CN114724091A - Method and device for identifying foreign matters on transmission line wire - Google Patents
Method and device for identifying foreign matters on transmission line wire Download PDFInfo
- Publication number
- CN114724091A CN114724091A CN202210631786.1A CN202210631786A CN114724091A CN 114724091 A CN114724091 A CN 114724091A CN 202210631786 A CN202210631786 A CN 202210631786A CN 114724091 A CN114724091 A CN 114724091A
- Authority
- CN
- China
- Prior art keywords
- wire
- transmission line
- foreign matter
- power transmission
- image
- 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
Images
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Electric Cable Installation (AREA)
Abstract
The invention belongs to the field of intelligent operation and detection of power transmission lines, and particularly relates to a method and a device for identifying foreign matters on a lead of a power transmission line. The method comprises the following steps: firstly, acquiring an image of a power transmission line shot by power transmission line monitoring equipment, detecting a wire in the image of the power transmission line by adopting a wire region identification algorithm, and intercepting a wire region; then, respectively inputting the cut conductor region images into a CascadeRCNN conductor foreign matter recognition model and a Yolov5 conductor foreign matter recognition model for conductor foreign matter recognition; and finally, fusing the recognition results of the two models, and filtering the coincidence frame to obtain a final recognition result. The method can accurately identify the foreign matters on the wire of the power transmission line, and can be applied to different power transmission line scenes. The accuracy and the real-time performance of the method can meet the actual requirements of the foreign matter identification of the wire of the power transmission line.
Description
Technical Field
The invention relates to the technical field of intelligent operation and detection of power transmission lines, in particular to a method and a device for identifying foreign matters on a lead of a power transmission line.
Background
The transmission line is one of important infrastructures in China, and normal power supply of various regions is guaranteed through normal operation of the transmission line. With the development requirements of China, more and more construction sites and greenhouse scenes are provided. Due to the change of time, a dust screen of a construction site is aged, a greenhouse film is loosened, and the film is easily hung on a wire of a power transmission line under the influence of strong wind weather, so that the accident of hidden danger of foreign matters of the wire is caused. In addition, kites, balloons, plastic bags and the like are also easy to cause foreign matter tripping accidents. The foreign matter of the lead is one of the important hidden troubles in the power transmission line and is paid attention by the operation and inspection personnel. In recent years, the automatic identification of hidden dangers is carried out in an artificial intelligence mode in the field of intelligent operation and maintenance of power transmission. However, due to the variety of types, forms and colors of the foreign matters of the wires, the detection of the foreign matters of the wires is difficult, so that the foreign matters of the wires are still difficult to find in time, and further, the accidents of the power transmission line caused by the foreign matters of the wires are caused.
Chinese patent CN106960438A discloses a method for identifying foreign matters in a power transmission line based on hough line transformation, which only utilizes threshold segmentation and edge detection to identify foreign matters in the power transmission line, and is difficult to accurately identify foreign matters, and particularly relates to a complex background area in a complex scene, which cannot effectively extract a target area in the foreground. Chinese patent CN111738307A discloses a foreign object identification method, system and computer readable storage medium in power transmission line environment based on fast RCNN, the technical solution thereof has the following disadvantages: because the types of the foreign matters on the lead are more, only the fast RCNN is used for identifying the foreign matters on the lead under the condition of fixed sample size, the detection effect of the foreign matters with few sample types cannot be improved even if data enhancement is carried out, and the foreign matters with multiple types cannot be identified. Chinese patent CN113076899A discloses a method for detecting foreign matter in a high-voltage power transmission line based on a target tracking algorithm, which processes a video stream by adopting a target detection and target tracking method, mainly aiming at identifying moving targets, but in this way, the video stream is misreported by small targets such as birds, and the difference of processing speed of the video stream is larger than that of an image.
In summary, those skilled in the art need to solve the above problems how to provide an accurate and efficient method and apparatus for identifying a foreign object on a power transmission line conductor, so as to improve the accuracy of identifying the foreign object on the power transmission line conductor and avoid power transmission line accidents caused by the foreign object on the power transmission line conductor.
Disclosure of Invention
The invention aims to provide a method and a device for identifying foreign matters on a wire of a power transmission line, which can overcome the defects in the prior art, can quickly identify the foreign matters on the wire by using a wire region identification method and combining a multi-model fusion technology, meet the real-time requirement, realize the accurate identification of the foreign matters on the wire, and effectively avoid the power transmission line accidents caused by the hidden danger of the foreign matters on the wire.
