CN107507172A - Merge the extra high voltage line insulator chain deep learning recognition methods of infrared visible ray - Google Patents
Merge the extra high voltage line insulator chain deep learning recognition methods of infrared visible ray Download PDFInfo
- Publication number
- CN107507172A CN107507172A CN201710671781.0A CN201710671781A CN107507172A CN 107507172 A CN107507172 A CN 107507172A CN 201710671781 A CN201710671781 A CN 201710671781A CN 107507172 A CN107507172 A CN 107507172A
- Authority
- CN
- China
- Prior art keywords
- component
- images
- insulator chain
- infrared
- visible
- 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 35
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000013135 deep learning Methods 0.000 title claims abstract description 14
- 230000004927 fusion Effects 0.000 claims abstract description 19
- 239000000284 extract Substances 0.000 claims abstract description 6
- 238000013527 convolutional neural network Methods 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 4
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 3
- 238000007500 overflow downdraw method Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 abstract description 6
- 238000001514 detection method Methods 0.000 abstract description 6
- 238000007689 inspection Methods 0.000 abstract description 6
- 238000012549 training Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000009413 insulation Methods 0.000 description 2
- 230000007935 neutral effect Effects 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000003475 lamination Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000035807 sensation Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
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
- G06T7/001—Industrial image inspection using an image reference approach
-
- G06T3/147—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration by the use of more than one image, e.g. averaging, subtraction
-
- 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
- G06T2207/10048—Infrared image
-
- 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
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Abstract
The present invention relates to a kind of extra high voltage line insulator chain deep learning recognition methods for merging infrared visible ray, including：YUV codings are carried out to the visible images of object to be identified, and obtain the Y-component of visible images, U components and V component；It will be seen that after obtaining fusion after Y-component and infrared image the progress fusion treatment of light image, its U component and V component with reference to visible images is obtained into fused images；Fused images are detected using the Faster Rcnn models trained, extract the insulator chain in object to be identified.Compared with prior art, the present invention uses Faster R CNN algorithm flows in the infrared visible ray fused images part detection of unmanned plane electric power line inspection, and every nearly 80ms recognition speed and 92.7% accuracy rate can be reached by carrying out the electric power widget identification positioning of plurality of classes.
Description
Technical field
The present invention relates to a kind of sub- extracting method of insulator string, more particularly, to a kind of extra-high line ball for merging infrared visible ray
Road insulator chain deep learning recognition methods.
Background technology
While UHV transmission line is effectively configured with the energy between regional power grid, new O＆M problem is also brought：It is special
Ultra-high-tension power transmission line is mostly high tower erection, and ground patrol officer is difficult to find the pin level defect such as bolt falling on high shaft tower；
Carrying out approaching shaft tower inspection using unmanned plane, unmanned plane shoots substantial amounts of video and picture data, if completely using artificial
Unmanned plane gathered data is carried out to check that identification defect has the low problem of efficiency.Convolutional neural networks algorithm is being used to nobody
Machine gathered data carries out running into unmanned plane gathered data background complexity while automatic identification extra high voltage line insulator chain again
Problem is this this patent by handling the infrared visible ray fused images of unmanned plane, for extra-high line ball under electriferous state
Road insulator chain feature is different from other in background and disturbed, and effectively extracts extra-high voltage under electriferous state from fused images
Pictorial information under the visible ray of line insulator string.
With application of the unmanned plane (UAV) in power-line patrolling operation, to the information excavating of unmanned plane inspection image
Or target identification demand is also more and more stronger.Traditional power components identification process often uses classical machine learning algorithm, such as
SVMs (SVM), random forest or adaboost, come with reference to the shallow-layer feature such as gradient, color or texture to power components
It is identified, it is difficult to make full use of the information of unmanned plane inspection image, and be difficult to reach higher accuracy rate.Convolutional Neural net
Network (CNN) shows excellent in target identification, turns into preference algorithm among many target identification scenes.Convolution based on region
Neutral net (RCNN) is extracted by using CNN from image to contain mesh target area to detect and identify target, still
Calculate complicated, it is difficult to meet the needs of identifying magnanimity electric inspection process picture.Fast R-CNN and Faster R-CNN utilize CNN nets
Network extracts characteristics of image, is followed by region and proposes layer, and optimizing extraction the mode containing target area and may improve identification
The grader of target, make the detection of target and identification almost real-time.
The content of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind fusion is infrared visible
The extra high voltage line insulator chain deep learning recognition methods of light.
