains a plurality of round pins, so that the connecting pull rod needs to be detected firstly, then the round pins at the two ends of the connecting pull rod are detected, the target of the connecting pull rod is large, the target of the round pin of the connecting pull rod is small, and when the fault detection effect detection is performed on the connecting pull rod with a large target in the existing mode of using deep learning, the better detection effect is difficult to achieve on the round pin of the connecting pull rod with a small target.
Disclosure of Invention
The invention aims to solve the problem that when the detection effect of a connecting pull rod with a larger target is detected by using the existing deep learning fault identification mode, the good detection effect of a connecting pull rod round pin with a smaller target is difficult to achieve, and provides a method and a system for identifying the fault of the broken connecting pull rod round pin of a railway wagon.
The invention provides a method for identifying the breaking fault of a round pin of a connecting pull rod of a railway wagon, which comprises the following specific steps:
step one, collecting an image to be detected, which comprises a part to be detected;
the part to be detected comprises a connecting pull rod and a round pin; the round pin comprises a broken round pin;
detecting an image to be detected through a pre-trained target detection model, and detecting the positions of the connecting pull rod and the broken round pin;
the classification detection function of the target detection model is as follows
Wherein a is i Is a class loss weight, and a i ＝e 1-max(w*h-γ,1) The method comprises the steps of carrying out a first treatment on the surface of the Pi is the label of sample i, positive class 1,negative class is 0; pi is the probability that sample i is predicted to be a positive class; w is the width of the target and h is the height of the target; γ=1600 to 6400;
and thirdly, calculating a foreground cross ratio IoF of the connecting pull rod and the broken round pin according to the positions of the connecting pull rod and the broken round pin, and carrying out round pin breaking fault alarming and showing the position of the corresponding broken round pin when IoF is larger than a set IoF threshold value.
Further, the method for obtaining the pre-training weight of the target detection model in the second step comprises the following steps:
filtering the sample images with the sizes smaller than the set size threshold in the data set;
the sample images are images containing connecting ties and images containing round pins.
Further, the method for obtaining the pre-training weight of the target detection model in the second step further comprises the following steps:
adding a sample image containing a small target image; the small target image is an image containing round pins.
Further, the foreground cross ratio threshold is 0.5.
Further, the size threshold is set to 100px x 100px.
The invention also provides a system for identifying the breaking fault of the round pin of the connecting pull rod of the railway wagon, which comprises the following steps:
the image acquisition module to be detected is used for acquiring an image to be detected containing the part to be detected;
the part to be detected comprises a connecting pull rod and a round pin; the round pin comprises a broken round pin;
the component position recognition module is used for detecting an image to be detected through a pre-trained target detection model and detecting the positions of the connecting pull rod and the broken round pin;
the classification detection function of the target detection model is as follows
Wherein a is i Is a class loss weight, and a i ＝e 1-max(w*h-γ,1) The method comprises the steps of carrying out a first treatment on the surface of the Pi is the label of sample i, positive class is1, negative class is 0; pi is the probability that sample i is predicted to be a positive class; w is the width of the target and h is the height of the target; γ=1600 to 6400;
the fault judging module is used for calculating the foreground crossing ratio IoF of the connecting pull rod and the broken round pin according to the positions of the connecting pull rod and the broken round pin, and carrying out round pin breaking fault alarming and showing the positions of the corresponding broken round pins when IoF is larger than a set IoF threshold value.
Further, the pre-training weight of the target detection model is obtained through a pre-training module, and the pre-training module comprises:
the image filtering module is used for filtering sample images with the size smaller than a set size threshold in the data set;
the sample images are images containing connecting ties and images containing round pins.
Further, the pre-training module further comprises:
a small target resampling module for adding a sample image containing a small target image; the small target image is an image containing round pins.
Further, the foreground cross ratio threshold is 0.5.
Further, the size threshold is set to 100px x 100px.
The beneficial effects of the invention are as follows:
the invention provides a method and a device for identifying breaking faults of round pins of a connecting pull rod based on a deep learning algorithm. When the classification loss function is designed by utilizing the mode of automatic image identification, the detection effect of the connecting pull rod with a larger target is detected, and meanwhile, the better detection effect of the connecting pull rod round pin with a smaller target can be achieved, so that the unified standard of the same faults is adopted, and the detection efficiency and the accuracy are improved.
Drawings
FIG. 1 is a flow chart of a method for identifying a break failure of a round pin of a railway wagon connecting pull rod.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that, without conflict, the embodiments of the present invention and features of the embodiments may be combined with each other.