In order to achieve the purpose, the invention adopts the following technical scheme:
a method for identifying a foreign matter on a wire of a power transmission line comprises the following steps:
s1: and acquiring an image of the power transmission line through the power transmission line monitoring device, processing the image by adopting a wire area identification algorithm to obtain a wire area coordinate, and storing the wire area coordinate in an equipment information table.
S2: and acquiring an image of the power transmission line with the conductor foreign matter through the power transmission line monitoring device, marking the conductor foreign matter, and constructing a conductor foreign matter data set.
S3: a CascadeRCNN lead foreign matter recognition model and a Yolov5 lead foreign matter recognition model based on Resnet101+ FPN are respectively constructed, and a lead foreign matter data set is utilized to respectively train the two lead foreign matter recognition models.
S4: and according to the wire region coordinates stored in the equipment information table, cutting the power transmission line image shot by the power transmission line monitoring device in real time, and respectively inputting the cut image into a CascadeRCNN wire foreign matter recognition model and a Yolov5 wire foreign matter recognition model for wire foreign matter recognition.
S5: taking the union of the recognition results of the CascadeRCNN conductor foreign matter recognition model and the Yolov5 conductor foreign matter recognition model as a final recognition result, and filtering the coincidence frame by using a non-maximum suppression algorithm.
Further, the wire area identification algorithm in step S1 specifically includes the following steps:
s11: and carrying out Gaussian filtering processing on the image of the power transmission line, and eliminating Gaussian noise in the image of the power transmission line to obtain a filtered image.
S12: and (4) carrying out binarization processing on the filtered image by using an adaptive threshold algorithm, and removing background information in the image to obtain a binarized image.
S13: and deleting the interference noise points in the binary image by using an expansion corrosion algorithm to obtain a foreground image.
S14: and detecting the conducting wire in the foreground image by using a Hough line transformation method, acquiring the position of the conducting wire, and determining the regional coordinate of the conducting wire according to the position of the conducting wire.
Further, the step S2, where the image of the power transmission line with the wire foreign matter is acquired by the power transmission line monitoring device, and the wire foreign matter is labeled to construct the wire foreign matter data set, specifically includes the following steps:
s21: and (3) using Labeling software to perform rectangular Labeling on the potential hazard target of the foreign matter of the wire in the image in the wire area in the image of the power transmission line. The common foreign body hidden danger targets of the forehead wires comprise dust screens, plastic greenhouse films, plastic bags and the like.
S22: according to the wire area coordinates in the equipment information table, cutting and storing the marked transmission line image, mixing the image shot by the transmission line monitoring equipment and the cut image to form a wire foreign matter data set, and according to the following steps: 1: 1 into a training set, a validation set and a test set.
Further, in step S3, "a CascadeRCNN wire foreign object recognition model and a YOLOV5 wire foreign object recognition model based on Resnet101+ FPN are respectively constructed, and the two wire foreign object recognition models are respectively trained by using the wire foreign object data set; ", which comprises the following steps:
s31: randomly turning the images in the training set according to the probability, adding the turned images into the training set, and enhancing the data of the training set; the present invention employs vertical inversion with a probability of 0.5 as data enhancement.
S32: the CascadeRCNN wire foreign body recognition model selects images with image input sizes of (1216, 1621) pixels in a training set to perform model training, and the YOLOV5 model selects images with image input sizes of (1280 ) pixels in the training set after data enhancement to perform model training.
The invention also relates to a device for identifying the foreign matters on the wire of the power transmission line, which comprises a wire area coordinate acquisition module, a wire foreign matter identification model construction module and a wire foreign matter identification module.
And the wire area coordinate acquisition module is used for processing the power transmission line image shot by the power transmission line monitoring device by adopting a wire area recognition algorithm to obtain a wire area coordinate, and storing the wire area coordinate in the equipment information table.
The wire foreign matter recognition model building module is used for conducting wire foreign matter labeling on a power transmission line image which is shot by the power transmission line monitoring device and provided with wire foreign matters, building a wire foreign matter data set, respectively building a CascadeRCNN wire foreign matter recognition model and a Yolov5 wire foreign matter recognition model based on Resnet101+ FPN, and respectively training the two wire foreign matter recognition models by utilizing the wire foreign matter data set.
