CN114445369A - Contact net sectional insulator identification method and device based on 3D imaging technology - Google Patents
Contact net sectional insulator identification method and device based on 3D imaging technology Download PDFInfo
- Publication number
- CN114445369A CN114445369A CN202210098587.9A CN202210098587A CN114445369A CN 114445369 A CN114445369 A CN 114445369A CN 202210098587 A CN202210098587 A CN 202210098587A CN 114445369 A CN114445369 A CN 114445369A
- Authority
- CN
- China
- Prior art keywords
- model
- insulator
- training
- data
- point cloud
- 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
- 239000012212 insulator Substances 0.000 title claims abstract description 72
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000005516 engineering process Methods 0.000 title claims abstract description 20
- 238000003384 imaging method Methods 0.000 title claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 48
- 239000011159 matrix material Substances 0.000 claims abstract description 37
- 238000012360 testing method Methods 0.000 claims abstract description 28
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 19
- 238000007781 pre-processing Methods 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 12
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000005070 sampling Methods 0.000 claims description 5
- 238000004422 calculation algorithm Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 8
- 238000001514 detection method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003137 locomotive effect Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013144 data compression Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- 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
-
- 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
- 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
- G06N3/047—Probabilistic or stochastic 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20076—Probabilistic image processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Molecular Biology (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a contact net sectional insulator identification method and a device based on a 3D imaging technology, wherein the method comprises the steps of collecting point cloud data of a contact net sectional insulator; preprocessing the collected point cloud data; converting the preprocessed point cloud data into a three-dimensional matrix; inputting the three-dimensional matrix as a training set into a convolutional neural network for model training; and testing the convolutional neural network model after training is finished, outputting the model to perform segmented insulator recognition if the test result meets the requirement, and otherwise, continuing training. The invention adopts the deep convolution neural network to intelligently identify the contact network section insulator, thereby improving the identification efficiency and precision.
Description
Technical Field
The invention belongs to the technical field of contact network insulator identification, and particularly relates to a contact network sectional insulator identification method and device based on a 3D imaging technology.
Background
The sectionalizing insulator is an insulating device adopted when the contact net carries out electric sectionalization, when a certain contact net sectionalization breaks down, the isolating switch at the sectionalizing insulator is opened to cut off the power of the part of the contact net, and the other parts normally supply power, so that the reliability and the flexibility of the operation of the contact net are improved. Due to the reasons of voltage difference, installation defects, pantograph abrasion and the like, an arcing phenomenon can occur when a train passes through the section insulator, so that the section insulator and a contact network are damaged, and the pantograph of the electric locomotive is subjected to pantograph striking in serious conditions, so that the operation safety of the electric locomotive is directly influenced by fault detection and identification of the section insulator.
At present, the method for detecting and identifying the section insulator mainly comprises a multi-point detection method and an on-line monitoring method, and the position information of the section insulator is identified by detecting at a plurality of points such as a slide connecting clamp, a contact line connecting clamp and the like on the section insulator so as to judge whether the section insulator is the section insulator; the online monitoring method uses a camera to continuously shoot, takes images with fixed frame number to manually check, judges whether the insulator is a segmented insulator or not, and has low efficiency.
Disclosure of Invention
In order to solve the problem of low efficiency of the existing sectional insulator identification technology, the invention provides a contact net sectional insulator identification method based on a 3D imaging technology. The invention adopts the deep convolution neural network to intelligently identify the contact network section insulator, thereby improving the identification efficiency and precision.
The invention is realized by the following technical scheme:
a contact net sectional insulator identification method based on a 3D imaging technology comprises the following steps:
collecting point cloud data of a sectional insulator of a contact network;
preprocessing the collected point cloud data;
converting the preprocessed point cloud data into a three-dimensional matrix;
inputting the three-dimensional matrix as a training set into a convolutional neural network for model training;
and testing the convolutional neural network model after training is finished, outputting the model to perform segmented insulator recognition if the test result meets the requirement, and otherwise, continuing training.
