CN110705397A - YOLOv3 pruning identification insulator defect method suitable for small field sample amount - Google Patents
YOLOv3 pruning identification insulator defect method suitable for small field sample amount Download PDFInfo
- Publication number
- CN110705397A CN110705397A CN201910884173.7A CN201910884173A CN110705397A CN 110705397 A CN110705397 A CN 110705397A CN 201910884173 A CN201910884173 A CN 201910884173A CN 110705397 A CN110705397 A CN 110705397A
- Authority
- CN
- China
- Prior art keywords
- insulator
- training
- pruning
- yolov3
- picture
- 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 59
- 238000013138 pruning Methods 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 title claims abstract description 22
- 230000007547 defect Effects 0.000 title claims abstract description 16
- 238000012549 training Methods 0.000 claims abstract description 41
- 238000007689 inspection Methods 0.000 claims abstract description 14
- 238000013136 deep learning model Methods 0.000 claims abstract description 6
- 230000004048 modification Effects 0.000 claims abstract description 6
- 238000012986 modification Methods 0.000 claims abstract description 6
- 238000004364 calculation method Methods 0.000 claims abstract description 5
- 238000012545 processing Methods 0.000 claims abstract description 5
- 238000013135 deep learning Methods 0.000 claims description 11
- 238000001514 detection method Methods 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 2
- 230000005540 biological transmission Effects 0.000 description 4
- 230000009467 reduction Effects 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 244000141353 Prunus domestica Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010292 electrical insulation Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000008439 repair process 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
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C1/00—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people
- G07C1/20—Checking timed patrols, e.g. of watchman
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Human Resources & Organizations (AREA)
- Water Supply & Treatment (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Public Health (AREA)
- General Business, Economics & Management (AREA)
- Software Systems (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Multimedia (AREA)
- Insulators (AREA)
Abstract
The invention provides a YOLOv3 pruning identification method for insulator defects with small field sample amount, which comprises the following steps: step 1, transmitting picture data in real time according to unmanned aerial vehicle routing inspection, and acquiring insulator images and corresponding XML tag data; step 2, processing the insulator image; step 3, establishing a deep learning model after pruning modification operation is carried out by taking YOLOv3 as a basic framework; step 4, introducing an SPPnet module consisting of 4 parallel maxpool layers into the pruned YOLOv 3; step 5, converting the processed insulator image and the modified TXT label data into training data; step 6, obtaining an insulator identification model; step 7, obtaining insulator picture frame information detected in each picture; and 8, judging the insulator fault. The advantages are that: under the condition of not reducing the average precision, the training time is reduced, the requirements on the picture quality are reduced, meanwhile, the calculation force requirements on a vehicle-mounted server are reduced, and the intelligent level of insulator routing inspection is improved.
Description
Technical Field
The invention relates to a power grid inspection maintenance technology, in particular to a YOLOv3 pruning identification insulator defect method suitable for small field sample amount.
Background
Along with rapid development of economy and progressive construction of the smart power grid, higher requirements are put forward on intellectualization and rapidness of power transmission line routing inspection. As is known, an insulator is a special insulating control element, which can play an important role in an overhead transmission line, and its main function is to achieve electrical insulation and mechanical fixation, for which various electrical and mechanical performance requirements are specified. However, the manual maintenance and repair process is very complicated and has a high risk coefficient, and according to survey statistics, the number of safety accidents caused by insulator defects is large. Therefore, the regular inspection of the pole and tower insulators is a very necessary work for power inspection.
In recent years, the power inspection technology is rapidly developed, an unmanned aerial vehicle is used for navigating and inspecting a power transmission line, and then pictures shot by the unmanned aerial vehicle are manually observed and screened to inspect defects. Although the method solves the danger of manual maintenance, the screening work at the later stage is still complicated and numerous, and the manual inspection of the picture causes many misjudgment results due to the problems of picture quality, definition and the like. Later, an artificial intelligence algorithm based on a deep learning model is gradually used for carrying out image identification and screening, but different inspection field characteristics are different, the same set of learning results cannot be always used, the field sample amount is small, and fewer effective samples are inevitable.
