CN115830416A - Method and system for identifying and early warning highway equipment facilities - Google Patents
Method and system for identifying and early warning highway equipment facilities Download PDFInfo
- Publication number
- CN115830416A CN115830416A CN202211642303.4A CN202211642303A CN115830416A CN 115830416 A CN115830416 A CN 115830416A CN 202211642303 A CN202211642303 A CN 202211642303A CN 115830416 A CN115830416 A CN 115830416A
- Authority
- CN
- China
- Prior art keywords
- data
- training
- model
- target
- equipment
- 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
-
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method and a system for identifying and early warning highway equipment facilities, and belongs to the technical field of data processing. The method comprises the following steps: acquiring basic data of expressway equipment facilities, and preprocessing the basic data to generate a data set; obtaining a training optimization model for recognizing and early warning highway equipment facilities; acquiring basic data of equipment facilities of a target highway section, preprocessing the basic data of the equipment facilities of the target highway section to obtain target data, inputting the target data into a training optimization model, performing operation to identify the equipment facilities of the target highway section to obtain an identification result, and if the identification result shows that any equipment facility of the target highway section has a fault, sending an early warning. According to the method and the system, the equipment can be identified and the equipment fault can be identified by using the training optimization model and the data of the target road section, so that the road maintenance level can be enhanced.
Description
Technical Field
The present invention relates to the field of data processing technologies, and more particularly, to a method and system for identifying and warning highway equipment facilities.
Background
Highway traffic, as an important component of a transportation system, is playing an increasingly important role in the development of national economy today. The highway usually adopts a bidirectional diversion and full interchange design, and has the characteristics of high driving speed, comfort, environmental protection and the like, so higher requirements are provided for the full guarantee of the highway and the maintenance of pavements and facilities.
The highway road property maintenance management is an important component of highway operation management, is a frequent and long-term work, and is also one of main means for ensuring the excellent service level of the highway. Because the capital of the current highway operation and management unit is limited, the expenditure for highway maintenance is limited, and the contradiction between the limited maintenance capital and the huge maintenance requirement becomes the bottleneck for restricting the improvement of the highway service level.
Disclosure of Invention
In view of the above problems, the present invention provides a method for identifying and warning facilities of highway equipment, comprising:
acquiring basic data of expressway equipment facilities, and preprocessing the basic data to generate a data set;
dividing the data set into a training set and a verification set according to a preset proportion, inputting the training set into a training network based on deep learning for training to obtain a training model and a training set loss function, verifying the training model by using the verification set to obtain the accuracy of the training model and the verification set loss function, and adjusting the parameters of the training model through the training set loss function, the accuracy of the training model and the verification set loss function to obtain a training optimization model for recognizing and early warning of expressway equipment facilities;
acquiring basic data of equipment facilities of a target highway section, preprocessing the basic data of the equipment facilities of the target highway section to obtain target data, inputting the target data into a training optimization model, performing operation to identify the equipment facilities of the target highway section to obtain an identification result, and if the identification result shows that any equipment facility of the target highway section has a fault, sending an early warning.
Optionally, the basic data includes at least one of the following: existing data, public data and field data;
the existing data comprises at least one of the following: missing, portal and camera data;
the disclosed data, comprising: sign plate data;
the field data comprises at least one of the following: the road lamp comprises a protective net, iron guardrails, dan Hulan, a sound insulation screen, a reflector and road lamp data.
Optionally, preprocessing the basic data to generate a data set, including:
and performing preliminary examination on the basic data, removing fuzzy data in the basic data to obtain preliminary examination data, extracting frames of the preliminary examination data to be stored as picture data, performing further examination on the picture data, removing fuzzy data in the picture data to obtain further examination data, converting the data format of the further examination data into a data format suitable for deep learning-based training network training, and generating a data set.
Optionally, a verification set is used to verify the training model, specifically, the accuracy of verification is performed on the detection algorithm of the training model.