Preferably, the pretreatment process of the present invention specifically comprises:
denoising the collected point cloud data;
and performing downsampling processing on the point cloud data subjected to denoising processing to ensure that the number of data points is consistent.
Preferably, the denoising processing of the invention specifically adopts KNN algorithm to perform clustering, and noise points are removed;
the down-sampling treatment specifically comprises the following steps: selecting a rectangular window with a preset size, dividing the rectangular window into small windows with the same size, taking one point for each window, and replacing the points of the whole small window with an average value.
Preferably, the step of converting the preprocessed point cloud data into the three-dimensional matrix specifically comprises:
dividing data into three matrixes of X multiplied by Y multiplied by 1, Y multiplied by Z multiplied by 1 and X multiplied by Z multiplied by 1 by respectively carrying out X, Y, Z axis data division;
splicing the three matrixes through dimensionality to obtain a W multiplied by H multiplied by 3 three-dimensional matrix; where W and H represent the width and height, respectively, of the three-dimensional matrix.
Preferably, the model training process of the present invention specifically includes:
carrying out normalization processing on the three-dimensional matrix, and uniformly mapping data to a [0,1] interval;
dividing the data set obtained after normalization processing into a training set and a test set;
and training the ResNet-18 convolutional neural network model by using a training set.
Preferably, the test process of the present invention specifically includes:
inputting the test set into the trained model;
obtaining a confusion matrix based on the model classification output result;
calculating to obtain the accuracy and the recall rate according to the confusion matrix;
and judging whether the model needs to be trained continuously according to the accuracy and the recall rate.
Preferably, the accuracy Accuarcy calculation formula of the invention is as follows:
the Recall rate Recall calculation formula is as follows:
wherein TP indicates the number of correctly recognized segment insulators, FP indicates the number of erroneously recognized other components as segment insulators, FN indicates the number of erroneously recognized segment insulators as other components, and TN indicates the number of correctly recognized non-segment insulators.
In a second aspect, the invention provides a contact network section insulator recognition device based on a 3D imaging technology, which comprises a data acquisition module, a data preprocessing module, a data conversion module, a model training module and a model testing module;
the data acquisition module is used for acquiring point cloud data of the contact net sectional insulator;
the data preprocessing module is used for preprocessing the collected point cloud data;
the data conversion module converts the preprocessed point cloud data into a three-dimensional matrix;
the model training module takes the three-dimensional matrix as a training set and inputs the three-dimensional matrix into a convolutional neural network for model training;
and the model testing module is used for testing the convolutional neural network model after training is finished, outputting the model to perform segmented insulator recognition if the test is qualified, and otherwise, continuing training.
In a third aspect, the invention proposes a computer device comprising a memory storing a computer program and a processor implementing the steps of the method of the invention when executing the computer program.
In a fourth aspect, the invention proposes a computer-readable storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method according to the invention.
The invention has the following advantages and beneficial effects:
1. according to the method, the three-dimensional point cloud data of the sectional insulator of the contact network are collected, preprocessed and converted, a training data set is constructed and used for training a convolutional neural network model, and the model is tested, so that an insulator recognition model is obtained and used for automatically recognizing the sectional insulator of the contact network. The invention has small calculation amount and low cost, and improves the identification efficiency and precision.
2. The invention adopts the ResNet-18 model, has high convergence speed and short time, and further improves the identification efficiency.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
fig. 1 is a schematic flow chart of an identification method according to an embodiment of the present invention.
Fig. 2 is a 3D image of a segmented insulator collected in accordance with an embodiment of the present invention.
Fig. 3 is a schematic diagram of clustering by distance in the field of the embodiment of the present invention.
Fig. 4 is a schematic diagram of down-sampling according to an embodiment of the present invention.
FIG. 5 is a schematic diagram illustrating a point cloud data preprocessing according to an embodiment of the invention.
FIG. 6 is a diagram of a feature extractor in accordance with an embodiment of the present invention.
Fig. 7 is a schematic diagram of a ResNet-18 network structure according to an embodiment of the present invention.
Fig. 8 is a block structure diagram according to an embodiment of the present invention.