The purpose of the present invention can be achieved through the following technical solutions：
A kind of extra high voltage line insulator chain deep learning recognition methods for merging infrared visible ray, including：
YUV codings are carried out to the visible images of object to be identified, and obtain the Y-component of visible images, U components and V
Component；
It will be seen that after obtaining fusion after Y-component and infrared image the progress fusion treatment of light image, it is combined into visible ray
The U components and V component of image obtain fused images；
Fused images are detected using the Faster-Rcnn models trained, extracted exhausted in object to be identified
Edge substring.
The visible images to object to be identified carry out YUV codings and specifically included：
Registration is carried out to visible images using affine transformation；
Images after registration is transformed into yuv space.
The Y-component of visible images and the fusion treatment process of infrared image use Wavelet Fusion method.
The Y-component and infrared image of visible images carry out fusion treatment：
Y'=w1 × Y+w2 × IR
Wherein：Y is the Y-component of visible images, and IR is infrared image, and w1 is the weight of visible images, and w2 is infrared
The weight of image, Y' are the Y-component after fusion.
The weight of the visible images and the weight of infrared image are 0.5.
The Faster-Rcnn models include RPN convolutional neural networks and Fast-rcnn convolutional neural networks.
It is described that fused images are detected using the Faster-Rcnn models trained, extract in object to be identified
Insulator chain, including：
The candidate region of insulator chain is identified from fused images using RPN convolutional neural networks；
Using Fast-rcnn convolutional neural networks insulator chain is identified from the candidate region of insulator chain.
Compared with prior art, the present invention has advantages below：
1) calculated in the infrared visible ray fused images part detection of unmanned plane electric power line inspection using Faster R-CNN
Method flow, every nearly 80ms recognition speed and 92.7% can be reached by carrying out the electric power widget identification positioning of plurality of classes
Accuracy rate.
2) amount of image information is maximum when the weight of visible images and the weight of infrared image are 0.5, and effect is preferable.
3) by compared with the fused images obtained by YUV color space transformations, wavelet transform fusion, as a result table
Bright, the image definition that the image interfusion method that the present invention uses obtains has larger improvement, so as to build insulator Sample Storehouse.
Brief description of the drawings
Fig. 1 is the key step schematic flow sheet of the inventive method；
Fig. 2 is the structural representation of RPN convolutional neural networks；
Fig. 3 is the block mold structural representation of Faster-Rcnn models.
Embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, give detailed embodiment and specific operating process, but protection scope of the present invention is not limited to
Following embodiments.
A kind of extra high voltage line insulator chain deep learning recognition methods for merging infrared visible ray, as shown in figure 1, bag
Include：
YUV codings are carried out to the visible images of object to be identified, and obtain the Y-component of visible images, U components and V
Component,
In YUV color spaces, each color has a luminance signal component Y, and two chroma signal components U and V.
Brightness is the sensation of intensity, and it and colourity are separation, if only y component of signals are without U, V component, then are so represented
Image be exactly black and white gray level image.Intensity, which changes can, does not influence color change.These three component combination cans are produced
A raw full-color image.
The visible images progress YUV codings of object to be identified are specifically included：
Registration is carried out to visible images using affine transformation；
Images after registration is transformed into yuv space.
It will be seen that after obtaining fusion after Y-component and infrared image the progress fusion treatment of light image, it is combined into visible ray
The U components and V component of image obtain fused images；
The Y-component of visible images and the fusion treatment process of infrared image use Wavelet Fusion method.
The Y-component and infrared image of visible images carry out fusion treatment：
Y'=w1 × Y+w2 × IR
Wherein：Y is the Y-component of visible images, and IR is infrared image, and w1 is the weight of visible images, and w2 is infrared
The weight of image, Y' are the Y-component after fusion.
The typically selection 0.1~0.9 of the weight of visible images, the weight of infrared image and the weight sum of visible images
For 1, it is preferred that the weight of visible images and the weight of infrared image are 0.5, and now amount of image information is maximum, effect compared with
It is good.
Fused images are detected using the Faster-Rcnn models trained, extracted exhausted in object to be identified
Edge substring.Faster-Rcnn models include RPN convolutional neural networks and Fast-rcnn convolutional neural networks.
Fused images are detected using the Faster-Rcnn models trained, extracted exhausted in object to be identified
Edge substring, including：
The candidate region of insulator chain is identified from fused images using RPN convolutional neural networks；
Using Fast-rcnn convolutional neural networks insulator chain is identified from the candidate region of insulator chain.