Disclosure of Invention
In order to solve the problems, the invention discloses a YOLOv3 pruning recognition insulator defect method suitable for small field sample quantity, which reduces the training speed, reduces the requirements on the picture quality, reduces the calculation force requirements on a vehicle-mounted server and improves the intelligent level of insulator routing inspection under the condition of not reducing the average precision.
In order to achieve the above purpose, the invention provides the following technical scheme:
a method for identifying insulator defects by YOLOv3 pruning, which is suitable for field sample quantity reduction, comprises the following specific steps:
step 2, performing primary processing on the insulator image to obtain a processed insulator image, and modifying pixel information corresponding to the position of a marking frame in the TXT tag data to obtain modified TXT tag data;
step 301, using Darknet-53 as a new basic backbone frame network to perform neural network feature extraction, adding more continuous convolution layers of 3 × 3 and 1 × 1 to Darknet-53, and forming them into a basic computing module, which is much more powerful than Darknet-19 in yollov 2;
step 302, performing sparse training, and distributing corresponding scale factors to each channel, wherein absolute values of the scale factors represent the importance of the channels, so that channel pruning is facilitated;
step 303, channel pruning is carried out, a maxpool layer irrelevant to a channel number is directly abandoned, a global threshold value is introduced to control the pruning rate, and meanwhile, a local safety threshold value is set to ensure the integrity of the model;
step 304, the output layer only keeps 8 × 8 output scales to reduce the size of the model and the training calculation amount;
step 4, introducing SPPnet modules consisting of 4 parallel maxpool layers (1 × 1,5 × 5,9 × 9 and 13 × 13) into the pruned YOLOv 3;
step 6, training a deep learning network model by using training data to obtain an insulator recognition model;
step 7, inputting the real-time pictures to be detected into the trained insulator detection model to obtain the insulator frame information detected in each picture;
and 8, judging the insulator fault according to the detected insulator picture frame information.
Further, the labeling tool LabelImg used in the step 1 is a visual image calibration tool, the data set required by the YOLOv3 algorithm needs to calibrate the target in the image, generate an XML file conforming to the format of PASCAL VOC, and convert the XML file into a TXT file.
Further, in the step 2, the insulator images obtained in the step 1 are uniformly scaled into images with the pixel width and height of 416 or other multiples of 32.
Further, the training data generated in step 5 is used in step 3 to train the deep learning network model in step 3; the method comprises the steps of detecting the change conditions of parameters such as an AP value and a loss value of the deep learning network every other training period.
Further, in the step 5, the TXT tag data obtained after modification in the step 2 is imported into an LISTSs file, so that the training picture is introduced into the model.
Further, the model weight after training in step 6 is saved in a WEIGHTS file form, and the training weight of this time can be directly called in the next training.
Compared with the prior art, the invention has the beneficial effects that:
1. the insulator and insulator fault defect is identified by using a YOLOv3 algorithm based on a Darknet frame, and pruning processing such as channel reduction, frame slimming, addition of an SPPnet module and the like is performed on the YOLOv3 algorithm on the basis, so that the lightweight performance of the YOLOv3 is further optimized, and the method is suitable for the condition that the number of effective samples is insufficient due to the reasons of less number of pictures shot by an unmanned aerial vehicle, fuzziness, angle deviation and the like in a power transmission line inspection site;
2. the invention can realize the training of shooting pictures on site as fast as possible, and simultaneously avoid the problem of accuracy reduction caused by too few samples of YOLOv 3;
3. the method only slightly prunes and modifies the YOLOv3 frame (50% of the number of channels of the YOLOv3 frame is cut, and an SPPnet module is added after a specific convolutional layer is added), but the average precision (mAP) is hardly reduced, and the training speed is greatly improved;
4. the invention reduces the computing power requirement on the vehicle-mounted server and lowers the hardware requirement for realizing the YOLOv3 algorithm.
Drawings
FIG. 1 is a flow chart of the detection method of the present invention.
