CN107729808B - An image intelligent acquisition system and method for unmanned aerial vehicle inspection of transmission lines - Google Patents
An image intelligent acquisition system and method for unmanned aerial vehicle inspection of transmission lines Download PDFInfo
- Publication number
- CN107729808B CN107729808B CN201710805540.0A CN201710805540A CN107729808B CN 107729808 B CN107729808 B CN 107729808B CN 201710805540 A CN201710805540 A CN 201710805540A CN 107729808 B CN107729808 B CN 107729808B
- Authority
- CN
- China
- Prior art keywords
- target
- network
- image
- 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.)
- Active
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
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
- G06V10/14—Optical characteristics of the device performing the acquisition or on the illumination arrangements
- G06V10/147—Details of sensors, e.g. sensor lenses
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30244—Camera pose
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Vascular Medicine (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
本发明公开了一种用于输电线路无人机巡检的图像智能采集系统及方法，当无人机巡检系统沿规划的航迹路线到达悬停点时，对杆塔进行粗定位，实现吊舱或云台的自动调整，将数据采集设备对准杆塔目标，获取一帧杆塔图像信息，对卷积神经网络的区域提取网络与分类网络的损失进行优化改进，利用改进的卷积神经网络算法快速识别杆塔上的附属典型部件；根据识别得到的目标位置，确定目标相对于整幅图像的位置，当目标位置不在图像中心时，计算目标距离中心的像素级偏移量，通过将像素值转化到相机转动量，完成图像中目标的位置调整；判断目标在图像中的面积占比，若占比小于设定值，对相机进行拉焦或者推进无人机距离目标的位置，进行拍摄。
The invention discloses an image intelligent acquisition system and method for unmanned aerial vehicle inspection of transmission lines. When the unmanned aerial vehicle inspection system arrives at a hovering point along a planned track route, rough positioning of the tower is carried out to realize hanging Automatic adjustment of cabin or PTZ, aiming the data acquisition equipment at the tower target, obtaining a frame of tower image information, optimizing and improving the loss of the area extraction network and classification network of the convolutional neural network, using the improved convolutional neural network algorithm Quickly identify the attached typical components on the tower; determine the position of the target relative to the entire image according to the identified target position. When the target position is not in the center of the image, calculate the pixel-level offset of the target from the center, and convert the pixel value by converting the pixel value. To the amount of camera rotation, complete the position adjustment of the target in the image; determine the area ratio of the target in the image, if the ratio is less than the set value, focus the camera or push the drone to the position of the target to shoot.
Description
Technical Field
The invention relates to the field of digital image processing, in particular to an intelligent image acquisition system and method for unmanned aerial vehicle inspection of a power transmission line.
Background
The transmission line is a framework support of a power grid, and with the increase of the transmission line history, especially the rapid development of ultrahigh voltage and extra-high voltage lines, more serious challenges are provided for the daily inspection work of the transmission line. Unmanned aerial vehicle patrols and examines the mode as a novel, efficient and is applied to daily patrolling and examining, but patrols and examines the in-process at unmanned aerial vehicle, needs two staff's cooperation to accomplish jointly and patrols and examines the task usually. One control hand is responsible for controlling the unmanned aerial vehicle platform, and one control hand is responsible for controlling the observation and the collection of airborne nacelle or cloud platform in order to accomplish the target information. In the inspection mode, on one hand, two control hands are required to be tightly matched, the requirement on the operation skill of an operator is high, and the labor intensity is high; on the other hand, the control is affected by communication delay, and the effect of target information acquisition is often not good. In order to reduce labor intensity and improve effectiveness of information acquisition, an intelligent acquisition system and method are urgently needed to achieve automatic acquisition of power transmission line information.
In the existing automatic acquisition system of the power transmission line, the unmanned aerial vehicle inspection tripod head control (ZL201210302421.0) of the power transmission line based on visual servo, which is proposed by the institute of electric power science of Shandong province electric power group, realizes the automatic control of the tripod head in a target matching mode, but can only be applied to the condition of a known target template, and has poor robustness.
