CN113052011A - Road target flow monitoring system based on computer vision - Google Patents
Road target flow monitoring system based on computer vision Download PDFInfo
- Publication number
- CN113052011A CN113052011A CN202110246689.6A CN202110246689A CN113052011A CN 113052011 A CN113052011 A CN 113052011A CN 202110246689 A CN202110246689 A CN 202110246689A CN 113052011 A CN113052011 A CN 113052011A
- Authority
- CN
- China
- Prior art keywords
- target
- data
- image
- detection
- computer vision
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 17
- 238000001514 detection method Methods 0.000 claims abstract description 38
- 238000007726 management method Methods 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims abstract description 6
- 238000000034 method Methods 0.000 claims description 13
- 238000011176 pooling Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000002146 bilateral effect Effects 0.000 claims description 3
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000007797 corrosion Effects 0.000 claims description 3
- 238000005260 corrosion Methods 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000003068 static effect Effects 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 230000003628 erosive effect Effects 0.000 claims description 2
- 238000003384 imaging method Methods 0.000 abstract description 3
- 238000013527 convolutional neural network Methods 0.000 abstract description 2
- 238000007781 pre-processing Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 3
- 230000006698 induction Effects 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
-
- 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
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Chemical & Material Sciences (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Analytical Chemistry (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Probability & Statistics with Applications (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a road target flow monitoring system based on computer vision, wherein a research target is image target monitoring, a yolov4 convolutional neural network model based on machine vision library opencv is used for carrying out batch preprocessing to draw a detection target, a loss function of original yolov4 is optimized, each drawn recognition object on a picture is selected by utilizing a modular matrix detection system frame, objects are classified and counted, and the coordinate of each monitored object is output; setting and comparing self-defined critical values for the deviation of the acquired data; the obtained video, data and identification results are stored and are imported into a database in real time, and big data are collected and analyzed, so that the classification of the targets is more refined, a foundation is provided for imaging management, the operation efficiency is improved while the identification precision is ensured, the detection speed is increased, and a realistic and reliable data source is provided for subsequent planning.
Description
Technical Field
The invention relates to the technical field of image target monitoring, in particular to a road target flow monitoring system based on computer vision.
Background
At present, domestic common traffic flow detection methods comprise ultrasonic detection, infrared detection, annular induction coil detection and computer vision detection, the ultrasonic detection precision is not high, the method is easily influenced by vehicle shielding and pedestrians, and the detection distance is short (generally not more than 12 m); the infrared detection is influenced by the heat source of the vehicle, the noise resistance is not strong, and the detection precision is not high; the annular induction coil is high in detection precision, but is required to be arranged in a pavement civil structure, damage is caused to a pavement, construction and installation are inconvenient, and the number of installation is large.
Disclosure of Invention
In order to overcome the problems, the invention provides a road target flow monitoring system based on computer vision, which comprises the following steps:
step one, data acquisition, wherein the data acquisition comprises static picture acquisition and dynamic video acquisition, the state and the position of a field detection target are acquired by a camera, and a sensor device is used for detecting the environment; the cloud end is connected with the intelligent terminal, researchers realize remote visual monitoring and control management on the traffic control system, the image acquisition and intelligent recognition system is presented in a C/S framework, and interaction between intelligent equipment and a background control program is realized;
step two, analyzing data, wherein in the data analysis process, a filter is in a self-defined mode, and background removal, critical value, corrosion, bilateral filtration, median blurring or form expansion processing is carried out according to the collection of a sample to obtain a more accurate result;
and step three, storing data, namely storing the acquired video, data and identification results, importing the stored data into a database in real time to acquire and analyze big data, so that the classification of the target is more detailed, connecting the database with an upper computer, and realizing data information transmission between the database and the upper computer.
Preferably, in the second step, the etching process is performed to remove boundary points of the image in the canvas, and the boundary is shrunk inwards according to the formulated parameters to remove objects which are meaningless than the detection target.
