CN115439412A - Bridge surface crack detection method and device of lightweight convolutional network - Google Patents
Bridge surface crack detection method and device of lightweight convolutional network Download PDFInfo
- Publication number
- CN115439412A CN115439412A CN202210936787.7A CN202210936787A CN115439412A CN 115439412 A CN115439412 A CN 115439412A CN 202210936787 A CN202210936787 A CN 202210936787A CN 115439412 A CN115439412 A CN 115439412A
- Authority
- CN
- China
- Prior art keywords
- network
- bridge
- crack
- image data
- data set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06T7/0008—Industrial image inspection checking presence/absence
-
- 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
- G06V10/40—Extraction of image or video features
-
- 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
-
- 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
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20088—Trinocular vision calculations; trifocal tensor
-
- 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
- G06T2207/30132—Masonry; Concrete
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Image Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Quality & Reliability (AREA)
- Life Sciences & Earth Sciences (AREA)
Abstract
The invention discloses a method and equipment for detecting cracks on the surface of a bridge of a lightweight convolution network, wherein the method comprises the following steps: constructing a lightweight convolution network model; wherein the lightweight convolutional neural network is a MobileNet-based YOLOv4 network; acquiring a bridge crack image data set, and training the lightweight convolution network model based on the bridge crack image data set to obtain a trained crack detection network; and detecting the bridge cracks in real time by using the crack detection network. According to the invention, the MobileNet-based YOLOv4 network is constructed, and the MobileNet-based YOLOv4 network is lightened, so that the MobileNet-based YOLOv4 network is trained for obtaining a bridge crack image data set, and the trained network is utilized to perform real-time detection on bridge cracks; the method can effectively overcome the defects of low detection precision and low detection speed of the existing crack detection method based on machine learning.
Description
Technical Field
The invention belongs to the field of intelligent bridge detection, and particularly relates to a method and equipment for detecting cracks on the surface of a concrete bridge through a lightweight convolution network.
Background
The Guizhou called "world bridge museum" is located on the noble plateau of the cloud in the southwest of China, and wide karst landforms are distributed in the interior, which can be called gullies, vertical and horizontal, and the ground surface is broken. With the large-scale construction of Guizhou in the traffic field, a high bridge with one seat appears on the Guizhou ground, so that the skben becomes a way. At present, 2.1 thousands of highway bridges are built in Guizhou province, 5000 more bridges are built, and almost all bridge types in the world are covered. According to statistics, among 100 bridges ranked all over the world, more than 80 bridges come from China; most of the 80 bridges come from Guizhou.
The structure of Guizhou bridges is mostly a concrete structure. The workload for health monitoring and maintenance associated with such a large number of concrete type bridges has increased. At present, the damage detection method of the bridge in China mainly comprises a manual observation method and a bridge nondestructive detection method. The failure of concrete structures is due to a number of factors, one of which is cracking. If the crack is allowed to expand, the concrete structure can be caused to fail, safety accidents are caused, and once the bridge has an accident, the accident is often a huge accident. In the existing detection method, the requirement on the professional skill of workers is high, and especially, special geographical positions, such as a north-disk river bridge, a duck pond river bridge and other high bridges, span between high peaks. If the weather is relatively bad, the high-altitude operation is performed by the survey personnel with great difficulty. Therefore, the method has great significance for detecting the bridge cracks.
In recent years, with the outbreak of deep learning, the development of computer vision technology is being driven to a peak. In the vision technology, image classification, target detection and semantic segmentation are the bases and the 3 fields which are developed most rapidly. Image classification is mainly to determine what object each image is. The object detection is to determine the location and type of the object. The semantic segmentation is to determine which category each pixel belongs to. The characteristics with better robustness and semantic property can be extracted from a large number of parameters of the deep neural network. In the prior art, an advanced target detection algorithm is applied to crack detection on the surface of a bridge in the prior art, so that potential safety hazards of aerial work are eliminated to a certain extent, the bridge crack real-time detection is realized, the crack identification precision is ensured, the detection efficiency is improved, and the time and labor cost are greatly reduced.