In another aspect, the present invention further provides a system for identifying and warning an expressway equipment facility, including:
the system comprises a data acquisition unit, a data processing unit and a data processing unit, wherein the data acquisition unit is used for acquiring basic data of expressway equipment facilities and preprocessing the basic data to generate a data set;
the model training unit is used for dividing the data set into a training set and a verification set according to a preset proportion, inputting the training set into a training network based on deep learning for training to obtain a training model and a training set loss function, verifying the training model by using the verification set to obtain the accuracy of the training model and the verification set loss function, and adjusting the parameters of the training model through the training set loss function, the accuracy of the training model and the verification set loss function to obtain a training optimization model for identifying and early warning of the expressway equipment facilities;
the identification early warning unit is used for acquiring basic data of equipment facilities of a target highway section, preprocessing the basic data of the equipment facilities of the target highway section to obtain target data, inputting the target data into a training optimization model, performing operation to identify the equipment facilities of the target highway section to obtain an identification result, and sending out early warning if the identification result shows that any equipment facility of the target highway section has a fault.
Optionally, the basic data includes at least one of the following: existing data, public data and field data;
the existing data comprises at least one of the following: missing, portal and camera data;
the disclosed data, comprising: sign plate data;
the field data comprises at least one of the following: the road lamp comprises a protective net, iron guardrails, dan Hulan, a sound insulation screen, a reflector and road lamp data.
Optionally, the data acquisition unit preprocesses the basic data to generate a data set, including:
and performing preliminary examination on the basic data, removing fuzzy data in the basic data to obtain preliminary examination data, extracting frames of the preliminary examination data to be stored as picture data, performing further examination on the picture data, removing fuzzy data in the picture data to obtain further examination data, converting the data format of the further examination data into a data format suitable for deep learning-based training network training, and generating a data set.
Optionally, the model training unit verifies the training model by using a verification set, specifically, verifies the accuracy of the detection algorithm of the training model.
In yet another aspect, the present invention also provides a computing device comprising: one or more processors;
a processor for executing one or more programs;
when executed by the one or more processors, implement the methods as described above.
In yet another aspect, the invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed, performs the method as described above.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a method for identifying and early warning highway equipment facilities, which comprises the following steps: acquiring basic data of expressway equipment facilities, and preprocessing the basic data to generate a data set; dividing the data set into a training set and a verification set according to a preset proportion, inputting the training set into a training network based on deep learning for training to obtain a training model and a training set loss function, verifying the training model by using the verification set to obtain the accuracy of the training model and the verification set loss function, and adjusting the parameters of the training model by the training set loss function, the accuracy of the training model and the verification set loss function to obtain a training optimization model for identifying and early warning of expressway equipment facilities; acquiring basic data of equipment facilities of a target highway section, preprocessing the basic data of the equipment facilities of the target highway section to obtain target data, inputting the target data into a training optimization model, performing operation to identify the equipment facilities of the target highway section to obtain an identification result, and if the identification result shows that any equipment facility of the target highway section has a fault, sending an early warning. According to the invention, the equipment can be identified and the equipment fault can be identified by using the training optimization model and the data of the target road section, so that the road maintenance level can be enhanced.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of a method of the present invention;
fig. 3 is a block diagram of the system of the present invention.
Detailed Description
The exemplary embodiments of the present invention will now be described with reference to the accompanying drawings, however, the present invention may be embodied in many different forms and is not limited to the embodiments described herein, which are provided for complete and complete disclosure of the present invention and to fully convey the scope of the present invention to those skilled in the art. The terms used in the exemplary embodiments shown in the drawings are not intended to limit the present invention. In the drawings, the same units/elements are denoted by the same reference numerals.
Unless otherwise defined, terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Further, it will be understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense.