Preferably, in step two, for the median blurring processing of the image, a medianburr function is adopted as a nonlinear filter, and the median of the current field is taken as the gray value of the current point.
Preferably, the object recognition is performed through a YOLO network, and specifically includes the following steps:
reading information in a camera in real time, sending the information into a YOLO network, and equally dividing a picture into a plurality of lattices;
step two, predicting whether a target central point exists in each grid, if so, realizing target detection, wherein the working content of the target needs to determine whether the target belongs to a detection object and the position of the detection object, and the following parameters need to be obtained when predicting the target position: the position of the central point of the target frame, the length and the width of the target frame and the confidence coefficient are used for judging whether the content in the selected frame belongs to the detection object or not;
and step three, calculating confidence, and adopting a probability statement method when sampling to estimate the overall parameters, namely that the estimated values and the overall parameters are within a certain allowable error range, and the corresponding probabilities are calculation targets.
Preferably, during the capturing, a single object with a given label in the detection image is located, after which all objects with a given label in the detection image are detected.
Preferably, setting a shooting area of a camera to be white within 60 seconds, wherein the traffic flow is less than 10; the traffic flow is more than or equal to 10 and less than 20, and the area is green; the traffic flow is more than or equal to 20 and less than 30, and the area is blue; the traffic flow is more than or equal to 30 and less than 50, and the area is yellow; the traffic flow is more than or equal to 50, and the area is red.
For the normalized index function softmax, the result of the multi-classification is expressed in the form of numerical probability, and to ensure the nonnegativity of the probability, softmax translates the prediction result of the model onto the index function.
The larger the loss function is, the smaller the occupation probability of the class on the label is, and the worse the performance is, so that the softmax function is easy to cause a problem of numerical stability in the aspect of processing details. YOLOv4 replaces the softmax function with multiple independent logic operators, greatly reducing the probability of this problem occurring.
To calculate the likelihood that the input belongs to a particular tag. When calculating the classification loss, binary cross entropy loss is used for each label, so that the use of a softmax function is avoided, and the complexity of calculation is reduced.
Recall increases with increasing predictive label, but Precision fluctuates up and down.
The network structure of YOLO has 5 layers: input layer-read in data from the detected image/video; convolution layer-a complete image obtained by reassembling the local characteristics of the object through a weight matrix; pooling layers, namely selecting the maximum value of the corresponding pixel points in the convolved feature image as the pooled feature, and using the pooled feature to reduce the dimension of the feature, compress the number of data and parameters, reduce overfitting and improve the fault tolerance of the model; a fully-connected layer; and (5) outputting the layer.
Compared with the prior art, the invention has the beneficial effects that:
the method is based on a yolov4 convolutional neural network model of a machine vision library opencv, performs batch preprocessing to draw detection targets, optimizes a loss function of an original yolov4, selects each drawn recognition object on a picture by using a modular matrix detection system frame, classifies and counts the objects, and outputs the coordinates of each monitored object. Setting and comparing self-defined critical values for the deviation of the acquired data; the obtained video, data and identification results are stored and are imported into a database in real time, and big data are collected and analyzed, so that the classification of the targets is more refined, a foundation is provided for imaging management, the operation efficiency is improved while the identification precision is ensured, the detection speed is increased, and a realistic and reliable data source is provided for subsequent planning.