The invention is further described below with reference to the drawings and specific examples, which are not intended to be limiting.
Detailed description of the preferred embodiments
The method for identifying the breaking fault of the round pin of the connecting pull rod of the railway wagon comprises the following specific steps:
step one, collecting an image to be detected, which comprises a part to be detected;
the part to be detected comprises a connecting pull rod and a round pin; the round pin comprises a broken round pin;
detecting an image to be detected through a pre-trained target detection model, and detecting the positions of the connecting pull rod and the broken round pin;
the classification detection function of the target detection model is as follows
Wherein a is i Is a class loss weight, and a i ＝e 1-max(w*h-γ,1) The method comprises the steps of carrying out a first treatment on the surface of the Pi is the label of sample i, positive class is 1, and negative class is 0; pi is the probability that sample i is predicted to be a positive class; w is the width of the target and h is the height of the target; γ=1600 to 6400;
and thirdly, calculating a foreground cross ratio IoF of the connecting pull rod and the broken round pin according to the positions of the connecting pull rod and the broken round pin, and carrying out round pin breaking fault alarming and showing the position of the corresponding broken round pin when IoF is larger than a set IoF threshold value.
Detailed description of the preferred embodiments
This embodiment is a further description of the first embodiment, where the method for obtaining the pre-training weight of the target detection model in the second step includes:
filtering the sample images with the sizes smaller than the set size threshold in the data set;
the sample images are images containing connecting ties and images containing round pins.
Other technical features of the present embodiment are exactly the same as those of the first embodiment.
Detailed description of the preferred embodiments
This embodiment is a further description of the first or second embodiment, where the method for obtaining the pre-training weight of the target detection model in the second step further includes:
adding a sample image containing a small target image; the small target image is an image containing round pins.
Other technical features of the present embodiment are exactly the same as those of the first or second embodiments.
Detailed description of the preferred embodiments
This embodiment is a further description of the first or second embodiment, in which the foreground cross ratio threshold is 0.5.
Other technical features of the present embodiment are exactly the same as those of the first or second embodiments.
Detailed description of the preferred embodiments
This embodiment is a further description of the second embodiment, in which the size threshold is set to 100px×100px.
Other technical features of the present embodiment are exactly the same as those of the second embodiment.
Detailed description of the preferred embodiments six
The utility model provides a railway freight car connecting rod round pin breaks fault identification system which this embodiment includes:
the image acquisition module to be detected is used for acquiring an image to be detected containing the part to be detected;
the part to be detected comprises a connecting pull rod and a round pin; the round pin comprises a broken round pin;
the component position recognition module is used for detecting an image to be detected through a pre-trained target detection model and detecting the positions of the connecting pull rod and the broken round pin;
the classification detection function of the target detection model is as follows
Wherein a is i Is a class loss weight, and a i ＝e 1-max(w*h-γ,1) The method comprises the steps of carrying out a first treatment on the surface of the Pi is the label of sample i, positive class is 1, and negative class is 0; pi is the probability that sample i is predicted to be a positive class; w is the width of the target and h is the height of the target; γ=1600 to 6400;
the fault judging module is used for calculating the foreground crossing ratio IoF of the connecting pull rod and the broken round pin according to the positions of the connecting pull rod and the broken round pin, and carrying out round pin breaking fault alarming and showing the positions of the corresponding broken round pins when IoF is larger than a set IoF threshold value.
Detailed description of the preferred embodiments
In this embodiment, the pre-training weight of the target detection model is obtained by a pre-training module, where the pre-training module includes:
the image filtering module is used for filtering sample images with the size smaller than a set size threshold in the data set;
the sample images are images containing connecting ties and images containing round pins.
Other technical features of the present embodiment are exactly the same as those of the sixth embodiment.
Detailed description of the preferred embodiments
This embodiment is a further description of the sixth or seventh embodiment, where the pre-training module further includes:
a small target resampling module for adding a sample image containing a small target image; the small target image is an image containing round pins.
Other technical features of the present embodiment are exactly the same as those of the sixth or seventh embodiment.
Detailed description of the preferred embodiments nine
This embodiment is a further description of the sixth or seventh embodiment, in which the foreground cross ratio threshold is 0.5.
Other technical features of the present embodiment are exactly the same as those of the sixth or seventh embodiment.
Detailed description of the preferred embodiments
This embodiment is a further description of the seventh embodiment, in which the size threshold is set to 100px×100px.
Other technical features of the present embodiment are exactly the same as those of the seventh embodiment.