The line inspection standardized acquisition method and system (CN201510504379.4) proposed by the aerospace scenic (Beijing) science and technology Limited company realizes fixed-point acquisition aiming at specific equipment by using set airway data and camera parameters, but the acquired transmission line image information is deviated due to the influence of field operation environment (illumination and wind speed) and GPS positioning precision.
Disclosure of Invention
The invention provides an intelligent image acquisition system and method for unmanned aerial vehicle inspection of a power transmission line, aiming at solving the problems.
The invention aims to provide an intelligent image acquisition system for unmanned aerial vehicle inspection of a power transmission line.
The second purpose of the invention is to provide an intelligent image acquisition method for unmanned aerial vehicle inspection of the power transmission line.
In order to achieve the purpose, the invention adopts the following technical scheme:
an intelligent image acquisition method for unmanned aerial vehicle inspection of a power transmission line comprises the following steps:
(1) when the unmanned aerial vehicle inspection system reaches a suspension point along a planned track route, performing coarse positioning on a tower, realizing automatic adjustment of a nacelle or a cloud deck, and aligning data acquisition equipment to a tower target;
(2) acquiring image information of a frame of tower, optimizing and improving the loss of an area extraction network and a classification network of a convolutional neural network, and quickly identifying an attached typical component on the tower by using an improved convolutional neural network algorithm;
(3) determining the position of the target relative to the whole image according to the identified target position, calculating the pixel-level offset of the target from the center when the target position is not in the center of the image, and completing the position adjustment of the target in the image by converting the pixel value into the rotation quantity of the camera;
(4) judging the area occupation ratio of the target in the image, if the occupation ratio is smaller than a set value, zooming the camera or pushing the unmanned aerial vehicle to the target to enable the occupation ratio of the target in the image to meet the set condition, and shooting;
(5) and (5) acquiring image information of a new position, executing the step (2) to the step (4), and continuously acquiring the inspection image.
In the step (1), the tower position coarse positioning specifically comprises the following steps:
(1-1) converting a geodetic coordinate system and a geocentric coordinate system according to GPS coordinates of the unmanned aerial vehicle platform and the tower to obtain position information of the tower relative to the unmanned aerial vehicle platform;
(1-2) when the unmanned aerial vehicle reaches the hovering position, converting a ground center coordinate system and an inertia coordinate system, and calculating to obtain the azimuth and the pitch angle of the airborne pod or the holder;
and (1-3) aligning the visual axis of the airborne nacelle or the holder to the tower through servo control, so that the tower appears in the large visual field range of the electric power line patrol nacelle.
And (1-2) calculating the azimuth angle and the pitch angle of the target relative to the airborne coordinate system according to the position coordinate information of the target in the airborne coordinate system after coordinate conversion.
In the step (2), based on the power transmission line inspection image database, the positions of the tower, the insulator, the vibration damper and the voltage-sharing ring component in the image are marked, category attribute labels are added to corresponding components, the marked power transmission line target component data are used as a training input data set, a VGG16 convolutional neural network front 13-layer network is adopted, the extraction of the power transmission line target image features is achieved, and finally the multi-dimensional high-level semantic features are obtained.
In the step (2), a reference window is constructed, a plurality of region extraction kernels are formed, region acquisition is performed on regions where targets may exist, and a region extraction network is formed to extract candidate positions of the target components.
In the step (2), the softmax algorithm is adopted to classify the extracted target area, the category attribute of the target is determined, and classification of multiple types of targets is realized.
In the step (2), the multidimensional high-level semantic feature data obtained by the feature extraction network is used as input data for training the area extraction network and the classification network.
In the step (2), the feature extraction network, the area extraction network and the classification network are combined and optimized, the feature extraction network is initialized by using a VGG16 convolutional neural network, and multi-dimensional feature data are extracted; respectively inputting the loss function into a regional extraction network and a classification network, optimizing the loss function by using a random gradient descent method, transmitting the loss to a feature extraction network through a feedback transmission network, and adjusting and optimizing internal parameters of the feature extraction network; and after multiple iterations, training the whole network is completed, and a power transmission line component detection model is obtained.