Example 1:
the invention provides a method for identifying and early warning highway equipment facilities, which comprises the following steps as shown in figure 1:
step 1, acquiring basic data of highway equipment facilities, and preprocessing the basic data to generate a data set;
step 2, dividing the data set into a training set and a verification set according to a preset proportion, inputting the training set into a training network based on deep learning for training to obtain a training model and a training set loss function, verifying the training model by using the verification set to obtain the accuracy of the training model and the verification set loss function, and adjusting the parameters of the training model by the training set loss function, the accuracy of the training model and the verification set loss function to obtain a training optimization model for identifying and early warning of expressway equipment facilities;
and 3, acquiring basic data of the equipment facilities of the target highway section, preprocessing the basic data of the equipment facilities of the target highway section to obtain target data, inputting the target data into a training optimization model, performing operation to identify the equipment facilities of the target highway section to obtain an identification result, and sending out an early warning if the identification result shows that any equipment facility of the target highway section has a fault.
Wherein, the basic data comprises at least one of the following data: existing data, public data and field data;
the existing data comprises at least one of the following: missing, portal and camera data;
the disclosed data, comprising: sign plate data;
the field data comprises at least one of the following: the road lamp comprises a protective net, iron guardrails, dan Hulan, a sound insulation screen, a reflector and road lamp data.
Wherein preprocessing the base data to generate a data set comprises:
and performing preliminary examination on the basic data, removing fuzzy data in the basic data to obtain preliminary examination data, extracting frames of the preliminary examination data to be stored as picture data, performing further examination on the picture data, removing fuzzy data in the picture data to obtain further examination data, converting the data format of the further examination data into a data format suitable for deep learning-based training network training, and generating a data set.
And verifying the training model by using a verification set, specifically verifying the accuracy of a detection algorithm of the training model.
The invention is further illustrated by the following examples:
the principle of the present invention is shown in fig. 2, and comprises:
1.1 acquiring data comprising:
1.1.1 existing data set
The category: left scattering, portal frame and camera.
Some of the data, including several of the above objectives, is screened from the company's existing data set.
1.1.2 public data sets
The category: and (4) marking a label.
The disclosed label tag data is used to enrich the data set, increasing the diversity of the data set.
1.1.3 on-site Collection
And aiming at the highway environment, taking the highway monitoring video and the historical AI project identification result as sources, collecting field materials, selecting the video with the asset target from the highway monitoring video, and downloading continuous videos. Historical AI project recognition results, such as road scattering in an expressway scene and road asset picture results, are collected and summarized as training material sources.
The category: the road lamp comprises a protective net, iron guardrails, dan Hulan, a sound insulation screen, a reflector and a road lamp.
Available video data is downloaded from a video source provided by a client. And processing the collected video, storing the video frames as pictures, and storing one picture in each 100 frames. The collected video mainly includes the above categories.
1.1.4 generating data sets
The category: losing and scattering
The scattered data set is monotonous and the data volume is small. To enrich the data set, "spill" pictures and pictures of the scene (highway) are collected. The "scattering" target is attached to the proper position of the picture using logic such as the relative position of objects in the picture.
1.2 data processing
And (3) data processing:
data collection: according to the requirements, the material collection tends to the real environment of the expressway;
data arrangement: classifying and sorting the materials according to asset classes and road disease classes;
and (3) screening data: according to algorithm training requirements, whether the material set has asset and disease characteristics is discriminated, fuzzy data without training conditions and characteristics are removed, and effective data are reserved;
data annotation: and labeling or converting the data set which accords with the algorithm training so as to enable the data set to meet the standard of the algorithm training input data set.
And (5) sorting, screening and marking the data. First, the collected data is examined, and some fuzzy, unclear and unsuitable data are deleted. And (3) extracting frames of the collected video to be stored as pictures, screening the pictures after the pictures are generated, and deleting the pictures which cannot be used (are fuzzy). For the public data set with labels, the labels are converted into a format required by model training. And then carrying out feature labeling on the data needing to be labeled. After the data labeling is completed, the labels and the data are converted into a format suitable for model training.