The wire foreign matter identification module is used for cutting the image of the power transmission line shot by the power transmission line monitoring device in real time according to the wire region coordinates stored in the equipment information table, and respectively inputting the cut image into a CascadeRCNN wire foreign matter identification model and a Yolov5 wire foreign matter identification model for wire foreign matter identification; taking the union of the recognition results of the CascadeRCNN conductor foreign matter recognition model and the Yolov5 conductor foreign matter recognition model as a final recognition result, and filtering the coincidence frame by using a non-maximum suppression algorithm.
Compared with the prior art, the invention has the following beneficial effects:
(1) according to the invention, a real-time image of the power transmission line is obtained through the power transmission line monitoring device, the conductor region is identified by adopting Hough line transformation, and the coordinate information of the conductor region is stored in the equipment information table. And intercepting the conductor area from an original image shot by equipment through the conductor area coordinates in the equipment information table, and inputting the intercepted image into a CascadeRCNN conductor foreign matter recognition model and a Yolov5 conductor foreign matter recognition model for conductor foreign matter recognition.
(2) The method combines a multi-model fusion strategy, uses a CascadeRCNN conductor foreign matter recognition model and a Yolov5 conductor foreign matter recognition model to recognize the foreign matters on the power transmission line conductor, can effectively recognize the conductor foreign matters, avoids the missing report of the conductor foreign matters, combines the recognition results of the CascadeRCNN conductor foreign matter recognition model and the Yolov5 conductor foreign matter recognition model, can effectively ensure the recognition rate of the conductor foreign matters, and avoids the hidden trouble accidents caused by the conductor foreign matters. The accuracy and the real-time performance of the method meet the actual requirement of identifying the foreign matters of the wires on the power transmission line.
(3) In the method for detecting the foreign matters on the lead based on deep learning in the prior art, for small targets with various types of foreign matters and difficulty in sample collection, adaptive local detection on a lead area is not considered, and a single model is difficult to identify all types of foreign matters on the lead. The invention carries out targeted strategy processing aiming at the special target type of the foreign body of the wire, has better effects on detection precision and processing speed, and has stronger robustness of the combination of the two models. The target detection algorithm adopted by the invention can adapt to various complex scenes, and the recognition rate of the foreign matters of the wires is greatly improved through a model fusion mode. In addition, the invention can train models aiming at different types of foreign matters by using a dual-model fusion mode, thereby improving the identification precision. The invention uses the mode of dual-model fusion detection, but is still much faster than the video stream processing, and the processing speed and the identification precision are increased aiming at the wire region identification.
Drawings
FIG. 1 is a flow chart of a method for identifying a foreign object on a wire of a power transmission line according to the present invention;
FIG. 2 is a block diagram of a device for identifying foreign matters on a wire of a power transmission line according to the present invention;
FIG. 3 is an image taken by the monitoring device for power transmission line according to the present invention;
FIG. 4 is an image of a lead area taken from FIG. 1;
FIG. 5 is a result image of the CascadeRCNN wire foreign object recognition model of the present invention recognizing FIG. 4;
FIG. 6 is a resulting image of the YOLOV5 wire foreign object identification model of FIG. 4 in accordance with the present invention;
fig. 7 is an image of the recognition result obtained by fusing the recognition results of the two wire foreign object recognition models of CascadeRCNN and yoloov 5 in the present invention.
Detailed Description
The present invention will be further described with reference to the following examples.
Fig. 1 shows a flow chart of a method for identifying a foreign matter on a power transmission line conductor in the invention.
And 102, identifying the wire area in the power transmission line image by using a wire area identification algorithm, acquiring the position coordinate of the wire area, and storing the position coordinate into an information table of the equipment.
And 103, cutting the image of the power transmission line by adopting an automatic threshold segmentation method, and cutting out the lead area.
And step 104 and step 105, respectively constructing a Cascade RCNN model and a Yolov5 model, and training the two models. The CascadeRCNN model and the Yolov5 model based on Resnet101+ FPN adopted by the invention are the existing open source models. The two models are subjected to model training by adopting the same training set, each round of precision testing is performed on the verification set in the training process, the optimal model in the training process is found, and after the training is finished, the generalization performance is tested on the testing set, so that the model is prevented from being over-fitted.
And 106 and 107, conducting wire foreign matter recognition on the power transmission line image shot by the monitoring equipment in real time and cut in the steps 102 and 103 by using the trained Cascade RCNN model and the trained Yolov5 model.
