CN112733610A - Power transmission line image recognition detection method based on unmanned aerial vehicle - Google Patents

Power transmission line image recognition detection method based on unmanned aerial vehicle Download PDF

Info

Publication number
CN112733610A
CN112733610A CN202011476612.XA CN202011476612A CN112733610A CN 112733610 A CN112733610 A CN 112733610A CN 202011476612 A CN202011476612 A CN 202011476612A CN 112733610 A CN112733610 A CN 112733610A
Authority
CN
China
Prior art keywords
image
transmission line
power transmission
aerial vehicle
unmanned aerial
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
Application number
CN202011476612.XA
Other languages
Chinese (zh)
Inventor
于常乐
朱远达
赵会
张肃
徐浩然
季彦辰
赵泓博
伊永飞
杨林
于明浩
李文文
万家
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Corp of China SGCC
State Grid Liaoning Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
State Grid Liaoning Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by State Grid Corp of China SGCC, State Grid Liaoning Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN202011476612.XA priority Critical patent/CN112733610A/en
Publication of CN112733610A publication Critical patent/CN112733610A/en
Pending legal-status Critical Current

Links

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
    • G06V20/176—Urban or other man-made structures
    • 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
    • G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • 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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/257—Belief theory, e.g. Dempster-Shafer
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/259—Fusion by voting

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Probability & Statistics with Applications (AREA)
  • Image Analysis (AREA)

Abstract

The invention relates to a power transmission line image identification and detection method based on an unmanned aerial vehicle. Firstly, a conductor image of a power transmission line is obtained through an unmanned aerial vehicle image acquisition system, an 3/4 denoising clustering method is provided for image enhancement, and a Canny edge detection method is adopted for extracting a conductor region outline. Then, an image segmentation identification method based on clustering and DS evidence fusion theory is provided, the state of wire breakage is simply and effectively identified, and an identification model base of the power transmission line in different states is established. And finally, the AdaBoost algorithm-based power transmission line image comparison method is used for quickly and accurately identifying the power transmission line image, so that the working efficiency and the detection quality of power transmission line overhauling and operation inspection are improved.