Fig. 9 is a diagram showing a configuration of a computer device according to an embodiment of the present invention.
Fig. 10 is a schematic block diagram of an identification device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
Example 1
The embodiment provides a method for identifying a contact line section insulator based on a 3D imaging technology, as shown in fig. 1, the method includes:
step 1, collecting point cloud data of a section insulator of a contact network.
In this embodiment, a 3D camera may be used to collect point cloud data of the catenary section insulator, as shown in fig. 2.
And 2, preprocessing the collected point cloud data.
In this embodiment, in order to make the number of the point cloud data of the segment insulator collected by the 3D camera the same, the collected point cloud data is denoised and downsampled.
The embodiment specifically performs denoising processing on the point cloud data as follows:
in the embodiment, a clustering algorithm is adopted for data clustering, so that noise points are removed. The present embodiment implements clustering by improving on the basis of the KNN algorithm, whose idea is that if a point is k nearest neighbors in the feature space, then these points belong to a class. In the embodiment, a K value is set for comparing the distance between two points, and the point cloud data acquired by the 3D camera often contains many points which do not exist in reality, and the points can greatly influence the processing of the point cloud. Therefore, data needs to be denoised, and the distance between two points is calculated by using euclidean clustering, wherein the calculation formula is as follows:
wherein (X)1，X2)，(Y1，Y2) Respectively representing the coordinates of two points.
Clustering was performed based on the K value to remove noise points, as shown in fig. 3. The circle in the graph represents a cluster of clusters, after the clusters are clustered, each cluster is combined according to the center point of the cluster, and the combination mode is as follows:
d1＝abs(A-B)
wherein, A and B respectively represent the coordinates of the central points of the two clusters.
The process of performing down-sampling processing to obtain an equal number of points in this embodiment specifically includes:
in this embodiment, a rectangular window with a fixed size is selected according to the data characteristics of the segmented insulator and the point cloud data analysis obtained by collection, and the point cloud data in the region is down-sampled, and the specific process is to divide the rectangular window into small windows with the same size, take one point for each window, and replace the points of the whole small window with an average value, as shown in fig. 4.
And 3, converting the preprocessed point cloud data into a three-dimensional matrix.
In this embodiment, the coordinate form of the data point is (X, Y, Z), three point coordinates are obtained by taking the X, Y, Z axes as dimensions, respectively, and after down-sampling is completed, the data point in the space is converted into three point coordinates in a rectangular coordinate system, and the three-dimensional matrix shown in fig. 5 is obtained by constructing in this way. In the figure, an X, Y matrix represents values of coordinates of a point obtained after X, Y coordinate conversion using point cloud data, an X, Z matrix represents values of coordinates of a point obtained after X, Z coordinate conversion using point cloud data, a Y, Z matrix represents values of coordinates of a point obtained after Y, Z coordinate conversion using point cloud data, and finally the three matrices are spliced to obtain a three-dimensional matrix of W × H × 3(W, H represents widths and heights of the matrices, respectively).
And 4, inputting the three-dimensional matrix serving as a training set into a convolutional neural network for model training.
In this embodiment, normalization processing is performed on the obtained three-dimensional matrix with the size of wxh × 3, and data is mapped to a [0,1] interval in a unified manner, so as to accelerate the convergence speed of the later-stage model, where the normalization processing formula is:
wherein a, b represent normalized intervals [ a, b ]]The present embodiment normalizes the data to between 0 and 1; max and Min respectively represent the maximum value and the minimum value in the matrix; x1Representing the value of the point to be normalized.
The normalized data set is divided into a training set (90%) and a test set (10%).
