CN113780130A - Vehicle type identification method based on magnetic field data - Google Patents
Vehicle type identification method based on magnetic field data Download PDFInfo
- Publication number
- CN113780130A CN113780130A CN202111010856.3A CN202111010856A CN113780130A CN 113780130 A CN113780130 A CN 113780130A CN 202111010856 A CN202111010856 A CN 202111010856A CN 113780130 A CN113780130 A CN 113780130A
- Authority
- CN
- China
- Prior art keywords
- magnetic field
- field data
- vehicle type
- encoder
- depth
- 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
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 238000012706 support-vector machine Methods 0.000 claims abstract description 12
- 238000012549 training Methods 0.000 claims abstract description 12
- 238000000605 extraction Methods 0.000 claims abstract description 6
- 238000012545 processing Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 5
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000013145 classification model Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a vehicle type identification method based on magnetic field data, which specifically comprises the following steps: collecting magnetic field data of different types of vehicles passing by using a magnetic field sensor, and manually marking the magnetic field data to mark the vehicle type; preprocessing the collected and labeled magnetic field data; extracting depth features of the preprocessed magnetic field data by adopting a pre-trained depth self-encoder; training the support vector machine classifier by using the extracted depth features to obtain a trained support vector machine classifier; acquiring magnetic field data of a vehicle to be identified by using a magnetic field sensor, and preprocessing the magnetic field data; and performing depth feature extraction on the preprocessed magnetic field data by adopting a pre-trained depth self-encoder, and performing vehicle type identification by using a trained support vector machine classifier. The invention can realize the recognition of vehicle types in traffic flow only by using the magnetic field sensor, thereby reducing the cost of vehicle type recognition.
Description
Technical Field
The invention relates to a vehicle type identification method based on magnetic field data, and belongs to the technical field of vehicle type identification.
Background
In the field of intelligent transportation, vehicle type identification plays a very critical role in the aspects of violation monitoring, pursuit tracking and the like. Through an automatic vehicle type identification method, a target vehicle can be effectively identified, manpower and material resources are saved, and the efficiency is greatly improved. Past work has implemented vehicle type recognition based on deep learning image processing, but has some limitations. On one hand, the accuracy of vehicle type recognition based on image processing in night and severe weather can be influenced; secondly, vehicle type identification based on image processing needs to transmit data back to a background server, and an identification algorithm is run on the background server, so that the requirements on hardware are high, and the deployment cost is high. Therefore, a new method for identifying the type of the vehicle and reducing the identification cost is needed.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the vehicle type identification method based on the magnetic field data is characterized in that the magnetic field data when a vehicle passes through is collected by a magnetic field sensor, the depth features are extracted by using a data preprocessing method and a depth self-encoder, and a support vector machine is used for vehicle type classification.
The invention adopts the following technical scheme for solving the technical problems:
a vehicle type recognition method based on magnetic field data comprises a model training stage and a vehicle type recognition stage, wherein,
the model training phase comprises the following steps:
step 1, collecting magnetic field data of different types of vehicles passing by using a magnetic field sensor, and manually marking the magnetic field data to mark the types of the vehicles;
step 2, preprocessing the magnetic field data collected and labeled in the step 1;
step 3, extracting the depth characteristics of the magnetic field data preprocessed in the step 2 by adopting a pre-trained depth self-encoder;
step 4, training the support vector machine classifier by using the depth features extracted in the step 3 to obtain a trained support vector machine classifier;
the vehicle type identification stage comprises the following steps:
step 5, collecting magnetic field data of the vehicle to be identified by using a magnetic field sensor, and preprocessing the magnetic field data;
and 6, performing depth feature extraction on the magnetic field data preprocessed in the step 5 by adopting a pre-trained depth self-encoder, and performing vehicle type identification by using a trained support vector machine classifier.
In a preferred embodiment of the present invention, the magnetic field sensor is buried under the ground of a road.
As a preferred scheme of the invention, the pretreatment process in the step 2 and the step 5 is as follows:
1) performing 0-1 standard normalization processing on the magnetic field data;
2) the method for segmenting the magnetic field data window by using the variance threshold method on the magnetic field data after the normalization processing specifically comprises the following steps:
sliding the magnetic field data by using a window with the size of 5000ms, wherein the window step length is 1000 ms; and calculating the mean value mu and the variance sigma of the magnetic field data in each window, and if the mean value mu of the magnetic field data in each window is greater than a threshold value mu 'and the variance sigma is greater than a threshold value sigma', storing the magnetic field data in the window, wherein the mu 'is 0.45, and the sigma' is 8.4.
In a preferred embodiment of the present invention, the depth self-encoder is composed of an encoder and a decoder, and the encoder and the decoder are symmetrical in structure.