Description

Power transmission line image recognition detection method based on unmanned aerial vehicle
Technical Field
The invention relates to the technical fields of information, big data, electric power systems, computer science and the like, in particular to a power transmission line image identification and detection method based on an unmanned aerial vehicle.
Background
The reliability, stability and power supply quality of the electric power energy play a crucial role in national economic development. The coverage of the transmission line in China is wide, the landform is complex, and greater challenges are provided for normal operation, detection and maintenance of the transmission line. The traditional manual inspection mode has the advantages of low efficiency, long period, high requirement and low safety, and cannot meet the maintenance requirement of the power transmission line under the current national conditions. And the unmanned aerial vehicle inspection tour mode taking the unmanned aerial vehicle as the core can observe the power transmission line at a short distance through the camera equipment carried at high altitude, and returns images for detection. The method has the advantages of low cost, easy operation, high safety, flexibility and the like, and thus, the method becomes the focus of power transmission line detection research in recent years. It is worth noting that at present, the inspection mode of unmanned aerial vehicle tour is still mostly in the stage of manually checking and shooting videos and images, and the research on computer automatic image recognition is very little.
Disclosure of Invention
The invention provides an unmanned aerial vehicle-based power transmission line image recognition detection method, which is used for checking and judging the running state of a power transmission line, so that the labor intensity of operators can be reduced, and the working efficiency, safety and accuracy of field recognition can be greatly improved.
An unmanned aerial vehicle-based power transmission line image identification detection method comprises the following steps:
step 1, acquiring a conductor image of a power transmission line through an unmanned aerial vehicle image acquisition system, performing image enhancement processing by adopting an 3/4 denoising clustering method, extracting a conductor region in the image, and extracting a conductor region outline by adopting a Canny edge detection method;
step 2, identifying the broken state of the lead by using an image segmentation identification method based on clustering and DS evidence fusion theory;
step 3, establishing an identification model library of the power transmission line in different states;
and 4, rapidly and accurately identifying the power transmission line by using the power transmission line image comparison method based on the AdaBoost algorithm.
In the step 1, the 3/4 denoising and clustering method comprises the following steps:
setting an original image as an mxn order matrix A:
Figure BDA0002835781740000011
in the formula: a isijRepresenting a gray value of a pixel of the image;
and (3) transforming the matrix A by 3/4 values to obtain an image matrix H:
Figure BDA0002835781740000021
order:
f(i,j)＝aij+ai+1,j+ai,j+1+ai+1,j+1
Figure BDA0002835781740000022
Figure BDA0002835781740000023
wherein d is a filtering threshold value, a sensitive area can be defined on the image, and the area is obtained by adopting a maximum variance threshold value method to calculate.
In the step 1, the Canny edge detection method comprises the following steps:
the Canny algorithm smoothes the image by using a two-dimensional gaussian function, which has the following form:
Figure BDA0002835781740000024
the corresponding gradient vector is:
Figure BDA0002835781740000025
two-dimensional gaussian filter functions are generally not directly convolved with an image, but are decomposed into two one-dimensional row and column filters:
Figure BDA0002835781740000026
Figure BDA0002835781740000027
wherein:
Figure BDA0002835781740000031
Figure BDA0002835781740000032
Figure BDA0002835781740000033
Figure BDA0002835781740000034
then, convolution operation is respectively carried out on the images to obtain:
Figure BDA0002835781740000035
Figure BDA0002835781740000036
where x is the convolution sign, k is a constant coefficient, and σ is a spatial scale parameter of the gaussian filter function.
In the step 2, the image clustering (K-means) segmentation comprises the following steps:
1) initializing, and determining a cluster number k and a cluster center initial value;
2) calculating Euclidean distance, and calculating each sample Z according to formula (1)iWith each cluster center OjEuclidean distance E ofij，
Figure BDA0002835781740000037
In the formula: zi,mIs the mth attribute of sample Z; m is a sample dimension, namely the number of attributes;
3) dividing clusters, namely dividing each sample into clusters with the minimum Euclidean distance to the sample;
4) updating the clusters, and calculating the average value of each sample in each cluster to be used as a new cluster center;
5) judging convergence, if the moving distance of the cluster center is smaller than a set value or reaches an iteration number limit value, judging the convergence, and ending clustering; otherwise, returning to the step 2 to continue calculation;
the method comprises the following steps of inputting a sample of the clustering algorithm as a pixel point of a power transmission line, compressing an image into H x w pixels before inputting in order to improve the calculation efficiency, selecting a hue value of an image segmentation target as a cluster center initial value, wherein the hue value H is an attribute of the pixel point, and the calculation method is shown as the following formula:
Figure BDA0002835781740000041
r, G and B are the values of the red, green and blue color channels, respectively: t ismax＝max(R,G,B)； Tmin＝min(R,G,B)。
In the step 2, the D-S evidence theory is that a finite set of all possible results is called an identification frame, elements in the finite set are mutually exclusive in pairs, and 2 is used in the given identification framenAnd satisfies the following conditions:
Figure BDA0002835781740000042
m(A1)＝1-m(A2)
mi(A1)＝1-mi(A2)＝Pi
m is 2nBasic probability distribution.
In step 4, the AdaBoost algorithm is as follows: the voting process with the best characteristics is that based on a weighted voting mechanism, the answer to a certain question is judged by a large number of weighted combinations of classification functions, the weight of the classifier with the better effect is increased in each calculation iteration, and the weight of the classifier with the relatively poor effect is gradually reduced.
The invention has the following advantages and effects:
after adopting 3/4 value transformation and iteration processing, the image clustering convergence can obtain good effect in the aspects of noise elimination and edge preservation. By adopting the 3/4 denoising clustering method, not only can noise be effectively removed, but also approximate gray level image points can be clustered.
The image segmentation and identification method based on clustering and DS evidence fusion theory is utilized to simply and effectively identify the state of wire breakage.
The AdaBoost algorithm is high in speed and strong in noise resistance, and the working efficiency and the detection quality of power transmission line maintenance and operation inspection are obviously improved.
Drawings
FIG. 1 is a flow chart of an image clustering (K-means) segmentation algorithm.
Fig. 2 is a flowchart showing the division result checking.
Fig. 3 is a flowchart of image comparison based on the AdaBoost algorithm.
Fig. 4 is a general architecture diagram of the unmanned aerial vehicle-based power transmission line image recognition detection method.
Detailed Description