And step 108, fusing the recognition results of the two models. And step 109, filtering the coincidence frame of the fusion result to obtain a final recognition result. The recognition results of the two models are coordinate information, and the coordinate information is put together, which is equivalent to the combination of the elements of the array. After the recognition results are combined, different algorithms can generate recognized coordinate information for the same target, namely two frames are arranged on the same target, only one frame is needed after the output of the combination, and redundant frames are removed through a non-maximum suppression algorithm.
Fig. 2 is a block diagram illustrating a power transmission line conductor foreign matter recognition apparatus according to the present invention. As shown in fig. 2, the transmission line conductor foreign object recognition apparatus 200 includes a conductor region coordinate acquisition module 201, a conductor foreign object recognition model construction module 202, and a conductor foreign object recognition module 203.
The lead area coordinate obtaining module 201 is configured to process the power transmission line image shot by the power transmission line monitoring device by using a lead area recognition algorithm, obtain a lead area coordinate, and store the lead area coordinate in the device information table.
The lead foreign matter identification model building module 202 is used for conducting lead foreign matter labeling on a power transmission line image with lead foreign matters shot by a power transmission line monitoring device, building a lead foreign matter data set, respectively building a CascadeRCNN lead foreign matter identification model and a Yolov5 lead foreign matter identification model based on Resnet101+ FPN, and respectively training the two lead foreign matter identification models by utilizing the lead foreign matter data set.
The lead foreign matter identification module 203 is used for cutting the electric transmission line image shot by the electric transmission line monitoring device in real time according to the lead region coordinates stored in the equipment information table, and respectively inputting the cut image into a CascadeRCNN lead foreign matter identification model and a Yolov5 lead foreign matter identification model for lead foreign matter identification; taking the union of the recognition results of the CascadeRCNN conductor foreign matter recognition model and the Yolov5 conductor foreign matter recognition model as a final recognition result, and filtering the coincidence frame by using a non-maximum suppression algorithm.
The following describes a method for identifying a foreign object on a wire of a power transmission line according to an embodiment of the present invention.
The method comprises the steps of obtaining an image shot by certain power transmission line monitoring equipment, inputting the image into a wire area identification algorithm, obtaining position coordinates of a wire area, and storing the position coordinates into an information table of the equipment. And continuously collecting images subsequently shot by the equipment, and intercepting the conductor area in each power transmission line image by combining the conductor position coordinates in the equipment information table. And respectively inputting the transmission line image and the wire region shot by the monitoring equipment into a CascadeRCNN wire foreign matter recognition model and a Yolov5 wire foreign matter recognition model to obtain detection results. And taking the union of the recognition results of the two models as a final recognition result.
In this embodiment, the method for identifying the foreign matter on the wire of the power transmission line includes the following steps:
s1: the power transmission line image is obtained through the power transmission line monitoring device, and the obtained power transmission line image is shown in fig. 3. The position of the wire is detected by Hough line transformation, and the coordinate information (xmin, ymin, xmax, ymax) of the wire area is obtained as (1, 1972, 1335), which is stored in a device information table. xmin represents the lower left abscissa of the wire position, ymin represents the lower left ordinate of the wire position, xmax represents the upper right abscissa of the wire position, and ymax represents the upper right ordinate of the wire position.
S2: acquiring 3 thousands of power transmission line images with the wire foreign matter through a power transmission line monitoring device, marking the wire foreign matter on each power transmission line image, constructing a wire foreign matter data set, and according to the following steps: 1: 1, the wire foreign body data set is divided into a training set, a verification set and a test set.
S3: and respectively training a Cascade RCNN lead foreign matter recognition model and a Yolov5 lead foreign matter recognition model which are built based on Resnet101+ FPN by using a training set, and training each model by 50 epochs to obtain the trained Cascade RCNN lead foreign matter recognition model and a Yolov5 lead foreign matter recognition model.
S4: the image in step S2 captured by the device is clipped by the wire area coordinates in the device information table, and as shown in fig. 4, the clipped image is input to the CascadeRCNN wire foreign matter recognition model and the YOLOV5 wire foreign matter recognition model, respectively, and wire foreign matter recognition is performed on the image using both models. The recognition result of the cassadercnn wire foreign matter recognition model is shown in fig. 5, and the recognition result of the YOLOV5 wire foreign matter recognition model is shown in fig. 5.