In the embodiment, an improved ResNet-18 convolutional neural network model is adopted for training, the network structure is shown in FIG. 7, and a residual block structure is introduced into the ResNet-18 model, so that the problem of gradient disappearance of the neural network in the training process is solved; in the embodiment, a ResNet-18 model is used, a feature extractor is added before being input into a network, the feature extractor performs feature extraction with 2409 × 3209 as an input size, as shown in FIG. 6, the image is compressed and down-sampled mainly by using convolution and pooling operations, and excessive pixels lost in the data compression process are reduced; meanwhile, a Batch Normalization (BN) layer is added before the output of each layer in the ResNet-18 model in the normalization operation in the step 4, data are normalized to be within a [0,1] interval, and a RELU activation function is added at the same time, as shown in a B/R layer of fig. 7, so that the data distribution caused by the increase of the network depth is prevented from changing, and the convergence speed of the model is accelerated; finally, the fully-connected (FC) layer is replaced with a 1 × 1 convolutional layer, so that the output image size is not limited by the input image size (the fully-connected layer needs a fixed input size to reconstruct the convolved image information), and the model parameters are reduced. Wherein the block structure is shown in fig. 8.
In this embodiment, the convolutional neural network model is trained by using a training set, the number of convolutional kernels is continuously adjusted in the training process, the bottom layer, middle layer and high layer features of data are learned by raising and lowering dimensions of the feature map of each layer, and finally, the input data is restored by using a 1 × 1 convolutional layer and is classified by using a softmax function.
And 5, testing the trained convolutional neural network model, if the test result meets the requirement, obtaining a segmented insulator identification model, and otherwise, returning to the step 3.
In this embodiment, after the model training is completed, the model after the training is tested by using the test set, and the model performance is evaluated by using the accuracy and the Recall ratio, where the accuracy Accuarcy and the Recall ratio Recall calculation formulas are as follows:
wherein TP represents the number of correctly recognized segment insulators, FP represents the number of erroneously recognized other components as segment insulators, FN represents the number of erroneously recognized segment insulators as other components, and TN represents the number of correctly recognized non-segment insulators; TP, FP, FN, TN are calculated by the confusion matrix as shown in table 1:
TABLE 1 confusion matrix for classification of section insulators
And judging whether the model needs to be trained continuously according to the indexes of accuracy and recall rate (the ratio of the number of the positively-identified sample section insulators to the number of all the predicted section insulators is correctly identified).
And 6, identifying the sectional insulator by adopting a sectional insulator identification model.
The embodiment also provides a computer device for executing the method of the embodiment.
As shown in fig. 9 in particular, the computer device includes a processor, an internal memory, and a system bus; various device components including internal memory and processors are connected to the system bus. A processor is hardware used to execute computer program instructions through basic arithmetic and logical operations in a computer system. An internal memory is a physical device used to temporarily or permanently store computing programs or data (e.g., program state information). The system bus may be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus. The processor and the internal memory may be in data communication via a system bus. Including read-only memory (ROM) or flash memory (not shown), and Random Access Memory (RAM), which typically refers to main memory loaded with an operating system and computer programs.
Computer devices typically include an external storage device. The external storage device may be selected from a variety of computer readable media, which refers to any available media that can be accessed by the computer device, including both removable and non-removable media. For example, computer-readable media includes, but is not limited to, flash memory (micro SD cards), CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer device.
A computer device may be logically connected in a network environment to one or more network terminals. The network terminal may be a personal computer, a server, a router, a smart phone, a tablet, or other common network node. The computer apparatus is connected to the network terminal through a network interface (local area network LAN interface). A Local Area Network (LAN) refers to a computer network formed by interconnecting within a limited area, such as a home, a school, a computer lab, or an office building using a network medium. WiFi and twisted pair wiring ethernet are the two most commonly used technologies to build local area networks.
It should be noted that other computer systems including more or less subsystems than computer devices can also be suitable for use with the invention.
As described in detail above, the computer device adapted to the present embodiment can perform the specified operations of the overhead line section insulator identification method based on the 3D imaging technology. The computer device performs these operations in the form of software instructions executed by a processor in a computer-readable medium. These software instructions may be read into memory from a storage device or from another device via a local area network interface. The software instructions stored in the memory cause the processor to perform the method of processing group membership information described above. Furthermore, the present invention can be implemented by hardware circuits or by a combination of hardware circuits and software instructions. Thus, implementation of the present embodiments is not limited to any specific combination of hardware circuitry and software.