2.1 Algorithm selection of training models
2.1.1 feature analysis
The algorithm needs to perform the functions of determining the position of the target in the video picture and outputting the category. First, a target detection model is selected. And carrying out characteristic analysis on the target to be detected. All objects are small parts of the picture in terms of spatial features, and each object is independent. The light reflecting plate, the door frame, the street lamp, the camera and the sign plate are vertical in the picture and occupy small area in the whole picture. The target detection frame is used for detection. For the guard net, iron guard rail, dan Hulan, several categories of sound barriers are inclined in the picture and long in length. If the target detection frame is used for detection, a lot of backgrounds are framed in the frame, and the backgrounds in the picture are complex. And the target frame occupies a large proportion of the whole picture, almost one frame can occupy 2/3 of the whole picture, so the segmentation model is selected and used by the several categories.
For the crack category, its shape is also very oblique, but its overall size is not large and is a road surface around its part, the background in the drawn frame is not complex. And the crack characteristics are not obvious, the whole volume is too small, and the method is not suitable for segmentation, so that a target detection method is preferentially selected.
2.1.2 training Algorithm selection
Firstly, a detection algorithm is selected, the environment on the highway is complex and various, and the variety of targets to be detected is various. According to introduction of a target detection algorithm and performance of a classical algorithm on a public data set, the deep learning-based target detection algorithm is adopted to automatically learn characteristics through a deep neural network, and the method can be well suitable for a complex high-speed environment. The YOLO algorithm uses a regression method, so that high accuracy can be achieved while rapid detection is realized. The scene on the highway is complex, the objects appearing in the picture are various, the YOLO adopts the full image information to predict, and the full image information is utilized in the training and predicting processes, so that the background block can be prevented from being detected as a target as much as possible. According to past project experience, several algorithms of the YOLO series are compared, and finally yolov4-tiny is selected to perform a target detection task.
ERFNet is selected as the segmentation algorithm, ERFnet is a lightweight real-time semantic segmentation algorithm, the parameter quantity is small, the processing time is short, accurate semantic segmentation can be provided while real-time operation is realized, and the algorithm can fuse the features of different layers to obtain more detailed output. The algorithm is selected.
2.2 Algorithm training
2.2.1 training tool
NVIDIA GeForce RTX 3090。
2.2.2 training Process
2.2.2.1 dataset partitioning
And dividing the data set into a training set and a verification set according to the proportion of 9:1, and arranging the training set and the verification set into an input format required by the model.
2.2.2.2 Algorithm training
And (3) sending the labeled data into a CNN (computer network-based neural network) training network, wherein the CNN network is used for comparison and training, and the GAN can actively generate more complex samples for learning and training according to the characteristics of the samples on the basis, and then continuously tuning and iterating to obtain a better training model.
Gan training network is a cyclic gaming system that combines the detection and recognition of networks with the generation of countermeasure networks. The great advantage is that development work can be started based on a limited number of sample images. The GAN training network working process is a circulating game process; firstly, the CNN network automatically detects target information in a sample picture through an algorithm model obtained by original training, such as: equipment facilities, sign lines, signs. The GAN network can take the output of the CNN network as input, generate the countermeasure principle by using the output, and simulate and produce a picture which is close to reality and contains a plurality of key elements by adjusting different parameters according to the results; the system can input the generated false pictures into the CNN network again for identification and authentication, and if the CNN network can not identify the false pictures, the pictures are returned to be regenerated again, and then the input is fed back again until the CNN system considers that the generated pictures can complete the process with a false degree. In this way, by repeating the steps, a plurality of or N generated fake sample pictures can be obtained through one original picture, so that the problems of scarcity of part of samples or high acquisition cost can be solved.
The output model provides a good basis for the production system to apply the model for detection and analysis.
2.2.2.3 parameter optimization
Parameters are adjusted by observing the training set loss function and the validation set accuracy.
The parameters mainly adjusted in the model training.
(1) Learning rate: the convergence of the model is controlled by adjusting the learning rate, and a dynamic learning rate conversion mode is adopted.
(2) An Epoch: observing the convergence condition of the model to adjust the Epoch, and stopping iteration if the loss of the verification sets of the continuous rounds is not reduced correspondingly. Overfitting is also prevented by controlling Epoch.