In addition, the network structure of structure Faster RCNN models；A RPN convolutional neural networks and one is respectively trained
Fast-rcnn convolutional neural networks, (a) obtain the thick candidate region of insulator using the RPN network processes samples trained.(b)
Thick candidate region is sent into the fast-rcnn networks trained and done and is differentiated, judges whether candidate region is most according to output vector
Excellent insulation subregion, and finally identify insulator chain.
The basic structure of Faster-Rcnn models is still convolutional neural networks, by convolutional neural networks last
A network for being called RPN (Region Proposal Network) is added to realize the partial function, RPN after layer characteristic pattern
Structure is as shown in Figure 2.RPN networks use the sliding window of different area and length-width ratio centered on each point on characteristic pattern
The feature come in acquisition characteristics figure specific region.RPN is connected to last volume by a sliding window (red block in figure)
In the Feature Mapping of lamination output, 256-d vector, the input as output layer are then adjusted to by full articulamentum.Simultaneously
Each sliding window corresponds to k anchor boxes, in the method using 3*3=9 anchor of 3 sizes and 3 ratios.
Each anchor corresponds to a receptive field in artwork, improves scale invariability by this method.
The block mold structures of Faster-Rcnn models is as shown in figure 3, by adjusting network structure, by stage by stage
Training, has all been incorporated into whole target detection identification process in neutral net.The input of model is original image, through excessive
After layer convolution obtains characteristic pattern, detection and the identification function of target are respectively completed by RPN networks and fully-connected network.The instruction of model
Practice process and be divided into 4 steps：
1) the CNN model initialization network parameters of pre-training are used, train RPN networks；
2) caused RoI regional trainings FastRCNN sorter networks in the first step are used；
3) fixed convolution layer parameter, adjusts RPN parameters；
4) fixed convolution layer parameter, full connection layer parameter is adjusted.
The application is directed to the design of the extraction algorithm of the insulation subregion under complex background condition, due to the subgraph face that insulate
Color characteristic is easy to surrounding environment and combined together, the difficulty larger to correct segmentation band under visible light, and infrared image pair
Temperature is more sensitive, and the temperature of insulator is generally greater than surrounding environment, therefore starts with from infrared image, merges visible images,
Complex background can be preferably removed, improves Faster Rcnn algorithm detection efficiencies, and RPN and fast-rcnn is by sharing convolution
Layer parameter, it is that system is simpler, amount of calculation is small, and loss is low, meets system real time requirement.
Claims (7)
- A kind of 1. extra high voltage line insulator chain deep learning recognition methods for merging infrared visible ray, it is characterised in that including：YUV codings are carried out to the visible images of object to be identified, and obtain the Y-component of visible images, U components and V component；It will be seen that after obtaining fusion after Y-component and infrared image the progress fusion treatment of light image, it is combined into visible images U components and V component obtain fused images；Fused images are detected using the Faster-Rcnn models trained, extract the insulator in object to be identified String.
- A kind of 2. extra high voltage line insulator chain deep learning identification side for merging infrared visible ray according to claim 1 Method, it is characterised in that the visible images to object to be identified carry out YUV codings and specifically included：Registration is carried out to visible images using affine transformation；Images after registration is transformed into yuv space.
- A kind of 3. extra high voltage line insulator chain deep learning identification side for merging infrared visible ray according to claim 1 Method, it is characterised in that the Y-component of visible images and the fusion treatment process of infrared image use Wavelet Fusion method.
- A kind of 4. extra high voltage line insulator chain deep learning identification side for merging infrared visible ray according to claim 3 Method, it is characterised in that the Y-component and infrared image of visible images carry out fusion treatment and be specially：Y'=w1 × Y+w2 × IRWherein：Y is the Y-component of visible images, and IR is infrared image, and w1 is the weight of visible images, and w2 is infrared image Weight, Y' be fusion after Y-component.
- A kind of 5. extra high voltage line insulator chain deep learning identification side for merging infrared visible ray according to claim 4 Method, it is characterised in that the weight of the visible images and the weight of infrared image are 0.5.
- A kind of 6. extra high voltage line insulator chain deep learning identification side for merging infrared visible ray according to claim 1 Method, it is characterised in that the Faster-Rcnn models include RPN convolutional neural networks and Fast-rcnn convolutional neural networks.