FIG. 2 is a diagram of the post-pruning Yolov3 model of the present invention.
FIG. 3 is a block diagram of the SPPnet module according to the present invention.
Detailed Description
The present invention will be further illustrated with reference to the accompanying drawings and specific embodiments, which are to be understood as merely illustrative of the invention and not as limiting the scope of the invention.
As shown in the figure, a method for identifying insulator defects by YOLOv3 pruning, which is suitable for field sample quantity reduction, comprises the following specific steps:
step 1: the method comprises the steps of transmitting picture data in real time according to unmanned aerial vehicle routing inspection, obtaining insulator images and corresponding XML label data, changing the insulator images and the corresponding XML label data into a TXT format according to the file characteristics of Darknet-53, wherein a used marking tool LabelImg is a visual image calibration tool, a data set required by a YOLOv3 algorithm needs to calibrate a target in an image through the tool, and an XML file conforming to the format of PASCAL VOC is generated;
step 2: performing primary processing on the insulator image, scaling the insulator image obtained in the step 1 into an image with the image pixel width and height being 416 or other multiples of 32 in a unified equal ratio manner to obtain a processed insulator image, and modifying pixel information corresponding to the position of a marking frame in the TXT label data to obtain modified TXT label data;
and step 3: with YOLOv3 as a basic framework, pruning modification operation is carried out to establish a deep learning model, and the deep learning model building method comprises the following steps:
step 301: the Darknet-53 is used as a new basic backbone frame network for neural network feature extraction, more continuous convolution layers of 3 x 3 and 1 x 1 are added into the Darknet-53, and the convolution layers form a basic computing module, so that the function is much stronger than that of Darknet-19 in Yolov 2;
step 302: sparse training is carried out, corresponding scale factors are distributed to each channel, and the absolute value of each scale factor represents the importance of the channel, so that channel pruning is facilitated;
step 303: channel pruning is carried out, a maxpool layer irrelevant to a channel number is directly abandoned, a global threshold value is introduced to control the pruning rate, and meanwhile, a local safety threshold value is set to ensure the integrity of the model;
step 304: the output layer only keeps 8 × 8 output scales to reduce the size of the model and the training calculation amount;
and 4, step 4: introducing SPPnet modules consisting of 4 parallel maxpool layers (1 × 1,5 × 5,9 × 9 and 13 × 13) into YOLOv 3;
and 5: converting the processed insulator image and the TXT label data modified in the step 3 into training data which can be used for deep learning network model training, and importing the TXT label data modified in the step 3 into an LISTS file, so that a training picture is introduced into a model;
step 6: training a deep learning network model by using training data to obtain an insulator identification model, wherein parameter change conditions such as an AP value, a loss value and the like of the deep learning network are detected every other training period, model WEIGHTS after training are saved in a WEIGHTS file form, and the training WEIGHTS can be directly called in the next training;
and 7: inputting the real-time pictures to be detected into the trained insulator detection model to obtain the insulator picture frame information detected in each picture;
and 8: and judging the insulator fault according to the detected insulator picture frame information.
The technical means disclosed in the invention scheme are not limited to the technical means disclosed in the above embodiments, but also include the technical scheme formed by any combination of the above technical features. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and such improvements and modifications are also considered to be within the scope of the present invention.