(3) And adjusting the size of the Batch _ size according to the video memory and data volume. The utilization rate of the memory and the processing speed of the data are improved, the gradient descending direction is more accurate, and the local optimum is prevented from falling.
2.2.2.4 model class adjustment
When the model training is started, the same detection model is used for training the categories of the reflector, the portal, the street lamp, the camera and the sign plate, which are scattered.
2.3 Algorithm testing
The system firstly preprocesses the images to enable the images to meet the basic requirements of detection and analysis.
And then, attention contents such as guardrails, slopes, pavements and the like in the video images are automatically extracted through deep learning, and are automatically compared and analyzed with the training model.
And (4) marking images and pictures which are found to be abnormal in the comparison analysis, identifying information such as abnormal points, positions, time and the like, and providing maintenance basic data for a highway maintenance management unit.
(1) Testing a validation set
For yolov4-tiny target detection algorithm: and observing the effect of the model by verifying indexes such as accuracy, recall rate, AP and mAP of the set.
For the ERFNet segmentation algorithm: the model effect is observed by verifying the mIOU of the set.
(2) Testing a test set
And collecting a small part of data as a test set, testing the test set by using the trained model, visualizing the result, and observing the test effect.
Example 2:
the present invention also provides a system 200 for highway equipment facility identification and early warning, as shown in fig. 3, comprising:
a data acquisition unit 201, configured to acquire basic data of an expressway equipment facility, and preprocess the basic data to generate a data set;
the model training unit 202 is used for dividing the data set into a training set and a verification set according to a preset proportion, inputting the training set into a training network based on deep learning for training to obtain a training model and a training set loss function, verifying the training model by using the verification set to obtain the accuracy of the training model and the verification set loss function, and adjusting the parameters of the training model by the training set loss function, the accuracy of the training model and the verification set loss function to obtain a training optimization model for identifying and early warning of the expressway equipment facilities;
the identification early warning unit 203 is used for acquiring basic data of equipment facilities of a target highway section, preprocessing the basic data of the equipment facilities of the target highway section to obtain target data, inputting the target data into a training optimization model, performing operation to identify the equipment facilities of the target highway section to obtain an identification result, and sending out early warning if the identification result shows that any equipment facility of the target highway section has a fault.
Wherein, the basic data comprises at least one of the following data: existing data, public data and field data;
the existing data comprises at least one of the following: missing, portal and camera data;
the disclosed data includes: sign plate data;
the field data comprises at least one of the following: the road lamp comprises a protective net, iron guardrails, dan Hulan, a sound insulation screen, a reflector and road lamp data.
Wherein the data acquisition unit 201 preprocesses the basic data to generate a data set, comprising:
and performing preliminary examination on the basic data, removing fuzzy data in the basic data to obtain preliminary examination data, extracting frames of the preliminary examination data to be stored as picture data, performing further examination on the picture data, removing fuzzy data in the picture data to obtain further examination data, converting the data format of the further examination data into a data format suitable for deep learning-based training network training, and generating a data set.
The model training unit 202 uses a verification set to verify the training model, specifically, to verify the accuracy of the detection algorithm of the training model.
According to the method and the system, the equipment can be identified and the equipment fault can be identified by using the training optimization model and the data of the target road section, so that the road maintenance level can be enhanced.
Example 3:
based on the same inventive concept, the present invention also provides a computer apparatus comprising a processor and a memory, the memory being configured to store a computer program comprising program instructions, the processor being configured to execute the program instructions stored by the computer storage medium. The Processor may be a Central Processing Unit (CPU), or may be other 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, etc., which is a computing core and a control core of the terminal, and is specifically adapted to load and execute one or more instructions in a computer storage medium to implement a corresponding method flow or a corresponding function, so as to implement the steps of the method in the above embodiments.