- A kind of 7. extra high voltage line insulator chain deep learning identification side for merging infrared visible ray according to claim 6 Method, it is characterised in that it is described that fused images are detected using the Faster-Rcnn models trained, extract to be identified Insulator chain in object, including：The candidate region of insulator chain is identified from fused images using RPN convolutional neural networks；Using Fast-rcnn convolutional neural networks insulator chain is identified from the candidate region of insulator chain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710671781.0A CN107507172A (en) | 2017-08-08 | 2017-08-08 | Merge the extra high voltage line insulator chain deep learning recognition methods of infrared visible ray |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710671781.0A CN107507172A (en) | 2017-08-08 | 2017-08-08 | Merge the extra high voltage line insulator chain deep learning recognition methods of infrared visible ray |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107507172A true CN107507172A (en) | 2017-12-22 |
Family
ID=60690675
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710671781.0A Pending CN107507172A (en) | 2017-08-08 | 2017-08-08 | Merge the extra high voltage line insulator chain deep learning recognition methods of infrared visible ray |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107507172A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107680090A (en) * | 2017-10-11 | 2018-02-09 | 电子科技大学 | Based on the electric transmission line isolator state identification method for improving full convolutional neural networks |
| CN108229440A (en) * | 2018-02-06 | 2018-06-29 | 北京奥开信息科技有限公司 | One kind is based on Multi-sensor Fusion indoor human body gesture recognition method |
| CN108509919A (en) * | 2018-04-03 | 2018-09-07 | 哈尔滨哈船智控科技有限责任公司 | A kind of detection and recognition methods based on deep learning to waterline in video or picture |
| CN108537170A (en) * | 2018-04-09 | 2018-09-14 | 电子科技大学 | A kind of power equipment firmware unmanned plane inspection pin missing detection method |
| CN109190633A (en) * | 2018-11-06 | 2019-01-11 | 西安文理学院 | A kind of intelligent object identifying system and control method based on deep learning |
| CN110866548A (en) * | 2019-10-31 | 2020-03-06 | 国网江苏省电力有限公司电力科学研究院 | Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line |
| CN111434494A (en) * | 2019-01-11 | 2020-07-21 | 海德堡印刷机械股份公司 | Missing nozzle detection in printed images |
| CN111611905A (en) * | 2020-05-18 | 2020-09-01 | 沈阳理工大学 | Visible light and infrared fused target identification method |
| WO2020181685A1 (en) * | 2019-03-12 | 2020-09-17 | 南京邮电大学 | Vehicle-mounted video target detection method based on deep learning |
| CN111680592A (en) * | 2020-05-28 | 2020-09-18 | 东风柳州汽车有限公司 | In-vehicle biological detection method, device, equipment and storage medium |
| CN111768436A (en) * | 2020-06-17 | 2020-10-13 | 哈尔滨理工大学 | Improved image feature block registration method based on fast-RCNN |
| CN111768372A (en) * | 2020-06-12 | 2020-10-13 | 国网智能科技股份有限公司 | Method and system for detecting foreign matters in GIS equipment cavity |
| WO2020232704A1 (en) * | 2019-05-23 | 2020-11-26 | 深圳市瑞立视多媒体科技有限公司 | Rigid body identification method and apparatus, system, and terminal device |
| CN112395972A (en) * | 2020-11-16 | 2021-02-23 | 中国科学院沈阳自动化研究所 | Electric power system insulator string identification method based on unmanned aerial vehicle image processing |
| CN112560763A (en) * | 2020-12-24 | 2021-03-26 | 国网上海市电力公司 | Target detection method fusing infrared and visible light images |
| CN112734692A (en) * | 2020-12-17 | 2021-04-30 | 安徽继远软件有限公司 | Transformer equipment defect identification method and device |
| CN112966576A (en) * | 2021-02-24 | 2021-06-15 | 西南交通大学 | System and method for aiming insulator water washing robot based on multi-light source image |
| CN114724042A (en) * | 2022-06-09 | 2022-07-08 | 国网江西省电力有限公司电力科学研究院 | Automatic detection method for zero-value insulator in power transmission line |
| CN116403057A (en) * | 2023-06-09 | 2023-07-07 | 山东瑞盈智能科技有限公司 | Power transmission line inspection method and system based on multi-source image fusion |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102982518A (en) * | 2012-11-06 | 2013-03-20 | 扬州万方电子技术有限责任公司 | Fusion method of infrared image and visible light dynamic image and fusion device of infrared image and visible light dynamic image |