Example 2
The embodiment provides a contact line section insulator recognition device based on a 3D imaging technology, as shown in fig. 10, including a data acquisition module 10, a data preprocessing module 20, a data conversion module 30, a model training module 40, and a model testing module 50.
The data acquisition module 10 is used for acquiring point cloud data of the contact line section insulator.
The data preprocessing module 20 preprocesses the collected point cloud data.
The data conversion module 30 converts the preprocessed point cloud data into a three-dimensional matrix.
The model training module 40 inputs the three-dimensional matrix as a training set into the convolutional neural network for model training.
The model testing module 50 is used for testing the trained convolutional neural network model, and if the test is qualified, the model is output to perform the segmentation insulator recognition, otherwise, the training is continued.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A contact net sectional insulator identification method based on a 3D imaging technology is characterized by comprising the following steps:
collecting point cloud data of a sectional insulator of a contact network;
preprocessing the collected point cloud data;
converting the preprocessed point cloud data into a three-dimensional matrix;
inputting the three-dimensional matrix as a training set into a convolutional neural network for model training;
and testing the convolutional neural network model after training is finished, outputting the model to perform segmented insulator recognition if the test result meets the requirement, and otherwise, continuing training.
2. The method for identifying the sectionalizing insulator of the overhead line system based on the 3D imaging technology according to claim 1, wherein the preprocessing process specifically comprises:
denoising the collected point cloud data;
and performing downsampling processing on the point cloud data subjected to denoising processing to ensure that the number of data points is consistent.
3. The contact line sectionalizing insulator identification method based on the 3D imaging technology as claimed in claim 2, wherein the denoising process specifically adopts a KNN algorithm for clustering to remove noise points;
the down-sampling treatment specifically comprises the following steps: selecting a rectangular window with a preset size, dividing the rectangular window into small windows with the same size, taking one point for each window, and replacing the points of the whole small window with an average value.
4. The method for identifying the sectionalized insulator of the overhead line system based on the 3D imaging technology as claimed in claim 1, wherein the step of converting the preprocessed point cloud data into the three-dimensional matrix specifically comprises:
dividing data into three matrixes of X multiplied by Y multiplied by 1, Y multiplied by Z multiplied by 1 and X multiplied by Z multiplied by 1 by respectively carrying out X, Y, Z axis data division;
splicing the three matrixes through dimensionality to obtain a W multiplied by H multiplied by 3 three-dimensional matrix; where W and H represent the width and height, respectively, of the three-dimensional matrix.
5. The method for identifying the sectionalized insulator of the overhead line system based on the 3D imaging technology, according to claim 1, wherein the model training process specifically comprises:
carrying out normalization processing on the three-dimensional matrix, and uniformly mapping data to a [0,1] interval;
dividing the data set obtained after normalization processing into a training set and a test set;
and training the ResNet-18 convolutional neural network model by using a training set.
6. The method for identifying the sectionalizing insulator of the overhead line system based on the 3D imaging technology according to claim 5, wherein the testing process specifically comprises:
inputting the test set into the trained model;
obtaining a confusion matrix based on the model classification output result;
calculating to obtain the accuracy and the recall rate according to the confusion matrix;
and judging whether the model needs to be trained continuously according to the accuracy and the recall rate.
7. The method for identifying the sectionalizing insulator of the overhead line system based on the 3D imaging technology as claimed in claim 6, wherein the accuracy Accuarcy calculation formula is as follows:
the Recall rate Recall calculation formula is as follows:
wherein TP indicates the number of correctly recognized segment insulators, FP indicates the number of erroneously recognized other components as segment insulators, FN indicates the number of erroneously recognized segment insulators as other components, and TN indicates the number of correctly recognized non-segment insulators.