Claims (6)
1. A YOLOv3 pruning identification insulator defect method suitable for small field sample amount is characterized by comprising the following specific steps:
step 1, transmitting picture data in real time according to unmanned aerial vehicle routing inspection, acquiring an insulator picture and corresponding XML tag data, and changing the insulator picture and the corresponding XML tag data into a TXT format according to the file characteristics of Darknet-53;
step 2, performing primary processing on the insulator image to obtain a processed insulator image, and modifying pixel information corresponding to the position of a marking frame in the TXT tag data to obtain modified TXT tag data;
step 3, with YOLOv3 as a basic framework, pruning and modifying operations are carried out to establish a deep learning model, and the method comprises the following steps:
step 301, using Darknet-53 as a basic backbone frame network to extract neural network features;
step 302, performing sparse training, and distributing corresponding scale factors to each channel, wherein absolute values of the scale factors represent the importance of the channels, so that channel pruning is facilitated;
step 303, channel pruning is carried out, a maxpool layer irrelevant to a channel number is directly abandoned, a global threshold value is introduced to control the pruning rate, and meanwhile, a local safety threshold value is set to ensure the integrity of the model;
step 304, the output layer only keeps 8 × 8 output scales to reduce the size of the model and the training calculation amount;
step 4, introducing SPPnet modules consisting of 4 parallel maxpool layers (1 × 1,5 × 5,9 × 9 and 13 × 13) into the pruned YOLOv 3;
step 5, converting the processed insulator image and the TXT label data modified in the step 3 into training data for deep learning network model training;
step 6, training a deep learning network model by using training data to obtain an insulator recognition model;
step 7, inputting the real-time pictures to be detected into the trained insulator detection model to obtain the insulator frame information detected in each picture;
and 8, judging the insulator fault according to the detected insulator picture frame information.
2. The method for pruning and identifying the insulator defects by using YOLOv3 with small field sample size according to claim 1, wherein the step 1 specifically comprises the following steps: the labeling tool LabelImg is a visual image calibration tool, and the data set required by the YOLOv3 algorithm calibrates the target in the image, generates an XML file conforming to the format of PASCAL VOC, and converts the XML file into a TXT file.
3. The method for pruning and identifying the insulator defects by using YOLOv3 with small field sample size according to claim 1, wherein the step 2 specifically comprises the following steps: and (3) scaling the insulator images obtained in the step (1) into images with the pixel width and the height of 416 or 32 times in a unified equal ratio mode.
4. The method for pruning and identifying the insulator defects by using YOLOv3 with small field sample size according to claim 1, wherein the step 3 specifically comprises the following steps: training the deep learning network model in the step 3 by using the training data generated in the step 5; the method comprises the steps of detecting the change conditions of parameters such as an AP value and a loss value of the deep learning network every other training period.
5. The method for pruning and identifying the insulator defects by using YOLOv3 with small field sample size according to claim 1, wherein the step 5 specifically comprises the following steps: and (3) importing the TXT label data obtained after modification in the step (2) into an LISTS file, so that the training picture is introduced into the model.
6. The method for pruning and identifying the insulator defects by using YOLOv3 with small field sample size according to claim 1, wherein the step 6 specifically comprises the following steps: the model weight after training is saved in a WEIGHTS file form, and the training weight can be directly called in the next training.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910884173.7A CN110705397A (en) | 2019-09-19 | 2019-09-19 | YOLOv3 pruning identification insulator defect method suitable for small field sample amount |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910884173.7A CN110705397A (en) | 2019-09-19 | 2019-09-19 | YOLOv3 pruning identification insulator defect method suitable for small field sample amount |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110705397A true CN110705397A (en) | 2020-01-17 |
Family
ID=69195507
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910884173.7A Pending CN110705397A (en) | 2019-09-19 | 2019-09-19 | YOLOv3 pruning identification insulator defect method suitable for small field sample amount |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110705397A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111239550A (en) * | 2020-02-27 | 2020-06-05 | 东南大学 | Unmanned aerial vehicle full-automatic multi-machine cooperative power transmission line fault identification method for inspection |