For example: zhengyun Xu et al, in an article named Application of Deep Convolution Neural Network in Crack Identification, propose a method for identifying bridge cracks using a Deep Convolution Neural Network, wherein the method uses a basic framework based on four Deep Convolution Neural networks to reconstruct a classifier thereof; it is a classification task that distinguishes cracked and non-cracked images by a convolutional neural network. However, the convolutional network adopted in the article has the problems of more parameters and poorer real-time running speed, and the convolutional network based on the classifier can only judge whether cracks exist and can not clearly provide positioning information of the cracks. For example, cheng Yang et al discloses a method for detecting cracks of a concrete structure bridge by using YOLO v3 in an article named Structural crack detection and Recognition Based on Deep Learning, and the method realizes bridge crack detection by training an open-source YOLO v3 model through a self-established data set. However, the characteristics of the YOLO v3 model are extracted by a plurality of network parameters, which makes it difficult to meet the requirement of real-time operation, and has the defects of low detection precision, low detection speed and incapability of meeting the requirement of timeliness of industrial application.
Disclosure of Invention
The invention aims to overcome the problems of inaccuracy, incompleteness and low efficiency in the existing crack detection method based on machine learning, and provides a method and equipment for detecting cracks on the surface of a concrete bridge through a lightweight convolutional network.
In order to achieve the above purpose, the invention provides the following technical scheme:
a bridge surface crack detection method of a lightweight convolution network comprises the following steps:
constructing a lightweight convolutional network model; wherein the lightweight convolutional neural network is a MobileNet-based YOLOv4 network;
acquiring a bridge crack image data set, and training the lightweight convolution network model based on the bridge crack image data set to obtain a trained crack detection network;
and detecting the bridge cracks in real time by using the crack detection network.
According to a specific embodiment, in the method for detecting cracks on a bridge surface using a lightweight convolutional network, the yollov 4 network based on MobileNet includes: the MobileNet sub-network, the SPP sub-network, the PANet sub-network and the Yolohead sub-network are connected in sequence.
According to a specific embodiment, in the method for detecting a crack on a bridge surface using a lightweight convolutional network, training the lightweight convolutional network model based on the bridge crack image dataset includes:
acquiring an open source image data set, and training the lightweight convolution network model based on the open source image data set to obtain a migration model;
and training the migration model based on the bridge crack image data set to obtain the trained crack detection model.
According to a specific implementation mode, in the method for detecting the bridge surface crack of the lightweight convolutional network, an SGD algorithm is adopted to train the migration model based on the bridge crack image data set.
According to a specific embodiment, in the method for detecting a crack on a bridge surface using a lightweight convolutional network, a CIOU function is used as a loss function in the training stage of the migration model.
According to a specific embodiment, in the method for detecting a crack on a bridge surface using a lightweight convolution network, the acquiring a bridge crack image dataset includes:
collecting a plurality of bridge crack pictures, and marking the collected bridge crack pictures to obtain a first bridge crack initial image dataset;
and adopting an image data augmentation method to perform data set augmentation on the bridge crack initial image data set to obtain the bridge crack image data set.
According to a specific embodiment, in the method for detecting cracks on a bridge surface using a lightweight convolutional network, the method for amplifying image data includes: like mirroring, translation, scaling, rotation, clipping, gaussian noise.
In another aspect of the present invention, an electronic device is provided, which includes at least one processor, and a memory communicatively coupled to the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the above-described method of bridge surface crack detection for a lightweight convolutional network.
Compared with the prior art, the invention has the following beneficial effects:
1. the method provided by the embodiment of the invention realizes training of the MobileNet-based YOLOv4 network on the acquired bridge crack image data set by constructing the MobileNet-based YOLOv4 network and based on the light weight of the MobileNet YOLOv4 network, and utilizes the trained network to detect the bridge cracks in real time; the method can effectively overcome the defects of low detection precision and low detection speed of the existing crack detection method based on machine learning.
2. The method provided by the embodiment of the invention effectively overcomes the defects of numerous parameters, difficulty in adjustment and deployment and low training efficiency of the existing crack detection method based on machine learning through transfer learning.
Drawings
Fig. 1 is a flowchart of a bridge surface crack detection method of a lightweight convolutional network according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a YOLOv4 network architecture according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a training platform according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a crack detection result using a lightweight convolutional network according to an embodiment of the present invention;
fig. 5 is a structural block diagram of a bridge surface crack detection device of a lightweight convolutional network according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to test examples and specific embodiments. It should be understood that the scope of the above-described subject matter is not limited to the following examples, and any techniques implemented based on the disclosure of the present invention are within the scope of the present invention.