The neural network can obtain the control of a proposed driving program through the analysis and statistics of big data, self-defines real-time optimized parameters, and can uniformly finish the extraction of characteristics and the classification of monitoring targets in a network environment; setting and comparing self-defined critical values for the deviation of the acquired data; the obtained video, data and identification results are stored and are imported into a database in real time, and big data are collected and analyzed, so that the classification of the targets is more refined, a foundation is provided for imaging management, the operation efficiency is improved while the identification precision is ensured, the detection speed is increased, and a realistic and reliable data source is provided for subsequent planning.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a data processing flow diagram of the present invention;
FIG. 2 is a confusion matrix diagram;
fig. 3 is a diagram of a YOLO network structure.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
Referring to fig. 1 to 3, the present invention provides the following technical solutions:
example one
A road target flow monitoring system based on computer vision comprises the following steps:
step one, data acquisition, wherein the data acquisition comprises static picture acquisition and dynamic video acquisition, the state and the position of a field detection target are acquired by a camera, and a sensor device is used for detecting the environment; the cloud end is connected with the intelligent terminal, researchers realize remote visual monitoring and control management on the traffic control system, the image acquisition and intelligent recognition system is presented in a C/S framework, and interaction between intelligent equipment and a background control program is realized;
step two, analyzing data, wherein in the data analysis process, a filter is in a self-defined mode, and background removal, critical value, corrosion, bilateral filtration, median blurring or form expansion processing is carried out according to the collection of a sample to obtain a more accurate result;
and step three, storing data, namely storing the acquired video, data and identification results, importing the stored data into a database in real time to acquire and analyze big data, so that the classification of the target is more detailed, connecting the database with an upper computer, and realizing data information transmission between the database and the upper computer.
In the second step, the purpose of the erosion processing is to eliminate the boundary points of the image in the canvas, and to shrink the boundary inward according to the formulated parameters, so as to eliminate objects that are meaningless than the detection target.
In the second step, for the median fuzzy processing of the image, a medianburr function is adopted as a nonlinear filter, and the median of the current field is taken as the gray value of the current point.
The process of identifying the object by the YOLO network is as follows:
1. reading information in a camera in real time, sending the information into a YOLO network, and equally dividing a picture into 7 × 7 grids;
2. predicting whether a target central point exists in each grid, if so, realizing target detection, wherein the working content of the target needs to determine whether the target belongs to a detection object and the position of the detection object, and the following parameters need to be obtained when predicting the position of the target: the position of the central point of the target frame, the length and the width of the target frame and the confidence coefficient are used for judging whether the content in the selected frame belongs to the detection object or not;
3. confidence is calculated, and when sampling estimates the overall parameters, the conclusion is always uncertain due to the randomness of the samples. Therefore, a statement method of probability is adopted, namely, the estimated value and the overall parameter are within a certain allowable error range, and the corresponding probability is the calculation target.
During shooting, a single object with a given label in the detection image is located, and then all objects with the given label in the detection image are detected.
Setting a shooting area of a camera for 60 seconds, wherein the traffic flow is less than 10, and the area is white; the traffic flow is more than or equal to 10 and less than 20, and the area is green; the traffic flow is more than or equal to 20 and less than 30, and the area is blue; the traffic flow is more than or equal to 30 and less than 50, and the area is yellow; the traffic flow is more than or equal to 50, and the area is red.
The network structure of the YOLO comprises an input layer, a convolution layer, a pooling layer, a full-link layer and an output layer; the input layer reads data from the detected image/video; the convolution layer reassembles the local characteristics of the object through a weight matrix to obtain a complete image; and the pooling layer selects the maximum value of the corresponding pixel points in the convolved feature image as the pooled feature, and is used for feature dimension reduction, compressing the number of data and parameters, reducing overfitting and improving the fault tolerance of the model.
It will be understood by those skilled in the art that all or part of the steps in the methods of the above embodiments may be implemented by hardware related to instructions of a program, and the program may be stored in a computer readable storage medium.
The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.
Claims (10)
1. A road target flow monitoring system based on computer vision is characterized by comprising the following steps:
step one, data acquisition, wherein the data acquisition comprises static picture acquisition and dynamic video acquisition, the state and the position of a field detection target are acquired by a camera, and a sensor device is used for detecting the environment; the cloud end is connected with the intelligent terminal, researchers realize remote visual monitoring and control management on the traffic control system, the image acquisition and intelligent recognition system is presented in a C/S framework, and interaction between intelligent equipment and a background control program is realized;
step two, analyzing data, wherein in the data analysis process, a filter is in a self-defined mode, and background removal, critical value, corrosion, bilateral filtration, median blurring or form expansion processing is carried out according to the collection of a sample to obtain a more accurate result;
and step three, storing data, namely storing the acquired video, data and identification results, importing the stored data into a database in real time to acquire and analyze big data, so that the classification of the target is more detailed, connecting the database with an upper computer, and realizing data information transmission between the database and the upper computer.