S5: taking a union of a CascadeRCNN conductor foreign matter recognition model and a Yolov5 conductor foreign matter recognition model as a recognition result, and filtering coincident frames in the recognition result by using a non-maximum suppression method to obtain a final recognition result. The final recognition result is shown in fig. 7, the image has the hidden danger of foreign matters in the wires, and the image is pushed to a power transmission line operator for processing.
In this embodiment, combine CascadeRCNN wire foreign matter recognition model and YOLOV5 wire foreign matter recognition model to use, adopt the regional recognition algorithm of wire and the wire foreign matter among the multi-model fusion technique discernment transmission line, realized the intelligent recognition of transmission line wire foreign matter, promoted the recognition rate of transmission line wire foreign matter, can effectively avoid the transmission line accident because of wire foreign matter hidden danger arouses.
The above-mentioned embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solution of the present invention by those skilled in the art should fall within the protection scope defined by the claims of the present invention without departing from the spirit of the present invention.
Claims (5)
1. A method for identifying a foreign matter on a wire of a power transmission line is characterized by comprising the following steps: the method comprises the following steps:
s1: acquiring a power transmission line image through a power transmission line monitoring device, processing the image by adopting a wire area identification algorithm to obtain a wire area coordinate, and storing the wire area coordinate in an equipment information table;
s2: acquiring an image of the power transmission line with the wire foreign matter through a power transmission line monitoring device, marking the wire foreign matter, and constructing a wire foreign matter data set;
s3: respectively constructing a CascadeRCNN lead foreign body recognition model and a Yolov5 lead foreign body recognition model based on Resnet101+ FPN, and respectively training the two lead foreign body recognition models by utilizing a lead foreign body data set;
s4: cutting the transmission line image shot by the transmission line monitoring device in real time according to the wire region coordinates stored in the equipment information table, and respectively inputting the cut image into a CascadeRCNN wire foreign matter recognition model and a Yolov5 wire foreign matter recognition model for wire foreign matter recognition;
s5: taking the union of the recognition results of the CascadeRCNN conductor foreign matter recognition model and the Yolov5 conductor foreign matter recognition model as a final recognition result, and filtering the coincidence frame by using a non-maximum suppression algorithm.
2. The method for identifying the foreign matter on the conducting wire of the power transmission line according to claim 1, wherein the method comprises the following steps: the wire area identification algorithm in step S1 specifically includes the following steps:
s11: carrying out Gaussian filtering processing on the image of the power transmission line, and eliminating Gaussian noise in the image of the power transmission line to obtain a filtering image;
s12: carrying out binarization processing on the filtered image by using a self-adaptive threshold algorithm, and removing background information in the image to obtain a binarized image;
s13: deleting interference noise points in the binary image by using an expansion corrosion algorithm to obtain a foreground image;
s14: and detecting the conducting wire in the foreground image by using a Hough line transformation method, acquiring the position of the conducting wire, and determining the regional coordinate of the conducting wire according to the position of the conducting wire.
3. The method for identifying the foreign matter on the conducting wire of the power transmission line according to claim 1, wherein the method comprises the following steps: in step S2, the method includes the steps of acquiring an image of the power transmission line with the wire foreign object through the power transmission line monitoring device, labeling the wire foreign object, and constructing a wire foreign object data set, and specifically includes the following steps:
s21: using Labeling software to perform rectangular Labeling on a lead foreign matter hidden danger target in an image in a lead area in a power transmission line image;
s22: according to the wire area coordinates in the equipment information table, cutting and storing the marked transmission line image, mixing the image shot by the transmission line monitoring equipment and the cut image to form a wire foreign matter data set, and according to the following steps: 1: 1 into a training set, a validation set and a test set.
4. The method for identifying the foreign matter on the conducting wire of the power transmission line according to claim 3, wherein the method comprises the following steps: step S3, establishing a CascadeRCNN lead foreign body recognition model and a Yolov5 lead foreign body recognition model based on Resnet101+ FPN respectively, and training the two lead foreign body recognition models respectively by utilizing lead foreign body data sets; ", which comprises the following steps:
s31: randomly turning the images in the training set according to the probability, adding the turned images into the training set, and enhancing the data of the training set;
s32: the CascadeRCNN wire foreign body recognition model selects images with image input sizes of (1216, 1621) pixels in a training set to perform model training, and the YOLOV5 model selects images with image input sizes of (1280 ) pixels in the training set after data enhancement to perform model training.