DETAILED DESCRIPTION OF EMBODIMENT (S) OF INVENTION
And in the running process of the truck, acquiring a high-definition linear array image of the truck through rail edge equipment. And finding out the area through the information obtained by the hardware and the position relation of the train parts. In this region, fault recognition is performed by using a deep learning method according to the form of the component. When a truck passes through on-line, the image of the target is intercepted by the same method, fault analysis is carried out on the image, and whether the round pin of the connecting pull rod is broken or not is judged. And alarming the area identified as the fault, outputting the position of the fault, and enabling a worker to quickly locate the target area according to the identification result, so as to perform corresponding processing and ensure the safe operation of the train.
The method comprises the following steps:
1. raw image acquisition
And (3) setting up high-speed imaging equipment at a fixed detection site to obtain high-definition linear array gray level images of all parts of the motor train. The method has the advantages that images in different environments in different time periods are collected, more sample data are obtained, various natural interferences such as illumination, rainwater, mud and the like in the data images are guaranteed, the diversity of the data is guaranteed, and therefore the designed algorithm has better robustness.
Intercepting fault subgraph according to component position information
The motor cars of the same type have the same structure, and in the large image of the fault, the sub-image containing the part to be detected can be intercepted according to the wheelbase information.
Target detection algorithm fault detection
The truncated sub-image contains a plurality of round pins, so that the connecting pull rod needs to be detected, and fault detection is performed on the round pins at the two ends of the connecting pull rod. When the data is marked, three types are marked, namely a connecting pull rod, a normal round pin and a broken round pin. Compared with a connecting pull rod, the round pin is a small target, and the detection difficulty is high, so that the small target is focused when the classification loss function is designed. The classification loss function used for the Faster-RCNN target detection is:
the classification loss weights of the large target and the small target are the same, and a weight a is set for increasing the classification precision of the small target i
a i ＝e 1-max(w*h-3600,1)
Using new class loss functions
Meanwhile, the pre-training weight is changed, the original pre-training weight is trained on the coco data set 80 targets, the sizes of the targets are various, in order to increase the detection capability of the small targets, when the pre-training weight is trained, targets with the sizes smaller than 100 x 100 are filtered, meanwhile, samples containing the small target images are added into the railway wagon images, the mixed data set is used for pre-training, and therefore the trained pre-training weight has a better detection effect on the small targets.
Training a Faster-RCNN target detection algorithm, detecting positions of a connecting pull rod, a normal round pin and a broken round pin, and if iof (Intersection over Foreground, foreground cross ratio) of the broken round pin and the connecting pull rod is larger than 0.5, performing fault alarm.
Although the invention herein has been described with reference to particular embodiments, it is to be understood that these embodiments are merely illustrative of the principles and applications of the present invention. It is therefore to be understood that numerous modifications may be made to the illustrative embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the different dependent claims and the features herein may be combined in ways other than as described in the original claims. It is also to be understood that features described in connection with separate embodiments may be used in other embodiments.
Claims (10)
1. A method for identifying the breaking fault of a round pin of a connecting pull rod of a railway wagon is characterized by comprising the following specific steps:
step one, collecting an image to be detected, which comprises a part to be detected;
the part to be detected comprises a connecting pull rod and a round pin; the round pin comprises a broken round pin;
detecting the image to be detected through a pre-trained target detection model, and detecting to obtain the positions of the connecting pull rod and the broken round pin;
the classification detection function of the target detection model is that
Wherein a is i Is a class loss weight, and a i ＝e 1-max(w*h-γ,1) The method comprises the steps of carrying out a first treatment on the surface of the Pi is the label of sample i, positive class is 1, and negative class is 0; pi is the probability that sample i is predicted to be a positive class; w is the width of the target and h is the height of the target; γ=1600 to 6400;
and thirdly, calculating a foreground cross ratio IoF of the connecting pull rod and the broken round pin according to the positions of the connecting pull rod and the broken round pin, and carrying out round pin breaking fault alarming and showing the position of the corresponding broken round pin when IoF is larger than a set IoF threshold value.
2. The method for identifying a break failure of a round pin of a tie rod for railway wagon connection according to claim 1, wherein the method for obtaining the pre-training weight of the target detection model in the second step comprises the following steps:
filtering the sample images with the sizes smaller than the set size threshold in the data set;
the sample images are images containing connecting ties and images containing round pins.
3. The method for identifying a break failure of a round pin of a tie rod for railway wagon connection according to claim 1 or 2, wherein the method for obtaining the pre-training weight of the target detection model in the second step further comprises:
adding a sample image containing a small target image; the small target image is an image containing a round pin.