The embodiment of the invention provides a system architecture based on a new energy cloud platform, which is used for solving the problem of new energy consumption and realizing unified scheduling and management of new energy. The problems that the existing platform is single in covering service, too large in analysis result and the like are effectively solved. In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the embodiments described below are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, an embodiment of the present invention provides an unmanned aerial vehicle-based power transmission line image recognition and detection method, and research focuses on building an overall algorithm framework of the unmanned aerial vehicle-based power transmission line image recognition and detection method. The system also comprises an image acquisition system and a monitoring center. The image acquisition system consisted of a 10 frame DJIJY-UAV model drone equipped with a digital camera with a resolution of 1200 ten thousand pixels with a shutter speed of 8-1/8000 seconds. The monitoring center consists of a computer, a video server and a database server. The communication signal adopts a 4G network. Please refer to fig. 1 for the flow of the image clustering (K-means) segmentation algorithm. Please refer to fig. 2 for the flow of checking the segmentation result. Please refer to fig. 3 for an image comparison process based on the AdaBoost algorithm.
Referring to fig. 4, for a general architecture of an unmanned aerial vehicle power transmission line image recognition detection method, 3/4 denoising clustering and Canny edge detection methods are provided for denoising images and extracting conductor region features of an unmanned aerial vehicle power transmission line inspection image recognition model. The image segmentation and identification method based on clustering and DS evidence fusion theory is provided, the state of wire breakage is simply and effectively identified, and an identification model base of the power transmission line in different states is established. Aiming at the establishment of an image dynamic rapid comparison identification model based on an identification model library, an AdaBoost algorithm is provided for rapidly and accurately comparing line images, so that the working efficiency and the detection quality of power transmission line inspection are remarkably improved.
The method specifically comprises the following steps:
step 1, acquiring a conductor image of a power transmission line through an unmanned aerial vehicle image acquisition system, performing image enhancement processing by adopting an 3/4 denoising clustering method, extracting a conductor region in the image, and extracting a conductor region outline by adopting a Canny edge detection method;
step 2, identifying the broken state of the lead by using an image segmentation identification method based on clustering and DS evidence fusion theory;
step 3, establishing an identification model library of the power transmission line in different states;
and 4, rapidly and accurately identifying the power transmission line by using the power transmission line image comparison method based on the AdaBoost algorithm.
The acquired initial image is noisy and must be eliminated using a filtering method. Common filtering methods are gaussian filtering, mean filtering, median filtering, minimum mean square error filtering, Gabor filtering, and the like. By adopting the 3/4 denoising clustering method, not only can noise be effectively removed, but also approximate gray level image points can be clustered. The idea is that the points adjacent to the image are gradually and closely related, noise points are uniformly distributed, and 1 distortion point is considered to be the most important point in 4 adjacent points of a certain point, so that the point can be represented by the mean value of other 3 points with higher quality. 3/4 the denoising and clustering method comprises the following steps:
setting an original image as an mxn order matrix A:
Figure BDA0002835781740000061
in the formula: a isijRepresenting a gray value of a pixel of the image;
and (3) transforming the matrix A by 3/4 values to obtain an image matrix H:
Figure BDA0002835781740000062
order:
f(i,j)＝aij+ai+1,j+ai,j+1+ai+1,j+1
Figure BDA0002835781740000063
Figure BDA0002835781740000064
wherein d is a filtering threshold value, a sensitive area can be defined on the image, and the area is obtained by adopting a maximum variance threshold value method to calculate.
In the step 1, the Canny edge detection method comprises the following steps:
the Canny algorithm smoothes the image by using a two-dimensional gaussian function, which has the following form:
Figure BDA0002835781740000071
the corresponding gradient vector is:
Figure BDA0002835781740000072
two-dimensional gaussian filter functions are generally not directly convolved with an image, but are decomposed into two one-dimensional row and column filters:
Figure BDA0002835781740000073
Figure BDA0002835781740000074
wherein:
Figure BDA0002835781740000075
Figure BDA0002835781740000076
Figure BDA0002835781740000077
Figure BDA0002835781740000078
then, convolution operation is respectively carried out on the images to obtain:
Figure BDA0002835781740000079
Figure BDA00028357817400000710
where x is the convolution sign, k is a constant coefficient, and σ is a spatial scale parameter of the gaussian filter function.
The clustering method can divide input samples into a plurality of clusters, wherein the samples in the same cluster have higher attribute similarity, and the samples in different clusters have larger difference. The similarity is described by the Euclidean distance of the K-means cluster, and the image cluster (K-means) segmentation comprises the following steps:
1) initializing, and determining a cluster number k and a cluster center initial value;
2) calculating Euclidean distance, and calculating each sample Z according to formula (1)iWith each cluster center OjEuclidean distance E ofij，
Figure BDA0002835781740000081
In the formula: zi,mIs the mth attribute of sample Z; m is a sample dimension, namely the number of attributes;
3) dividing clusters, namely dividing each sample into clusters with the minimum Euclidean distance to the sample;
4) updating the clusters, and calculating the average value of each sample in each cluster to be used as a new cluster center;
5) judging convergence, if the moving distance of the cluster center is smaller than a set value or reaches an iteration number limit value, judging the convergence, and ending clustering; otherwise, returning to the step 2 to continue calculation;
the method comprises the following steps of inputting a sample of the clustering algorithm as a pixel point of a power transmission line, compressing an image into H x w pixels before inputting in order to improve the calculation efficiency, selecting a hue value of an image segmentation target as a cluster center initial value, wherein the hue value H is an attribute of the pixel point, and the calculation method is shown as the following formula:
Figure BDA0002835781740000082
r, G and B are the values of the red, green and blue color channels, respectively: t ismax＝max(R,G,B)； Tmin＝min(R,G,B)。
In the step 2, the D-S evidence theory is that a finite set of all possible results is called an identification frame, elements in the finite set are mutually exclusive in pairs, and the identification frame is givenIn the shelf 2nAnd satisfies the following conditions:
Figure BDA0002835781740000083
m(A1)＝1-m(A2)
mi(A1)＝1-mi(A2)＝Pi
m is 2nBasic probability distribution.
In step 4, the AdaBoost algorithm is as follows: the voting process with the best characteristics is that based on a weighted voting mechanism, the answer to a certain question is judged by a large number of weighted combinations of classification functions, the weight of the classifier with the better effect is increased in each calculation iteration, and the weight of the classifier with the relatively poor effect is gradually reduced.