8. The contact network section insulator recognition device based on the 3D imaging technology is characterized by comprising a data acquisition module, a data preprocessing module, a data conversion module, a model training module and a model testing module;
the data acquisition module is used for acquiring point cloud data of the contact net sectional insulator;
the data preprocessing module is used for preprocessing the collected point cloud data;
the data conversion module converts the preprocessed point cloud data into a three-dimensional matrix;
the model training module takes the three-dimensional matrix as a training set and inputs the three-dimensional matrix into a convolutional neural network for model training;
and the model testing module is used for testing the convolutional neural network model after training is finished, outputting the model to perform segmented insulator recognition if the test is qualified, and otherwise, continuing training.
9. A computer arrangement comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method according to any of claims 1-7.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210098587.9A CN114445369A (en) | 2022-01-27 | 2022-01-27 | Contact net sectional insulator identification method and device based on 3D imaging technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210098587.9A CN114445369A (en) | 2022-01-27 | 2022-01-27 | Contact net sectional insulator identification method and device based on 3D imaging technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114445369A true CN114445369A (en) | 2022-05-06 |
Family
ID=81369275
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210098587.9A Pending CN114445369A (en) | 2022-01-27 | 2022-01-27 | Contact net sectional insulator identification method and device based on 3D imaging technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114445369A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114972042A (en) * | 2022-08-01 | 2022-08-30 | 成都唐源电气股份有限公司 | Pantograph point cloud splicing method, system, equipment and medium based on standard model |
| CN116051565A (en) * | 2023-04-03 | 2023-05-02 | 广州水木星尘信息科技有限公司 | Contact net defect target detection method and device based on structured light 3D point cloud |
-
2022
- 2022-01-27 CN CN202210098587.9A patent/CN114445369A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114972042A (en) * | 2022-08-01 | 2022-08-30 | 成都唐源电气股份有限公司 | Pantograph point cloud splicing method, system, equipment and medium based on standard model |
| CN116051565A (en) * | 2023-04-03 | 2023-05-02 | 广州水木星尘信息科技有限公司 | Contact net defect target detection method and device based on structured light 3D point cloud |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107564025B (en) | Electric power equipment infrared image semantic segmentation method based on deep neural network | |
| CN114445369A (en) | Contact net sectional insulator identification method and device based on 3D imaging technology | |
| CN111680614B (en) | Abnormal behavior detection method based on video monitoring | |
| CN111291817B (en) | Image recognition method, image recognition device, electronic equipment and computer readable medium | |
| CN111931592A (en) | Object recognition method, device and storage medium | |
| CN110728300A (en) | Method and system for identifying fault type based on turnout action current curve | |
| WO2023179099A1 (en) | Image detection method and apparatus, and device and readable storage medium | |
| CN114998695A (en) | Method and system for improving image recognition speed | |
| CN110852327A (en) | Image processing method, image processing device, electronic equipment and storage medium | |
| CN115861210B (en) | Transformer substation equipment abnormality detection method and system based on twin network | |
| CN114996258A (en) | Contact network fault diagnosis method based on data warehouse | |
| CN112419243B (en) | Power distribution room equipment fault identification method based on infrared image analysis | |
| WO2022222036A1 (en) | Method and apparatus for determining parking space | |
| CN114445663A (en) | Method, apparatus and computer program product for detecting challenge samples | |
| CN117171634A (en) | ECG (ECG) identity recognition method based on multi-scale feature selection fusion residual error network | |
| CN112037174A (en) | Chromosome abnormality detection method, device, equipment and computer readable storage medium | |
| CN116304824A (en) | Feature classification method, device, system and medium of electroencephalogram signals | |
| CN113591787A (en) | Method, device, equipment and storage medium for identifying optical fiber link component | |
| CN114299572A (en) | Face age estimation method and device and electronic equipment | |
| CN114550288A (en) | Event data based action identification method and device | |
| CN112435185A (en) | Fuzzy pollen picture processing method and device and electronic equipment | |
| CN111597934A (en) | System and method for processing training data for statistical applications | |
| CN117911792B (en) | Pin detecting system for voltage reference source chip production | |
| CN111798376A (en) | Image recognition method and device, electronic equipment and storage medium | |
| CN117333926B (en) | Picture aggregation method and device, electronic equipment and readable storage medium |
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 |