4. A method of identifying a rail wagon tie-down pin break failure in accordance with claim 1 or 2, wherein the threshold foreground crossing ratio is 0.5.
5. The method for identifying a break failure of a round pin of a tie rod for railway wagon connection according to claim 2, wherein the size threshold is set to 100px x 100px.
6. A rail wagon tie rod round pin break fault identification system, comprising:
the image acquisition module to be detected is used for acquiring an image to be detected containing the part to be detected;
the part to be detected comprises a connecting pull rod and a round pin; the round pin comprises a broken round pin;
the component position identification module is used for detecting the image to be detected through a pre-trained target detection model and detecting the positions of the connecting pull rod and the broken round pin;
the classification detection function of the target detection model is that
Wherein a is i Is a class loss weight, and a i ＝e 1-max(w*h-γ,1) The method comprises the steps of carrying out a first treatment on the surface of the Pi is the label of sample i, positive class is 1, and negative class is 0; pi is the probability that sample i is predicted to be a positive class; w is the width of the target and h is the height of the target; γ=1600 to 6400;
the fault judging module is used for calculating the foreground crossing ratio IoF of the connecting pull rod and the broken round pin according to the positions of the connecting pull rod and the broken round pin, and carrying out round pin breaking fault alarming and showing the positions of the corresponding broken round pins when IoF is larger than a set IoF threshold value.
7. The railroad freight car tie rod round pin break failure recognition system of claim 6, wherein the pre-training weights of the target detection model are obtained by a pre-training module, the pre-training module comprising:
the image filtering module is used for filtering sample images with the size smaller than a set size threshold in the data set;
the sample images are images containing connecting ties and images containing round pins.
8. A rail wagon tie bar round pin break fault identification system as claimed in claim 6 or 7, wherein the pre-training module further comprises:
a small target resampling module for adding a sample image containing a small target image; the small target image is an image containing a round pin.
9. A rail wagon coupling tie-rod round pin break failure recognition system according to claim 6 or 7, wherein the foreground crossing ratio threshold is 0.5.
10. The railroad car tie rod round pin break fault identification system of claim 7, wherein the size threshold is 100px x 100px.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310302963.6A CN116452858B (en) | 2023-03-24 | 2023-03-24 | Rail wagon connecting pull rod round pin breaking fault identification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310302963.6A CN116452858B (en) | 2023-03-24 | 2023-03-24 | Rail wagon connecting pull rod round pin breaking fault identification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116452858A true CN116452858A (en) | 2023-07-18 |
| CN116452858B CN116452858B (en) | 2023-12-15 |
Family
ID=87129422
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310302963.6A Active CN116452858B (en) | 2023-03-24 | 2023-03-24 | Rail wagon connecting pull rod round pin breaking fault identification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116452858B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111080611A (en) * | 2019-12-12 | 2020-04-28 | 哈尔滨市科佳通用机电股份有限公司 | Railway wagon bolster spring fracture fault image identification method |
| CN111091544A (en) * | 2019-12-12 | 2020-05-01 | 哈尔滨市科佳通用机电股份有限公司 | Method for detecting breakage fault of side integrated framework of railway wagon bogie |
| WO2020164282A1 (en) * | 2019-02-14 | 2020-08-20 | 平安科技（深圳）有限公司 | Yolo-based image target recognition method and apparatus, electronic device, and storage medium |
| CN111652296A (en) * | 2020-05-21 | 2020-09-11 | 哈尔滨市科佳通用机电股份有限公司 | Deep learning-based rail wagon lower pull rod fracture fault detection method |
| CN112733742A (en) * | 2021-01-14 | 2021-04-30 | 哈尔滨市科佳通用机电股份有限公司 | Deep learning-based fault detection method for round pin of lower pull rod of railway wagon |
| CN113506269A (en) * | 2021-07-12 | 2021-10-15 | 东南大学 | Turnout and non-turnout rail fastener positioning method based on deep learning |
| CN114841257A (en) * | 2022-04-21 | 2022-08-02 | 北京交通大学 | Small sample target detection method based on self-supervision contrast constraint |
| CN115170882A (en) * | 2022-07-19 | 2022-10-11 | 哈尔滨市科佳通用机电股份有限公司 | Optimization method of rail wagon part detection network and guardrail breaking fault identification method |