Claims (6)

1. An electric transmission line image identification detection method based on an unmanned aerial vehicle is characterized by comprising the following steps:
step 1, acquiring a conductor image of a power transmission line through an unmanned aerial vehicle image acquisition system, performing image enhancement processing by adopting an 3/4 denoising clustering method, extracting a conductor region in the image, and extracting a conductor region outline by adopting a Canny edge detection method;
step 2, identifying the broken state of the lead by using an image segmentation identification method based on clustering and DS evidence fusion theory;
step 3, establishing an identification model library of the power transmission line in different states;
and 4, rapidly and accurately identifying the power transmission line by using the power transmission line image comparison method based on the AdaBoost algorithm.
2. The unmanned aerial vehicle-based power transmission line image identification and detection method as claimed in claim 1, wherein in the step 1, the 3/4 denoising and clustering method comprises the following steps:
setting an original image as an mxn order matrix A:
Figure FDA0002835781730000011
in the formula: a isijRepresenting a gray value of a pixel of the image;
and (3) transforming the matrix A by 3/4 values to obtain an image matrix H:
Figure FDA0002835781730000012
order:
f(i,j)＝aij+ai+1,j+ai,j+1+ai+1,j+1
Figure FDA0002835781730000013
Figure FDA0002835781730000014
wherein d is a filtering threshold value, a sensitive area can be defined on the image, and the area is obtained by adopting a maximum variance threshold value method to calculate.
3. The unmanned aerial vehicle-based power transmission line image recognition detection method according to claim 1, wherein in the step 1, the Canny edge detection method comprises the following steps:
the Canny algorithm smoothes the image by using a two-dimensional gaussian function, which has the following form:
Figure FDA0002835781730000021
the corresponding gradient vector is:
Figure FDA0002835781730000022
two-dimensional gaussian filter functions are generally not directly convolved with an image, but are decomposed into two one-dimensional row and column filters:
Figure FDA0002835781730000023
Figure FDA0002835781730000024
wherein:
Figure FDA0002835781730000025
Figure FDA0002835781730000026
Figure FDA0002835781730000027
Figure FDA0002835781730000028
then, convolution operation is respectively carried out on the images to obtain:
Figure FDA0002835781730000029
Figure FDA00028357817300000210
where x is the convolution sign, k is a constant coefficient, and σ is a spatial scale parameter of the gaussian filter function.
4. The method according to claim 1, wherein in step 2, the image clustering (K-means) segmentation comprises the following steps:
1) initializing, and determining a cluster number k and a cluster center initial value;
2) calculating Euclidean distance, and calculating each sample Z according to formula (1)iWith each cluster center OjEuclidean distance E ofij，
Figure FDA0002835781730000031
In the formula: zi,mIs the mth attribute of sample Z; m is a sample dimension, namely the number of attributes;
3) dividing clusters, namely dividing each sample into clusters with the minimum Euclidean distance to the sample;
4) updating the clusters, and calculating the average value of each sample in each cluster to be used as a new cluster center;
5) judging convergence, if the moving distance of the cluster center is smaller than a set value or reaches an iteration number limit value, judging the convergence, and ending clustering; otherwise, returning to the step 2 to continue calculation;
the method comprises the following steps of inputting a sample of the clustering algorithm as a pixel point of a power transmission line, compressing an image into H x w pixels before inputting in order to improve the calculation efficiency, selecting a hue value of an image segmentation target as a cluster center initial value, wherein the hue value H is an attribute of the pixel point, and the calculation method is shown as the following formula:
Figure FDA0002835781730000032
r, G and B are the values of the red, green and blue color channels, respectively: t ismax＝max(R,G,B)；Tmin＝min(R,G,B)。
5. The unmanned aerial vehicle-based power transmission line image identification detection method as claimed in claim 1, wherein in step 2, the D-S evidence theory is that a finite set of all possible results is called an identification frame, wherein two elements are mutually exclusive, and 2 is in a given identification framenAnd satisfies the following conditions:
Figure FDA0002835781730000033
m(A1)＝1-m(A2)
mi(A1)＝1-mi(A2)＝Pi
m is 2nBasic probability distribution.
6. The unmanned aerial vehicle-based power transmission line image identification and detection method according to claim 1, wherein in the step 4, the AdaBoost algorithm is as follows: the voting process with the best characteristics is that based on a weighted voting mechanism, the answer to a certain question is judged by a large number of weighted combinations of classification functions, the weight of the classifier with the better effect is increased in each calculation iteration, and the weight of the classifier with the relatively poor effect is gradually reduced.
CN202011476612.XA 2020-12-15 2020-12-15 Power transmission line image recognition detection method based on unmanned aerial vehicle Pending CN112733610A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011476612.XA CN112733610A (en) 2020-12-15 2020-12-15 Power transmission line image recognition detection method based on unmanned aerial vehicle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011476612.XA CN112733610A (en) 2020-12-15 2020-12-15 Power transmission line image recognition detection method based on unmanned aerial vehicle