2. The system of claim 1, wherein in step two, the erosion process is aimed at eliminating boundary points of the image in the canvas, and the boundary is shrunk inwards according to the parameters set forth to eliminate meaningless objects compared with the detected object.
3. The system for monitoring the target flow rate of the road based on the computer vision as claimed in claim 1, wherein in the second step, the median fuzzy processing of the image adopts medianburr function as a nonlinear filter, and the median of the current domain is taken as the gray value of the current point.
4. The system for monitoring the road target flow based on the computer vision as claimed in claim 1, wherein the object recognition is performed through a YOLO network, which includes the following steps:
reading information in a camera in real time, sending the information into a YOLO network, and equally dividing a picture into a plurality of lattices;
step two, predicting whether a target central point exists in each grid, if so, realizing target detection, wherein the working content of the target needs to determine whether the target belongs to a detection object and the position of the detection object, and the following parameters need to be obtained when predicting the target position: the position of the central point of the target frame, the length and the width of the target frame and the confidence coefficient are used for judging whether the content in the selected frame belongs to the detection object or not;
and step three, calculating confidence, and adopting a probability statement method when sampling to estimate the overall parameters, namely that the estimated values and the overall parameters are within a certain allowable error range, and the corresponding probabilities are calculation targets.
5. The system of claim 4, wherein during the capturing process, a single object with a given tag in the detected image is located, and then all objects with a given tag in the detected image are detected.
6. The system for monitoring the road target flow based on the computer vision as claimed in claim 4, wherein a camera is planned to shoot an area within 60 seconds, the traffic flow is less than 10, and the area is white; the traffic flow is more than or equal to 10 and less than 20, and the area is green; the traffic flow is more than or equal to 20 and less than 30, and the area is blue; the traffic flow is more than or equal to 30 and less than 50, and the area is yellow; the traffic flow is more than or equal to 50, and the area is red.
7. The computer vision based road target flow monitoring system of claim 4, wherein the network structure of YOLO comprises an input layer, a convolutional layer, a pooling layer, a fully-connected layer, and an output layer.
8. The system of claim 7, wherein the input layer reads data from the detected image/video.
9. The system of claim 7, wherein the convolutional layer is a complete image obtained by reassembling local features of the object through a weight matrix.
10. The system of claim 7, wherein the pooling layer selects the maximum value of the corresponding pixel points in the convolved feature image as the pooled feature for feature dimension reduction, compresses the number of data and parameters, reduces overfitting, and improves the fault tolerance of the model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110246689.6A CN113052011A (en) | 2021-03-05 | 2021-03-05 | Road target flow monitoring system based on computer vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110246689.6A CN113052011A (en) | 2021-03-05 | 2021-03-05 | Road target flow monitoring system based on computer vision |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113052011A true CN113052011A (en) | 2021-06-29 |
Family
ID=76510284
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110246689.6A Pending CN113052011A (en) | 2021-03-05 | 2021-03-05 | Road target flow monitoring system based on computer vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113052011A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107358803A (en) * | 2017-08-22 | 2017-11-17 | 哈尔滨理工大学 | A kind of traffic signal control system and its control method |
CN108205891A (en) * | 2018-01-02 | 2018-06-26 | 霍*** | A kind of vehicle monitoring method of monitoring area |