5. The utility model provides a transmission line wire foreign matter recognition device which characterized in that: the system comprises a wire region coordinate acquisition module, a wire foreign matter identification model construction module and a wire foreign matter identification module;
the wire area coordinate acquisition module is used for processing the power transmission line image shot by the power transmission line monitoring device by adopting a wire area recognition algorithm to obtain a wire area coordinate, and storing the wire area coordinate in an equipment information table;
the wire foreign matter identification model building module is used for conducting wire foreign matter labeling on a power transmission line image with wire foreign matters shot by a power transmission line monitoring device, building a wire foreign matter data set, respectively building a CascadeRCNN wire foreign matter identification model and a Yolov5 wire foreign matter identification model based on Resnet101+ FPN, and respectively training the two wire foreign matter identification models by utilizing the wire foreign matter data set;
the wire foreign matter identification module is used for cutting the image of the power transmission line shot by the power transmission line monitoring device in real time according to the wire region coordinates stored in the equipment information table, and respectively inputting the cut image into a CascadeRCNN wire foreign matter identification model and a Yolov5 wire foreign matter identification model for wire foreign matter identification; taking the union of the recognition results of the CascadeRCNN conductor foreign matter recognition model and the YoloV5 conductor foreign matter recognition model as a final recognition result, and filtering the coincidence frame by using a non-maximum suppression algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210631786.1A CN114724091A (en) | 2022-06-07 | 2022-06-07 | Method and device for identifying foreign matters on transmission line wire |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210631786.1A CN114724091A (en) | 2022-06-07 | 2022-06-07 | Method and device for identifying foreign matters on transmission line wire |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114724091A true CN114724091A (en) | 2022-07-08 |
Family
ID=82232488
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210631786.1A Pending CN114724091A (en) | 2022-06-07 | 2022-06-07 | Method and device for identifying foreign matters on transmission line wire |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114724091A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115294740A (en) * | 2022-07-25 | 2022-11-04 | 国网河北省电力有限公司雄安新区供电公司 | Grid calibration method for overhead transmission line channel protection area |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106960438A (en) * | 2017-03-25 | 2017-07-18 | 安徽继远软件有限公司 | Method for recognizing impurities to transmission line of electricity is converted based on Hough straight line |
| CN108921826A (en) * | 2018-06-13 | 2018-11-30 | 山东信通电子股份有限公司 | The transmission line of electricity that super-pixel segmentation is combined with deep learning invades object detecting method |
| CN109801302A (en) * | 2018-12-14 | 2019-05-24 | 华南理工大学 | A kind of ultra-high-tension power transmission line foreign matter detecting method based on binocular vision |
| CN110689531A (en) * | 2019-09-23 | 2020-01-14 | 云南电网有限责任公司电力科学研究院 | Automatic power transmission line machine inspection image defect identification method based on yolo |
| CN111353413A (en) * | 2020-02-25 | 2020-06-30 | 武汉大学 | Low-missing-report-rate defect identification method for power transmission equipment |
| CN112069894A (en) * | 2020-08-03 | 2020-12-11 | 许继集团有限公司 | Wire strand scattering identification method based on fast-RCNN model |
| CN113076899A (en) * | 2021-04-12 | 2021-07-06 | 华南理工大学 | High-voltage transmission line foreign matter detection method based on target tracking algorithm |
| CN113673514A (en) * | 2021-08-11 | 2021-11-19 | 国网山东省电力公司微山县供电公司 | Method and system for detecting invasion of foreign matters into power transmission line |
| CN113744267A (en) * | 2021-11-04 | 2021-12-03 | 智洋创新科技股份有限公司 | Method for detecting icing and estimating thickness of transmission conductor based on deep learning |
| CN113903009A (en) * | 2021-12-10 | 2022-01-07 | 华东交通大学 | Railway foreign matter detection method and system based on improved YOLOv3 network |
-
2022
- 2022-06-07 CN CN202210631786.1A patent/CN114724091A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106960438A (en) * | 2017-03-25 | 2017-07-18 | 安徽继远软件有限公司 | Method for recognizing impurities to transmission line of electricity is converted based on Hough straight line |