Publications (1)

Publication Number Publication Date
CN112733610A true CN112733610A (en) 2021-04-30

Family

ID=75602116

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011476612.XA Pending CN112733610A (en) 2020-12-15 2020-12-15 Power transmission line image recognition detection method based on unmanned aerial vehicle

Country Status (1)

Country Link
CN (1) CN112733610A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116823804A (en) * 2023-07-21 2023-09-29 北京化工大学 Knowledge and data combined driving-based power transmission channel safety monitoring method

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116823804A (en) * 2023-07-21 2023-09-29 北京化工大学 Knowledge and data combined driving-based power transmission channel safety monitoring method
CN116823804B (en) * 2023-07-21 2024-02-09 北京化工大学 Knowledge and data combined driving-based power transmission channel safety monitoring method

Similar Documents

Publication Publication Date Title
CN108446630B (en) Intelligent monitoring method for airport runway, application server and computer storage medium
CN111797653B (en) Image labeling method and device based on high-dimensional image
CN108615226B (en) Image defogging method based on generation type countermeasure network
CN107153817B (en) Pedestrian re-identification data labeling method and device
CN109685045B (en) Moving target video tracking method and system
JP6192271B2 (en) Image processing apparatus, image processing method, and program
CN109345547B (en) Traffic lane line detection method and device based on deep learning multitask network
KR20180065889A (en) Method and apparatus for detecting target
CN111639564B (en) Video pedestrian re-identification method based on multi-attention heterogeneous network
CN110930411B (en) Human body segmentation method and system based on depth camera
CN109389609B (en) Interactive self-feedback infrared target detection method based on FART neural network
WO2023082453A1 (en) Image processing method and device
CN111695373B (en) Zebra stripes positioning method, system, medium and equipment
CN108345835B (en) Target identification method based on compound eye imitation perception
CN111597933A (en) Face recognition method and device
CN114581709A (en) Model training, method, apparatus, and medium for recognizing target in medical image
CN115841633A (en) Power tower and power line associated correction power tower and power line detection method
CN104299234B (en) The method and system that rain field removes in video data
CN112733610A (en) Power transmission line image recognition detection method based on unmanned aerial vehicle
CN112115737B (en) Vehicle orientation determining method and device and vehicle-mounted terminal
CN112396582A (en) Mask RCNN-based equalizing ring skew detection method
WO2024016632A1 (en) Bright spot location method, bright spot location apparatus, electronic device and storage medium
CN113129329A (en) Method and device for constructing dense point cloud based on base station target segmentation
CN107341456B (en) Weather sunny and cloudy classification method based on single outdoor color image
CN114924260A (en) Multi-laser radar point cloud registration method

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