Example 4:
based on the same inventive concept, the present invention further provides a storage medium, in particular a computer readable storage medium (Memory), which is a Memory device in a computer device and is used for storing programs and data. It is understood that the computer readable storage medium herein can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space storing an operating system of the terminal. Also, one or more instructions, which may be one or more computer programs (including program code), are stored in the memory space and are adapted to be loaded and executed by the processor. It should be noted that the computer readable storage medium may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk memory. One or more instructions stored in a computer-readable storage medium may be loaded and executed by a processor to implement the steps of the method in the above-described embodiments.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The scheme in the embodiment of the invention can be realized by adopting various computer languages, such as object-oriented programming language Java and transliterated scripting language JavaScript.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (10)
1. A method for highway equipment identification and forewarning, the method comprising:
acquiring basic data of expressway equipment facilities, and preprocessing the basic data to generate a data set;
dividing the data set into a training set and a verification set according to a preset proportion, inputting the training set into a training network based on deep learning for training to obtain a training model and a training set loss function, verifying the training model by using the verification set to obtain the accuracy of the training model and the verification set loss function, and adjusting the parameters of the training model by the training set loss function, the accuracy of the training model and the verification set loss function to obtain a training optimization model for identifying and early warning of expressway equipment facilities;
acquiring basic data of equipment facilities of a target highway section, preprocessing the basic data of the equipment facilities of the target highway section to obtain target data, inputting the target data into a training optimization model, performing operation to identify the equipment facilities of the target highway section to obtain an identification result, and if the identification result shows that any equipment facility of the target highway section has a fault, sending an early warning.
2. The method of claim 1, wherein the basic data comprises at least one of: existing data, public data and field data;
the existing data includes at least one of the following: missing, portal and camera data;
the disclosed data, comprising: sign data;
the field data comprises at least one of the following: the road lamp comprises a protective net, iron guardrails, dan Hulan, a sound insulation screen, a reflector and road lamp data.
3. The method of claim 1, wherein preprocessing the base data to generate a data set comprises:
and performing preliminary examination on the basic data, removing fuzzy data in the basic data to obtain preliminary examination data, extracting frames of the preliminary examination data to store as picture data, performing further examination on the picture data, removing fuzzy data in the picture data to obtain further examination data, converting the data format of the further examination data into a data format suitable for deep learning-based training network training, and generating a data set.
4. The method according to claim 1, wherein the training model is validated using a validation set, in particular a validation accuracy of a detection algorithm of the training model.
5. A system for highway equipment identification and forewarning, the system comprising:
the system comprises a data acquisition unit, a data processing unit and a data processing unit, wherein the data acquisition unit is used for acquiring basic data of expressway equipment facilities and preprocessing the basic data to generate a data set;
the model training unit is used for dividing the data set into a training set and a verification set according to a preset proportion, inputting the training set into a training network based on deep learning for training to obtain a training model and a training set loss function, verifying the training model by using the verification set to obtain the accuracy of the training model and the verification set loss function, and adjusting the parameters of the training model through the training set loss function, the accuracy of the training model and the verification set loss function to obtain a training optimization model for identifying and early warning of the expressway equipment facilities;
the system comprises an identification early warning unit and a warning unit, wherein the identification early warning unit is used for acquiring basic data of equipment facilities on a target highway section, preprocessing the basic data of the equipment facilities on the target highway section to obtain target data, inputting the target data into a training optimization model, performing operation to identify the equipment facilities on the target highway section to obtain an identification result, and giving an early warning if the identification result shows that any equipment facility on the target highway section has a fault.
6. The system of claim 5, wherein the base data comprises at least one of: existing data, public data and field data;
the existing data comprises at least one of the following: missing, portal and camera data;
the disclosed data, comprising: sign plate data;
the field data comprises at least one of the following: the road lamp comprises a protective net, iron guardrails, dan Hulan, a sound insulation screen, a light reflecting plate and road lamp data.
7. The system of claim 5, wherein the data acquisition unit preprocesses the base data to generate a data set, comprising:
and performing preliminary examination on the basic data, removing fuzzy data in the basic data to obtain preliminary examination data, extracting frames of the preliminary examination data to store as picture data, performing further examination on the picture data, removing fuzzy data in the picture data to obtain further examination data, converting the data format of the further examination data into a data format suitable for deep learning-based training network training, and generating a data set.