| CN108921826A (en) * | 2018-06-13 | 2018-11-30 | 山东信通电子股份有限公司 | The transmission line of electricity that super-pixel segmentation is combined with deep learning invades object detecting method |
| CN109801302A (en) * | 2018-12-14 | 2019-05-24 | 华南理工大学 | A kind of ultra-high-tension power transmission line foreign matter detecting method based on binocular vision |
| CN110689531A (en) * | 2019-09-23 | 2020-01-14 | 云南电网有限责任公司电力科学研究院 | Automatic power transmission line machine inspection image defect identification method based on yolo |
| CN111353413A (en) * | 2020-02-25 | 2020-06-30 | 武汉大学 | Low-missing-report-rate defect identification method for power transmission equipment |
| CN112069894A (en) * | 2020-08-03 | 2020-12-11 | 许继集团有限公司 | Wire strand scattering identification method based on fast-RCNN model |
| CN113076899A (en) * | 2021-04-12 | 2021-07-06 | 华南理工大学 | High-voltage transmission line foreign matter detection method based on target tracking algorithm |
| CN113673514A (en) * | 2021-08-11 | 2021-11-19 | 国网山东省电力公司微山县供电公司 | Method and system for detecting invasion of foreign matters into power transmission line |
| CN113744267A (en) * | 2021-11-04 | 2021-12-03 | 智洋创新科技股份有限公司 | Method for detecting icing and estimating thickness of transmission conductor based on deep learning |
| CN113903009A (en) * | 2021-12-10 | 2022-01-07 | 华东交通大学 | Railway foreign matter detection method and system based on improved YOLOv3 network |
Non-Patent Citations (1)
| Title |
|---|
| 金立军等: "基于航拍图像的输电线路异物识别", 《同济大学学报(自然科学版)》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115294740A (en) * | 2022-07-25 | 2022-11-04 | 国网河北省电力有限公司雄安新区供电公司 | Grid calibration method for overhead transmission line channel protection area |
| CN115294740B (en) * | 2022-07-25 | 2023-11-07 | 国网河北省电力有限公司雄安新区供电公司 | Gridding calibration method for overhead transmission line channel protection area |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110826538B (en) | Abnormal off-duty identification system for electric power business hall | |
| WO2021042682A1 (en) | Method, apparatus and system for recognizing transformer substation foreign mattter, and electronic device and storage medium | |
| CN110084165B (en) | Intelligent identification and early warning method for abnormal events in open scene of power field based on edge calculation | |
| CN103093201B (en) | Vehicle-logo location recognition methods and system | |
| CN106331636A (en) | Intelligent video monitoring system and method of oil pipelines based on behavioral event triggering | |
| CN111161312B (en) | Object trajectory tracking and identifying device and system based on computer vision | |
| CN113947731B (en) | Foreign matter identification method and system based on contact net safety inspection | |
| CN103745224A (en) | Image-based railway contact net bird-nest abnormal condition detection method | |
| CN104506819A (en) | Multi-camera real-time linkage mutual feedback tracing system and method | |
| CN112364778A (en) | Power plant safety behavior information automatic detection method based on deep learning | |
| CN113298077A (en) | Transformer substation foreign matter identification and positioning method and device based on deep learning | |
| CN112084892B (en) | Road abnormal event detection management device and method thereof | |
| CN111145222A (en) | Fire detection method combining smoke movement trend and textural features | |
| CN116052222A (en) | Cattle face recognition method for naturally collecting cattle face image | |
| CN113052894A (en) | Door opening and closing state detection method and system based on image semantic segmentation | |
| CN114724091A (en) | Method and device for identifying foreign matters on transmission line wire | |
| CN114998815A (en) | Traffic vehicle identification tracking method and system based on video analysis | |
| CN114863311A (en) | Automatic tracking method and system for inspection target of transformer substation robot | |
| CN114332781A (en) | Intelligent license plate recognition method and system based on deep learning | |
| CN111597939B (en) | High-speed rail line nest defect detection method based on deep learning | |
| WO2022121021A1 (en) | Identity card number detection method and apparatus, and readable storage medium and terminal | |
| CN109948618A (en) | A kind of terminal, the system and method for remote Car license recognition | |
| CN117475353A (en) | Video-based abnormal smoke identification method and system | |
| CN112669269A (en) | Pipeline defect classification and classification method and system based on image recognition | |
| CN113221603A (en) | Method and device for detecting shielding of monitoring equipment by foreign matters |
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: 20220708 |
|
| RJ01 | Rejection of invention patent application after publication |