Example 1
Fig. 1 shows a bridge surface crack detection method of a lightweight convolutional network according to an exemplary embodiment of the present invention, including:
constructing a lightweight convolution network model; wherein the lightweight convolutional neural network is a MobileNet-based YOLOv4 network;
acquiring a bridge crack image data set, and training the lightweight convolution network model based on the bridge crack image data set to obtain a trained crack detection network;
and detecting the bridge cracks in real time by using the crack detection network.
The method provided by the embodiment of the invention realizes training of the MobileNet-based YOLOv4 network on the acquired bridge crack image data set by constructing the MobileNet-based YOLOv4 network and realizing real-time detection of bridge cracks by using the trained network; the method can effectively overcome the defects of low detection precision and low detection speed of the existing crack detection method based on machine learning.
Example 2
In one possible implementation, the yollov 4 network based on MobileNet includes: the mobile network sub-network, the SPP sub-network and the PANet sub-network are connected in sequence.
Specifically, as shown in fig. 2, the entire network structure of the standard YOLO v4 network can be divided into three parts, including: CSPDarkNet, SPP, PANet, yolohead. Wherein, CSPDarkNet is a trunk extraction network, SPP and PANET are reinforced feature extraction networks; the YoloHead is a prediction network and is used for predicting the obtained characteristics.
It is understood that MobileNet, a classical network of lightweight networks, uses deep separable convolutions that are approximately 9 times less computationally intensive than standard convolutions and can be used for classification, with the main part acting as feature extraction. Therefore, in this embodiment, a MobileNet series network is used to replace CSPdarknet53 in Yolov4 for feature extraction, and feature layers with the same shape in three preliminary effective feature layers are subjected to enhanced feature extraction, so that the MobileNet series can be replaced in Yolov4, and the Yolov4 is light. By using the MobileNet as the feature extraction network, the model is greatly reduced compared with the original model. The original model is 245M, and the model provided by the embodiment of the invention is only 53.7M. In addition, the detection speed is greatly improved, and the FSP is greatly improved. In this embodiment, mobileNet v3 is selected as the deep separable convolution sub-network of this embodiment.
In one possible implementation, the training the lightweight convolutional network model based on the bridge crack image dataset includes:
acquiring an open source image data set, and training the lightweight convolution network model based on the open source image data set to obtain a migration model;
and training the migration model based on the bridge crack image data set to obtain the trained crack detection model.
After the model is constructed, pre-training the model based on the open-source data set to obtain a migration model; the hardware platform for model training is shown in fig. 3. And training the migration model based on the bridge crack image data set after the pre-training is finished to obtain the trained crack detection model.
Specifically, the training weights are pre-trained on the public data set by using a transfer learning method; the training parameters were as follows:
1. input image size: 416, 416;
2.anchors_mask＝[[6,7,8],[3,4,5],[0,1,2]]；
3. using CUDA;
4. the training is divided into two phases, namely a freezing phase and a thawing phase:
Init_Epoch＝0；
Freeze_Epoch＝50；
Freeze_batch_size＝16；
UnFreeze_Epoch 300；
Unfreeze_batch_size＝8；
5. using sgd optimization, the parameters used in the optimizer:
momentum＝0.937；
weight_decay＝5e-4；
6. learning rate reduction mode: cos;
7. loss function CIOU:
in a possible implementation manner, in the network training process, a clustering algorithm of a priori frame and a mosaic data enhancement algorithm are adopted to improve the training rate of the model and accelerate the convergence rate of the model.
In a possible implementation manner, the acquiring a bridge crack image dataset includes:
collecting a plurality of bridge crack pictures, and marking the collected bridge crack pictures to obtain a first bridge crack initial image dataset;
and adopting an image data augmentation method to perform data set augmentation on the bridge crack initial image data set to obtain the bridge crack image data set.
Specifically, crack image data of a multi-concrete bridge and bridge image data without cracks are collected through image collection equipment, and a preliminary image data set is established; collecting and sorting the preliminary image data set to obtain a sample set which is the original data to be processed; the sample capacity in the preliminary image dataset is greater than 600 images;
then, carrying out data annotation on each image in the preliminary image dataset by taking positioning areas such as bridge cracks as label contents (namely, labels with cracks) to obtain an annotated preliminary bridge image dataset;
finally, adopting a plurality of image data augmentation methods to expand the marked preliminary image data set to obtain a final image data set; the image data augmentation method includes: image mirroring, translation, scaling, rotation, clipping, gaussian noise, and the like; wherein the image data in the final image dataset is greater than 2000; according to the following steps of 8:2, dividing the image into a training data set and a test data set to complete the construction of the image data set.