CN108537117A (en) * | 2018-03-06 | 2018-09-14 | 哈尔滨思派科技有限公司 | A kind of occupant detection method and system based on deep learning |
CN111554105A (en) * | 2020-05-29 | 2020-08-18 | 浙江科技学院 | Intelligent traffic identification and statistics method for complex traffic intersection |
CN112101433A (en) * | 2020-09-04 | 2020-12-18 | 东南大学 | Automatic lane-dividing vehicle counting method based on YOLO V4 and DeepsORT |
-
2021
- 2021-03-05 CN CN202110246689.6A patent/CN113052011A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107358803A (en) * | 2017-08-22 | 2017-11-17 | 哈尔滨理工大学 | A kind of traffic signal control system and its control method |
CN108205891A (en) * | 2018-01-02 | 2018-06-26 | 霍*** | A kind of vehicle monitoring method of monitoring area |
CN108537117A (en) * | 2018-03-06 | 2018-09-14 | 哈尔滨思派科技有限公司 | A kind of occupant detection method and system based on deep learning |
CN111554105A (en) * | 2020-05-29 | 2020-08-18 | 浙江科技学院 | Intelligent traffic identification and statistics method for complex traffic intersection |
CN112101433A (en) * | 2020-09-04 | 2020-12-18 | 东南大学 | Automatic lane-dividing vehicle counting method based on YOLO V4 and DeepsORT |
Non-Patent Citations (3)
Title |
---|
ZICONG JIANG等: "Real-time object detection method for embedded devices", 《HTTPS://ARXIV.ORG/ABS/2011.04244V2》 * |
徐子睿等: "基于YOLOv4的车辆检测与流量统计研究", 《现代信息科技》 * |
黄健等: "深度卷积神经网络的目标检测算法综述", 《计算机工程与应用》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108961235B (en) | Defective insulator identification method based on YOLOv3 network and particle filter algorithm | |
CN109345547B (en) | Traffic lane line detection method and device based on deep learning multitask network | |
CN112380952A (en) | Power equipment infrared image real-time detection and identification method based on artificial intelligence | |
CN112016461B (en) | Multi-target behavior recognition method and system | |
CN111832398B (en) | Unmanned aerial vehicle image distribution line pole tower ground wire broken strand image detection method | |
CN113436184B (en) | Power equipment image defect discriminating method and system based on improved twin network | |
CN111462155B (en) | Motion detection method, device, computer equipment and storage medium | |
CN111223129A (en) | Detection method, detection device, monitoring equipment and computer readable storage medium | |
CN110599458A (en) | Underground pipe network detection and evaluation cloud system based on convolutional neural network | |
CN110826577A (en) | High-voltage isolating switch state tracking identification method based on target tracking | |
CN114155474A (en) | Damage identification technology based on video semantic segmentation algorithm | |
CN114627406A (en) | Method, system, equipment and medium for identifying rapid crowd gathering behaviors | |
CN115311623A (en) | Equipment oil leakage detection method and system based on infrared thermal imaging | |
CN118038021A (en) | Transformer substation operation site foreign matter intrusion detection method based on improvement yolov4 | |
CN117475327A (en) | Multi-target detection positioning method and system based on remote sensing image in city | |
CN116205905A (en) | Power distribution network construction safety and quality image detection method and system based on mobile terminal | |
CN116908184A (en) | Ground wire crimping detection system and detection method thereof | |
CN110599460A (en) | Underground pipe network detection and evaluation cloud system based on hybrid convolutional neural network | |
CN113052011A (en) | Road target flow monitoring system based on computer vision | |
CN116168213A (en) | People flow data identification method and training method of people flow data identification model | |
CN116051539A (en) | Diagnosis method for heating fault of power transformation equipment | |
CN114648736A (en) | Robust engineering vehicle identification method and system based on target detection | |
CN114283323A (en) | Marine target recognition system based on image deep learning | |
CN114549628A (en) | Power pole inclination detection method, device, equipment and storage medium | |
CN115100546A (en) | Mobile-based small target defect identification method and system for power 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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210629 |
|
RJ01 | Rejection of invention patent application after publication |