As a preferred scheme of the present invention, the depth self-encoder is a fully connected neural network, the dimension of the depth feature is 50, and before the depth self-encoder is used, the depth self-encoder is trained by collecting magnetic field data in advance to obtain the capability of extracting the depth feature.
Compared with the prior art, the invention adopting the technical scheme has the following technical effects:
the invention provides a vehicle type identification method based on magnetic field data, which analyzes the magnetic field data to obtain vehicle types in traffic flow, improves the usability and reduces the cost compared with the traditional vehicle type identification based on image identification.
Drawings
Fig. 1 is an overall architecture diagram of a vehicle type recognition method based on magnetic field data according to the present invention.
Fig. 2 is a schematic view of an embedded magnetic field sensor of the present invention.
Fig. 3 is a schematic diagram of a depth self-encoder in the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
As shown in fig. 1, the overall architecture diagram of a vehicle type recognition method based on magnetic field data proposed by the present invention is classified into two parts: the method comprises a model training stage and a vehicle type recognition stage, wherein the two stages comprise three common steps: a data acquisition step, a data preprocessing step and a feature extraction step; and simultaneously, the model training stage also comprises training of the classification model, and the vehicle type recognition stage also comprises vehicle type recognition by using the trained classification model.
A data acquisition step: as shown in fig. 2, a magnetic field sensor is buried in a road, magnetic field data signals of vehicles are collected in real time, and manual data labeling is performed to mark the types of the vehicles;
a data preprocessing step: the data preprocessing method comprises two steps: firstly, carrying out 0-1 standard normalization on the collected magnetic field data; the normalized magnetic field data is subjected to segmentation of a magnetic field data window by using a variance threshold method, which comprises the following specific steps:
1) scanning real-time magnetic field data by using a sliding window, and calculating the mean value mu and the variance sigma of the magnetic field data in the window, wherein the window size is 5000ms, and the step length is 1000 ms;
2) if the mean value μ of the magnetic field data in the window is greater than the threshold value μ 'and the variance σ is greater than the threshold value σ', the magnetic field data in the window is saved for further processing, wherein μ 'is 0.45 and σ' is 8.4.
A characteristic extraction step: the extraction of depth features is performed on the magnetic field data within the window using a depth self-encoder, the structure of which is shown in fig. 3. The depth self-encoder is composed of an encoder and a decoder, the general structure is symmetrical, the self-encoder used by the invention is composed of a fully-connected neural network, and the dimension of the depth feature is 50. Before the self-encoder is used, a part of magnetic field data is collected in advance, and the self-encoder is trained to acquire the capability of extracting depth features.
Training a classification model: in the model training phase, the support vector machine classifier is trained using the extracted depth features.
Vehicle type identification: in the vehicle type identification stage, magnetic field data of a vehicle to be identified are collected by using a magnetic field sensor, after the magnetic field data are preprocessed by using a preprocessing method, depth features are extracted by using a depth self-encoder, and finally, the extracted depth features are classified by using a trained support vector machine classifier so as to obtain a corresponding vehicle type.
The above embodiments are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and any modifications made on the basis of the technical scheme according to the technical idea of the present invention fall within the protection scope of the present invention.
Claims (5)
1. A vehicle type recognition method based on magnetic field data is characterized by comprising a model training stage and a vehicle type recognition stage, wherein,
the model training phase comprises the following steps:
step 1, collecting magnetic field data of different types of vehicles passing by using a magnetic field sensor, and manually marking the magnetic field data to mark the types of the vehicles;
step 2, preprocessing the magnetic field data collected and labeled in the step 1;
step 3, extracting the depth characteristics of the magnetic field data preprocessed in the step 2 by adopting a pre-trained depth self-encoder;
step 4, training the support vector machine classifier by using the depth features extracted in the step 3 to obtain a trained support vector machine classifier;
the vehicle type identification stage comprises the following steps:
step 5, collecting magnetic field data of the vehicle to be identified by using a magnetic field sensor, and preprocessing the magnetic field data;
and 6, performing depth feature extraction on the magnetic field data preprocessed in the step 5 by adopting a pre-trained depth self-encoder, and performing vehicle type identification by using a trained support vector machine classifier.
2. The vehicle type recognition method based on magnetic field data according to claim 1, wherein the magnetic field sensor is buried under a road surface.
3. The vehicle type identification method based on magnetic field data as claimed in claim 1, wherein the preprocessing procedures in step 2 and step 5 are as follows:
1) performing 0-1 standard normalization processing on the magnetic field data;
2) the method for segmenting the magnetic field data window by using the variance threshold method on the magnetic field data after the normalization processing specifically comprises the following steps:
sliding the magnetic field data by using a window with the size of 5000ms, wherein the window step length is 1000 ms; and calculating the mean value mu and the variance sigma of the magnetic field data in each window, and if the mean value mu of the magnetic field data in each window is greater than a threshold value mu 'and the variance sigma is greater than a threshold value sigma', storing the magnetic field data in the window, wherein the mu 'is 0.45, and the sigma' is 8.4.