| CN106023129A (en) * | 2016-05-26 | 2016-10-12 | 西安工业大学 | Infrared and visible light image fused automobile anti-blooming video image processing method |
| CN106504233A (en) * | 2016-10-18 | 2017-03-15 | 国网山东省电力公司电力科学研究院 | Image electric power widget recognition methodss and system are patrolled and examined based on the unmanned plane of Faster R CNN |
| CN106503742A (en) * | 2016-11-01 | 2017-03-15 | 广东电网有限责任公司电力科学研究院 | A kind of visible images insulator recognition methods |
-
2017
- 2017-08-08 CN CN201710671781.0A patent/CN107507172A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102982518A (en) * | 2012-11-06 | 2013-03-20 | 扬州万方电子技术有限责任公司 | Fusion method of infrared image and visible light dynamic image and fusion device of infrared image and visible light dynamic image |
| CN106023129A (en) * | 2016-05-26 | 2016-10-12 | 西安工业大学 | Infrared and visible light image fused automobile anti-blooming video image processing method |
| CN106504233A (en) * | 2016-10-18 | 2017-03-15 | 国网山东省电力公司电力科学研究院 | Image electric power widget recognition methodss and system are patrolled and examined based on the unmanned plane of Faster R CNN |
| CN106503742A (en) * | 2016-11-01 | 2017-03-15 | 广东电网有限责任公司电力科学研究院 | A kind of visible images insulator recognition methods |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107680090A (en) * | 2017-10-11 | 2018-02-09 | 电子科技大学 | Based on the electric transmission line isolator state identification method for improving full convolutional neural networks |
| CN108229440A (en) * | 2018-02-06 | 2018-06-29 | 北京奥开信息科技有限公司 | One kind is based on Multi-sensor Fusion indoor human body gesture recognition method |
| CN108509919A (en) * | 2018-04-03 | 2018-09-07 | 哈尔滨哈船智控科技有限责任公司 | A kind of detection and recognition methods based on deep learning to waterline in video or picture |
| CN108509919B (en) * | 2018-04-03 | 2022-04-29 | 哈尔滨哈船智控科技有限责任公司 | Method for detecting and identifying waterline in video or picture based on deep learning |
| CN108537170A (en) * | 2018-04-09 | 2018-09-14 | 电子科技大学 | A kind of power equipment firmware unmanned plane inspection pin missing detection method |
| CN109190633A (en) * | 2018-11-06 | 2019-01-11 | 西安文理学院 | A kind of intelligent object identifying system and control method based on deep learning |
| CN111434494A (en) * | 2019-01-11 | 2020-07-21 | 海德堡印刷机械股份公司 | Missing nozzle detection in printed images |
| US11752775B2 (en) | 2019-01-11 | 2023-09-12 | Heidelberger Druckmaschinen Ag | Method for determining print defects in a printing operation carried out on an inkjet printing machine for processing a print job |
| CN111434494B (en) * | 2019-01-11 | 2021-12-17 | 海德堡印刷机械股份公司 | Missing nozzle detection in printed images |
| JP2021530062A (en) * | 2019-03-12 | 2021-11-04 | 南京郵電大学Ｎａｎｊｉｎｇ Ｕｎｉｖｅｒｓｉｔｙ Ｏｆ Ｐｏｓｔｓ Ａｎｄ Ｔｅｌｅｃｏｍｍｕｎｉｃａｔｉｏｎｓ | In-vehicle video target detection method based on deep learning |
| WO2020181685A1 (en) * | 2019-03-12 | 2020-09-17 | 南京邮电大学 | Vehicle-mounted video target detection method based on deep learning |
| JP7120689B2 (en) | 2019-03-12 | 2022-08-17 | 南京郵電大学 | In-Vehicle Video Target Detection Method Based on Deep Learning |
| WO2020232704A1 (en) * | 2019-05-23 | 2020-11-26 | 深圳市瑞立视多媒体科技有限公司 | Rigid body identification method and apparatus, system, and terminal device |
| CN110866548A (en) * | 2019-10-31 | 2020-03-06 | 国网江苏省电力有限公司电力科学研究院 | Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line |
| CN111611905A (en) * | 2020-05-18 | 2020-09-01 | 沈阳理工大学 | Visible light and infrared fused target identification method |
| CN111611905B (en) * | 2020-05-18 | 2023-04-18 | 沈阳理工大学 | Visible light and infrared fused target identification method |
| CN111680592A (en) * | 2020-05-28 | 2020-09-18 | 东风柳州汽车有限公司 | In-vehicle biological detection method, device, equipment and storage medium |
| CN111768372B (en) * | 2020-06-12 | 2024-03-12 | 国网智能科技股份有限公司 | Method and system for detecting foreign matters in cavity of GIS (gas insulated switchgear) |
| CN111768372A (en) * | 2020-06-12 | 2020-10-13 | 国网智能科技股份有限公司 | Method and system for detecting foreign matters in GIS equipment cavity |