Example 3
In this embodiment, taking a large concrete bridge as an example, fig. 4 shows a crack detection result diagram of the method provided by the embodiment of the present invention, and it can be seen from the diagram that the method provided by the embodiment of the present invention can clearly and accurately locate a bridge crack. Further, based on the same image data set, the detection result of the method is compared with the monitoring result of the traditional model, and the comparison result is as follows:
TABLE 1 comparison of crack detection results for different models
| Method | Precision(％) | Module Size(MB) | FPS(f/s) |
| Faser RCNN | 78.21 | 109 | 23.7 |
| SSD | 76.65 | 91.6 | 86.5 |
| CenterNet | 72.13 | 124 | 75.6 |
| YOLO v3 | 82.46 | 235 | 41.6 |
| YOLO v4 | 84.26 | 245 | 44.7 |
| YOLO v5 | 81.39 | 176 | 49.2 |
| Ours | 89.13 | 53.7 | 120.9 |
According to the table, the size of the detection model provided by the embodiment of the invention is reduced by 50% -80% compared with the existing model, the FPS of the model is improved by multiple times compared with the existing model, and meanwhile, the accuracy of the model can be equal to or slightly higher than that of the existing model.
In another aspect of the present invention, as shown in fig. 5, there is also provided an electronic device, which includes a processor, a network interface, and a memory, where the processor, the network interface, and the memory are connected to each other, where the memory is used to store a computer program, and the computer program includes program instructions, and the processor is configured to call the program instructions to execute the above search optimization method.
In another aspect of the present invention, a computer storage medium is provided, in which program instructions are stored, and the program instructions are executed by at least one processor, and the method for guiding a vehicle yard to actively open an electronic invoice is provided.
In an embodiment of the invention, the processor may be an integrated circuit chip having signal processing capabilities. The Processor may be a general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, discrete hardware component.
The various methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The processor reads the information in the storage medium and completes the steps of the method in combination with the hardware.
The storage medium may be a memory, for example, which may be volatile memory or nonvolatile memory, or may include both volatile and nonvolatile memory.
The nonvolatile Memory may be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable PROM (EEPROM), or a flash Memory.
The volatile Memory may be a Random Access Memory (RAM) which serves as an external cache. By way of example, and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), SLDRAM (SLDRAM), and Direct Rambus RAM (DRRAM).
The storage media described in connection with the embodiments of the invention are intended to comprise, without being limited to, these and any other suitable types of memory.
It should be understood that the disclosed system may be implemented in other ways. For example, the division of the modules into only one logical functional division may be implemented in practice in other ways, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not implemented. In addition, the communication connection between the modules may be an indirect coupling or communication connection between servers or units through some interfaces, and may be electrical or in other forms.
In addition, functional modules in the embodiments of the present invention may be integrated into one processing unit, or each module may exist alone physically, or two or more modules are integrated into one processing unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention, which is substantially or partly contributed by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.
The above description is intended to be illustrative of the preferred embodiment of the present invention and should not be taken as limiting the invention, but rather, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.
The above description is intended to be illustrative of the preferred embodiment of the present invention and should not be taken as limiting the invention, but rather, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.
Claims (8)
1. A bridge surface crack detection method of a lightweight convolution network is characterized by comprising the following steps:
constructing a lightweight convolution network model; wherein the lightweight convolutional neural network is a MobileNet-based YOLOv4 network;
acquiring a bridge crack image data set, and training the lightweight convolution network model based on the bridge crack image data set to obtain a trained crack detection network;
and detecting the bridge cracks in real time by using the crack detection network.
2. The method for detecting cracks on the bridge surface of a light-weight convolutional network as claimed in claim 1, wherein the MobileNet-based YOLOv4 network comprises: the MobileNet sub-network, the SPP sub-network, the PANet sub-network and the YoloHead sub-network are connected in sequence.
3. The method of claim 1, wherein training the lightweight convolutional network model based on the bridge crack image dataset comprises:
acquiring an open source image data set, and training the lightweight convolution network model based on the open source image data set to obtain a migration model;
and training the migration model based on the bridge crack image data set to obtain the trained crack detection model.
4. The method for detecting the cracks on the bridge surface of the light-weight convolutional network as claimed in claim 3, wherein the migration model is trained based on the bridge crack image dataset by adopting an SGD algorithm.