4. The vehicle type recognition method based on magnetic field data as claimed in claim 1, wherein said depth self-encoder is composed of an encoder and a decoder, and the encoder and decoder are symmetrical in structure.
5. The vehicle type identification method based on the magnetic field data as claimed in claim 1, wherein the depth self-encoder is a fully-connected neural network, the dimension of the depth feature is 50, and the depth self-encoder acquires the magnetic field data in advance before use to train the depth self-encoder so as to obtain the capability of the depth self-encoder to extract the depth feature.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111010856.3A CN113780130A (en) | 2021-08-31 | 2021-08-31 | Vehicle type identification method based on magnetic field data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111010856.3A CN113780130A (en) | 2021-08-31 | 2021-08-31 | Vehicle type identification method based on magnetic field data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113780130A true CN113780130A (en) | 2021-12-10 |
Family
ID=78840182
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111010856.3A Pending CN113780130A (en) | 2021-08-31 | 2021-08-31 | Vehicle type identification method based on magnetic field data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113780130A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779281A (en) * | 2012-06-25 | 2012-11-14 | 同济大学 | Vehicle type identification method based on support vector machine and used for earth inductor |
| CN105321354A (en) * | 2015-09-22 | 2016-02-10 | 中国科学院上海微系统与信息技术研究所 | Simple vehicle detection, classification and identification method based on geomagnetic sensor |
| CN110378236A (en) * | 2019-06-20 | 2019-10-25 | 西安电子科技大学 | Testing vehicle register identification model construction, recognition methods and system based on deep learning |
-
2021
- 2021-08-31 CN CN202111010856.3A patent/CN113780130A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779281A (en) * | 2012-06-25 | 2012-11-14 | 同济大学 | Vehicle type identification method based on support vector machine and used for earth inductor |
| CN105321354A (en) * | 2015-09-22 | 2016-02-10 | 中国科学院上海微系统与信息技术研究所 | Simple vehicle detection, classification and identification method based on geomagnetic sensor |
| CN110378236A (en) * | 2019-06-20 | 2019-10-25 | 西安电子科技大学 | Testing vehicle register identification model construction, recognition methods and system based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 余胜;陈敬东;王新余;: "基于深度学习的复杂场景下车辆识别方法", 计算机与数字工程, no. 09 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Silva et al. | An extended center-symmetric local binary pattern for background modeling and subtraction in videos | |
| CN100485710C (en) | Method for recognizing vehicle type by digital picture processing technology | |
| CN105930791B (en) | The pavement marking recognition methods of multi-cam fusion based on DS evidence theory | |
| CN109101924B (en) | Machine learning-based road traffic sign identification method | |
| CN103761531B (en) | The sparse coding license plate character recognition method of Shape-based interpolation contour feature | |
| CN108090429B (en) | Vehicle type recognition method for graded front face bayonet | |
| CN106650731B (en) | Robust license plate and vehicle logo recognition method | |
| Li et al. | Robust people counting in video surveillance: Dataset and system | |
| CN105335743A (en) | Vehicle license plate recognition method | |
| Yogheedha et al. | Automatic vehicle license plate recognition system based on image processing and template matching approach | |
| CN105373794A (en) | Vehicle license plate recognition method | |
| CN111260629A (en) | Pantograph structure abnormity detection algorithm based on image processing | |
| CN103268489A (en) | Motor vehicle plate identification method based on sliding window searching | |
| CN102156983A (en) | Pattern recognition and target tracking based method for detecting abnormal pedestrian positions | |
| CN110659649A (en) | Image processing and character recognition algorithm based on near infrared light imaging | |
| CN103745224A (en) | Image-based railway contact net bird-nest abnormal condition detection method | |
| CN108509950B (en) | Railway contact net support number plate detection and identification method based on probability feature weighted fusion | |
| CN104200228A (en) | Recognizing method and system for safety belt | |
| Islam et al. | Automatic vehicle number plate recognition using structured elements | |
| CN105117707A (en) | Regional image-based facial expression recognition method | |
| CN106803087A (en) | A kind of car number automatic identification method and system | |
| CN107358236A (en) | A kind of number-plate number identifying system and method based on camera device | |
| CN107633251A (en) | A kind of vehicle identification system based on image enhaucament | |
| CN107578414B (en) | Method for processing pavement crack image | |
| CN104063682A (en) | Pedestrian detection method based on edge grading and CENTRIST characteristic |
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 |