| CN111768436A (en) * | 2020-06-17 | 2020-10-13 | 哈尔滨理工大学 | Improved image feature block registration method based on fast-RCNN |
| CN112395972A (en) * | 2020-11-16 | 2021-02-23 | 中国科学院沈阳自动化研究所 | Electric power system insulator string identification method based on unmanned aerial vehicle image processing |
| CN112734692A (en) * | 2020-12-17 | 2021-04-30 | 安徽继远软件有限公司 | Transformer equipment defect identification method and device |
| CN112734692B (en) * | 2020-12-17 | 2023-12-22 | 国网信息通信产业集团有限公司 | Defect identification method and device for power transformation equipment |
| CN112560763A (en) * | 2020-12-24 | 2021-03-26 | 国网上海市电力公司 | Target detection method fusing infrared and visible light images |
| CN112966576A (en) * | 2021-02-24 | 2021-06-15 | 西南交通大学 | System and method for aiming insulator water washing robot based on multi-light source image |
| CN114724042A (en) * | 2022-06-09 | 2022-07-08 | 国网江西省电力有限公司电力科学研究院 | Automatic detection method for zero-value insulator in power transmission line |
| CN116403057A (en) * | 2023-06-09 | 2023-07-07 | 山东瑞盈智能科技有限公司 | Power transmission line inspection method and system based on multi-source image fusion |
| CN116403057B (en) * | 2023-06-09 | 2023-08-18 | 山东瑞盈智能科技有限公司 | Power transmission line inspection method and system based on multi-source image fusion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107507172A (en) | Merge the extra high voltage line insulator chain deep learning recognition methods of infrared visible ray | |
| CN105631880B (en) | Lane line dividing method and device | |
| CN106683046B (en) | Image real-time splicing method for police unmanned aerial vehicle reconnaissance and evidence obtaining | |
| CN106997597B (en) | It is a kind of based on have supervision conspicuousness detection method for tracking target | |
| CN103745449B (en) | Rapid and automatic mosaic technology of aerial video in search and tracking system | |
| CN113065558A (en) | Lightweight small target detection method combined with attention mechanism | |
| CN106504233A (en) | Image electric power widget recognition methodss and system are patrolled and examined based on the unmanned plane of Faster R CNN | |
| CN104598924A (en) | Target matching detection method | |
| CN106845408A (en) | A kind of street refuse recognition methods under complex environment | |
| CN108399361A (en) | A kind of pedestrian detection method based on convolutional neural networks CNN and semantic segmentation | |
| CN107844795A (en) | Convolutional neural networks feature extracting method based on principal component analysis | |
| CN105046206B (en) | Based on the pedestrian detection method and device for moving prior information in video | |
| CN109559310A (en) | Power transmission and transformation inspection image quality evaluating method and system based on conspicuousness detection | |
| CN108154102A (en) | A kind of traffic sign recognition method | |
| CN110210608A (en) | The enhancement method of low-illumination image merged based on attention mechanism and multi-level features | |
| CN106650668A (en) | Method and system for detecting movable target object in real time | |
| CN108229587A (en) | A kind of autonomous scan method of transmission tower based on aircraft floating state | |
| CN109583349A (en) | A kind of method and system for being identified in color of the true environment to target vehicle | |
| CN107705254A (en) | A kind of urban environment appraisal procedure based on streetscape figure | |
| CN105787470A (en) | Method for detecting power transmission line tower in image based on polymerization multichannel characteristic | |
| CN105678318A (en) | Traffic label matching method and apparatus | |
| CN112487981A (en) | MA-YOLO dynamic gesture rapid recognition method based on two-way segmentation | |
| CN108647689A (en) | Magic square restored method and its device based on GoogLeNet neural networks | |
| Zhu et al. | Fast detection of moving object based on improved frame-difference method | |
| CN109284759A (en) | One kind being based on the magic square color identification method of support vector machines (svm) |
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: 20171222 |
|
| RJ01 | Rejection of invention patent application after publication |