8. The system according to claim 5, wherein the model training unit verifies the training model using a verification set, in particular a verification accuracy of a detection algorithm of the training model.
9. A computer device, comprising:
one or more processors;
a processor for executing one or more programs;
the one or more programs, when executed by the one or more processors, implement the method of any of claims 1-4.
10. A computer-readable storage medium, having stored thereon a computer program which, when executed, implements the method of any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211642303.4A CN115830416A (en) | 2022-12-20 | 2022-12-20 | Method and system for identifying and early warning highway equipment facilities |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211642303.4A CN115830416A (en) | 2022-12-20 | 2022-12-20 | Method and system for identifying and early warning highway equipment facilities |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115830416A true CN115830416A (en) | 2023-03-21 |
Family
ID=85517104
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211642303.4A Pending CN115830416A (en) | 2022-12-20 | 2022-12-20 | Method and system for identifying and early warning highway equipment facilities |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115830416A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116758259A (en) * | 2023-04-26 | 2023-09-15 | 中国公路工程咨询集团有限公司 | Highway asset information identification method and system |
| CN117351708A (en) * | 2023-10-08 | 2024-01-05 | 北京迈道科技有限公司 | Expressway safety operation management early warning method, system and storage medium |
-
2022
- 2022-12-20 CN CN202211642303.4A patent/CN115830416A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116758259A (en) * | 2023-04-26 | 2023-09-15 | 中国公路工程咨询集团有限公司 | Highway asset information identification method and system |
| CN117351708A (en) * | 2023-10-08 | 2024-01-05 | 北京迈道科技有限公司 | Expressway safety operation management early warning method, system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110364008B (en) | Road condition determining method and device, computer equipment and storage medium | |
| CN115830416A (en) | Method and system for identifying and early warning highway equipment facilities | |
| CN110188807A (en) | Tunnel pedestrian target detection method based on cascade super-resolution network and improvement Faster R-CNN | |
| Guo et al. | A pavement distresses identification method optimized for YOLOv5s | |
| CN109325429A (en) | A kind of method, apparatus, storage medium and the terminal of linked character data | |
| CN110533046A (en) | A kind of image instance dividing method and device | |
| CN110929806A (en) | Picture processing method and device based on artificial intelligence and electronic equipment | |
| CN108229289A (en) | Target retrieval method, apparatus and electronic equipment | |
| CN109376868A (en) | Information management system | |
| CN110458794A (en) | Fittings quality detection method and device for track train | |
| CN115719475A (en) | Three-stage trackside equipment fault automatic detection method based on deep learning | |
| CN114372532A (en) | Method, device, equipment, medium and product for determining label marking quality | |
| CN111339950B (en) | Remote sensing image target detection method | |
| CN113223200A (en) | Road stealing and escaping intelligent prevention and control method, storage medium and system based on index dimension | |
| CN116824488A (en) | Target detection method based on transfer learning | |
| CN116740495A (en) | Training method and defect detection method for defect detection model of road and bridge tunnel | |
| Khan | Vehicle and pedestrian detection using YOLOv3 and YOLOv4 for self-driving cars | |
| CN110765900A (en) | DSSD-based automatic illegal building detection method and system | |
| CN112529116B (en) | Scene element fusion processing method, device and equipment and computer storage medium | |
| Gao et al. | Detection method of potholes on highway pavement based on yolov5 | |
| CN113936300A (en) | Construction site personnel identification method, readable storage medium and electronic device | |
| CN110659384B (en) | Video structured analysis method and device | |
| CN111767945A (en) | License plate sample generation method and system based on CycleGAN network model | |
| Sekar et al. | A Novel SGD-U-Network-Based Pixel-Level Road Crack Segmentation and Classification | |
| CN116246128B (en) | Training method and device of detection model crossing data sets and electronic equipment |
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 |