-
2023
- 2023-03-24 CN CN202310302963.6A patent/CN116452858B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020164282A1 (en) * | 2019-02-14 | 2020-08-20 | 平安科技（深圳）有限公司 | Yolo-based image target recognition method and apparatus, electronic device, and storage medium |
| CN111080611A (en) * | 2019-12-12 | 2020-04-28 | 哈尔滨市科佳通用机电股份有限公司 | Railway wagon bolster spring fracture fault image identification method |
| CN111091544A (en) * | 2019-12-12 | 2020-05-01 | 哈尔滨市科佳通用机电股份有限公司 | Method for detecting breakage fault of side integrated framework of railway wagon bogie |
| CN111652296A (en) * | 2020-05-21 | 2020-09-11 | 哈尔滨市科佳通用机电股份有限公司 | Deep learning-based rail wagon lower pull rod fracture fault detection method |
| CN112733742A (en) * | 2021-01-14 | 2021-04-30 | 哈尔滨市科佳通用机电股份有限公司 | Deep learning-based fault detection method for round pin of lower pull rod of railway wagon |
| CN113506269A (en) * | 2021-07-12 | 2021-10-15 | 东南大学 | Turnout and non-turnout rail fastener positioning method based on deep learning |
| CN114841257A (en) * | 2022-04-21 | 2022-08-02 | 北京交通大学 | Small sample target detection method based on self-supervision contrast constraint |
| CN115170882A (en) * | 2022-07-19 | 2022-10-11 | 哈尔滨市科佳通用机电股份有限公司 | Optimization method of rail wagon part detection network and guardrail breaking fault identification method |
Non-Patent Citations (4)
| Title |
|---|
| CHI-HAO WU ET.AL: "Boosted Convolutional Neural Networks (BCNN) for Pedestrian DetectionA novel Faster-RCNN Mutant network for breast lesions screening", 《2017 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2017)》, pages 540 - 549 * |
| 冶静: "青藏铁路运行货车故障诊断系统与智能图像检测方法", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》, pages 033 - 298 * |
| 刘云鹏;裴少通;武建华;纪欣欣;梁利辉;: "基于深度学习的输变电设备异常发热点红外图片目标检测方法", 南方电网技术, no. 02, pages 33 - 39 * |
| 耿振宇: "基于Faster R-CNN的PCB表面缺陷检测模型研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》, pages 135 - 342 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116452858B (en) | 2023-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111079819B (en) | Method for judging state of coupler knuckle pin of railway wagon based on image recognition and deep learning | |
| CN105719482A (en) | Dynamic vehicle information collecting device | |
| CN113516629A (en) | Intelligent detection system for TFDS passing operation | |
| CN111080598A (en) | Bolt and nut missing detection method for coupler yoke key safety crane | |
| CN111080611A (en) | Railway wagon bolster spring fracture fault image identification method | |
| Guo et al. | Image-based seat belt detection | |
| CN103745224A (en) | Image-based railway contact net bird-nest abnormal condition detection method | |
| CN205541424U (en) | Developments vehicle information acquisition device | |
| CN111080600A (en) | Fault identification method for split pin on spring supporting plate of railway wagon | |
| CN111080601A (en) | Method for identifying fault image of pull ring grinding shaft of derailment brake device of railway wagon | |
| CN111080607B (en) | Rolling bearing oil slinging fault detection method based on image recognition | |
| CN111723697A (en) | Improved driver background segmentation method based on Mask-RCNN | |
| CN108357517A (en) | Motor train unit train operation troubles information detecting method and system | |
| CN115527170A (en) | Method and system for identifying closing fault of door stopper handle of automatic freight car derailing brake device | |
| CN111079629A (en) | Railway wagon center plate out-of-position fault image identification method | |
| CN111967384A (en) | Vehicle information processing method, device, equipment and computer readable storage medium | |
| CN115272850A (en) | Railway wagon BAB type brake adjuster pull rod head breaking fault identification method | |
| CN111080599A (en) | Fault identification method for hook lifting rod of railway wagon | |
| CN112329858B (en) | Image recognition method for breakage fault of anti-loosening iron wire of railway motor car | |
| CN116452858B (en) | Rail wagon connecting pull rod round pin breaking fault identification method and system | |
| CN112418323B (en) | Railway wagon coupler knuckle pin fault detection method based on image processing | |
| CN111832562B (en) | Spring supporting plate fault identification method based on image processing | |
| CN116164984A (en) | Road test device and test method for speed limit sign auxiliary system | |
| CN116631187A (en) | Intelligent acquisition and analysis system for case on-site investigation information | |
| DE102019118106B4 (en) | Method for determining a visibility |
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 |