CN111784692A (en) * | 2020-08-11 | 2020-10-16 | 国网内蒙古东部电力有限公司 | Method and device for detecting insulator defects in power system and electronic equipment |
CN112229845A (en) * | 2020-10-12 | 2021-01-15 | 国网河南省电力公司濮阳供电公司 | Unmanned aerial vehicle high-precision winding tower intelligent inspection method based on visual navigation technology |
CN112465114A (en) * | 2020-11-25 | 2021-03-09 | 重庆大学 | Rapid target detection method and system based on optimized channel pruning |
CN113592814A (en) * | 2021-07-30 | 2021-11-02 | 深圳大学 | Laser welding surface defect detection method for safety explosion-proof valve of new energy power battery |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110033453A (en) * | 2019-04-18 | 2019-07-19 | 国网山西省电力公司电力科学研究院 | Based on the power transmission and transformation line insulator Aerial Images fault detection method for improving YOLOv3 |
-
2019
- 2019-09-19 CN CN201910884173.7A patent/CN110705397A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110033453A (en) * | 2019-04-18 | 2019-07-19 | 国网山西省电力公司电力科学研究院 | Based on the power transmission and transformation line insulator Aerial Images fault detection method for improving YOLOv3 |
Non-Patent Citations (1)
Title |
---|
PENGYI ZHANG等: "SlimYOLOv3: Narrower, Faster and Better for Real-Time UAV Applications", 《ARXIV》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111239550A (en) * | 2020-02-27 | 2020-06-05 | 东南大学 | Unmanned aerial vehicle full-automatic multi-machine cooperative power transmission line fault identification method for inspection |
CN111784692A (en) * | 2020-08-11 | 2020-10-16 | 国网内蒙古东部电力有限公司 | Method and device for detecting insulator defects in power system and electronic equipment |
CN112229845A (en) * | 2020-10-12 | 2021-01-15 | 国网河南省电力公司濮阳供电公司 | Unmanned aerial vehicle high-precision winding tower intelligent inspection method based on visual navigation technology |
CN112465114A (en) * | 2020-11-25 | 2021-03-09 | 重庆大学 | Rapid target detection method and system based on optimized channel pruning |
CN113592814A (en) * | 2021-07-30 | 2021-11-02 | 深圳大学 | Laser welding surface defect detection method for safety explosion-proof valve of new energy power battery |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110705397A (en) | YOLOv3 pruning identification insulator defect method suitable for small field sample amount | |
CN110059694B (en) | Intelligent identification method for character data in complex scene of power industry | |
CN109376606B (en) | Power inspection image tower foundation fault detection method | |
CN112183788A (en) | Domain adaptive equipment operation detection system and method | |
CN113436169A (en) | Industrial equipment surface crack detection method and system based on semi-supervised semantic segmentation | |
CN110910440B (en) | Power transmission line length determination method and system based on power image data | |
CN111431986A (en) | Industrial intelligent quality inspection system based on 5G and AI cloud edge cooperation | |
CN113486865A (en) | Power transmission line suspended foreign object target detection method based on deep learning | |
CN114973002A (en) | Improved YOLOv 5-based ear detection method | |
CN111598942A (en) | Method and system for automatically positioning electric power facility instrument | |
CN111639530B (en) | Method and system for detecting and identifying power transmission tower and insulator of power transmission line | |
CN112541389A (en) | Power transmission line fault detection method based on EfficientDet network | |
CN114240868A (en) | Unmanned aerial vehicle-based inspection analysis system and method | |
CN112070715A (en) | Transmission line small-size hardware defect detection method based on improved SSD model | |
CN113255590A (en) | Defect detection model training method, defect detection method, device and system | |
CN112101138A (en) | Bridge inhaul cable surface defect real-time identification system and method based on deep learning | |
CN111008641A (en) | Power transmission line tower external force damage detection method based on convolutional neural network | |
CN112906654A (en) | Anti-vibration hammer detection method based on deep learning algorithm | |
CN112614130A (en) | Unmanned aerial vehicle power transmission line insulator fault detection method based on 5G transmission and YOLOv3 | |
CN115619719A (en) | Pine wood nematode infected wood detection method based on improved Yolo v3 network model | |
CN115546664A (en) | Cascaded network-based insulator self-explosion detection method and system | |
CN114119454A (en) | Device and method for smoke detection of power transmission line | |
CN112132088B (en) | Inspection point missing inspection identification method | |
CN111915565B (en) | Method for analyzing cracks of porcelain insulator of power transmission and transformation line in real time based on YOLACT algorithm | |
CN112683916A (en) | Method and device for identifying missing or mounting error of small hardware fittings of current collecting line tower |
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: 20200117 |
|
RJ01 | Rejection of invention patent application after publication |