5. The method for detecting the cracks on the bridge surface of the light-weight convolutional network as claimed in claim 2, wherein a CIOU function is used as a loss function in the training phase of the migration model.
6. The method for detecting cracks on a bridge surface using a light-weight convolutional network as claimed in any of claims 1-5, wherein the acquiring of the bridge crack image dataset comprises:
collecting a plurality of bridge crack pictures, and labeling the collected bridge crack pictures to obtain a first bridge crack initial image data set;
and adopting an image data augmentation method to perform data set augmentation on the bridge crack initial image data set to obtain the bridge crack image data set.
7. The method for detecting cracks on the bridge surface of a light-weight convolutional network of claim 6, wherein the image data augmentation method comprises: like mirroring, translation, scaling, rotation, clipping, gaussian noise.
8. An electronic device comprising at least one processor, and a memory communicatively coupled to the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of bridge surface crack detection for a lightweight convolutional network of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210936787.7A CN115439412A (en) | 2022-08-05 | 2022-08-05 | Bridge surface crack detection method and device of lightweight convolutional network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210936787.7A CN115439412A (en) | 2022-08-05 | 2022-08-05 | Bridge surface crack detection method and device of lightweight convolutional network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115439412A true CN115439412A (en) | 2022-12-06 |
Family
ID=84243497
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210936787.7A Pending CN115439412A (en) | 2022-08-05 | 2022-08-05 | Bridge surface crack detection method and device of lightweight convolutional network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115439412A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117036348A (en) * | 2023-10-08 | 2023-11-10 | 中国石油大学(华东) | Metal fatigue crack detection method based on image processing and crack recognition model |
-
2022
- 2022-08-05 CN CN202210936787.7A patent/CN115439412A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117036348A (en) * | 2023-10-08 | 2023-11-10 | 中国石油大学(华东) | Metal fatigue crack detection method based on image processing and crack recognition model |
| CN117036348B (en) * | 2023-10-08 | 2024-01-09 | 中国石油大学(华东) | Metal fatigue crack detection method based on image processing and crack recognition model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Automatic defect detection of metro tunnel surfaces using a vision-based inspection system | |
| US20210319561A1 (en) | Image segmentation method and system for pavement disease based on deep learning | |
| Sony et al. | A systematic review of convolutional neural network-based structural condition assessment techniques | |
| Pereira et al. | A deep learning-based approach for road pothole detection in timor leste | |
| CN110222701B (en) | Automatic bridge disease identification method | |
| Wang et al. | RENet: Rectangular convolution pyramid and edge enhancement network for salient object detection of pavement cracks | |
| CN113378763B (en) | SAR image-oriented target automatic detection method and system | |
| CN110931112B (en) | Brain medical image analysis method based on multi-dimensional information fusion and deep learning | |
| Li et al. | Automatic bridge crack identification from concrete surface using ResNeXt with postprocessing | |
| CN114359130A (en) | Road crack detection method based on unmanned aerial vehicle image | |
| Pereira et al. | Semantic segmentation of paved road and pothole image using u-net architecture | |
| CN115719475B (en) | Three-stage trackside equipment fault automatic detection method based on deep learning | |
| Li et al. | Pixel-level recognition of pavement distresses based on U-Net | |
| CN115457277A (en) | Intelligent pavement disease identification and detection method and system | |
| CN115439412A (en) | Bridge surface crack detection method and device of lightweight convolutional network | |
| CN116206112A (en) | Remote sensing image semantic segmentation method based on multi-scale feature fusion and SAM | |
| Wang et al. | Automatic identification and location of tunnel lining cracks | |
| CN113902793A (en) | End-to-end building height prediction method and system based on single vision remote sensing image and electronic equipment | |
| CN113487638A (en) | Ground feature edge detection method of high-precision semantic segmentation algorithm U2-net | |
| CN111179278B (en) | Image detection method, device, equipment and storage medium | |
| CN116052110B (en) | Intelligent positioning method and system for pavement marking defects | |
| CN110765900B (en) | Automatic detection illegal building method and system based on DSSD | |
| CN111897993A (en) | Efficient target person track generation method based on pedestrian re-recognition | |
| CN115761613B (en) | Automatic tunnel crack detection method based on convolutional network | |
| Wu et al. | Deep learning-based crack monitoring for ultra-high performance concrete (UHPC) |
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 |