In the step (2), a power transmission line component detection process is completed by fusing a feature extraction network, a regional extraction network and a classification network, multi-dimensional image abstract features are obtained by performing feature extraction operation only once in the detection process, the regional extraction network quickly obtains possible position information of a target, and the classification network is used for classifying the target category.
In the step (2), a feature extraction convolution operation is carried out on the input image by using a feature extraction network to obtain a multi-dimensional feature map; generating a large number of transmission line target candidate region frames on the feature map by using a region extraction network; and taking out the features in the candidate region frames on the feature map and forming a high-dimensional feature vector, calculating the category score of the features of each target region by using a classification network, only keeping the target candidate frames with the scores higher than a set value, carrying out non-maximum value suppression on the target candidate region frames of the power transmission line, and predicting the more appropriate target peripheral frame position.
In the step (4), the servo control of the pan-tilt or the pod is performed based on the eye-in-hand vision servo control system.
In the step (4), the camera and the video camera are simultaneously installed on the cloud deck or the nacelle and are located at coaxial positions, the position conversion of the cloud deck is converted into the change of image parameters through a camera perspective projection matrix, so that the conversion relation between an image characteristic space and a cloud deck tail end position space is obtained, the position and angle information of the electric transmission line target relative to the image center is obtained, the pixel recorded offset is obtained, the rotating angular speed and linear speed of the camera are determined according to the linear mapping relation between the target pixel level offset and the cloud deck rotation quantity, the camera is adjusted through the servo control system to place the target at the center position, and therefore the automatic acquisition process is completed.
The utility model provides an image intelligent acquisition system for transmission line unmanned aerial vehicle patrols and examines, but off-line or on-line operation on processor equipment such as server specifically include:
the rough positioning module is configured to perform rough positioning on the tower when the unmanned aerial vehicle inspection system reaches a suspension point along a planned track route, so that automatic adjustment of a nacelle or a holder is realized, and the data acquisition equipment is aligned to a tower target;
the target identification module is configured to acquire image information of a frame of tower, optimize and improve the loss of an area extraction network and a classification network of the convolutional neural network, and quickly identify an attached typical component on the tower by using an improved convolutional neural network algorithm;
the computing module is configured to determine the position of the target relative to the whole image according to the identified target position, calculate the pixel-level offset of the target from the center when the target position is not in the center of the image, and complete the position adjustment of the target in the image by converting the pixel value into the rotation amount of the camera;
and the servo control module is configured to judge the area occupation ratio of the target in the image, and if the occupation ratio is smaller than a set value, the camera is zoomed in or is pushed close to the position of the unmanned aerial vehicle away from the target, so that the occupation ratio of the target in the image meets a set condition, and shooting is performed.
Compared with the prior art, the invention has the beneficial effects that:
1. the invention realizes the rough positioning of the tower through the rapid positioning technology, then utilizes the target recognition algorithm to determine the position information of the target, and finally realizes the high-definition image information acquisition of the target by converting the target pixel information into the servo control system quantity;
2. according to the invention, the losses of the regional extraction network and the classification network are optimized, the convolution network is further improved, the training of the whole network is completed, and the power transmission line detection model is obtained and used for the actual detection process, so that the method has a good application prospect and can be helpful for completing the unmanned aerial vehicle inspection task of the power transmission line.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the application and, together with the description, serve to explain the application and are not intended to limit the application.
FIG. 1 is a flow chart of a tower position coarse positioning coordinate transformation relationship of the present invention;
FIG. 2 is a block diagram of the area extraction network and classification network of the present invention;
fig. 3 is a flow chart of the intelligent acquisition system for the unmanned aerial vehicle inspection image of the power transmission line.
The specific implementation mode is as follows:
the invention is further described with reference to the following figures and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
As introduced by the background art, in the prior art, the image intelligent acquisition system and method for unmanned aerial vehicle inspection of the power transmission line have the defects of poor robustness and easiness in being influenced by other factors when being applied to the known target template, and in order to solve the technical problems, the technology realizes the rough positioning of a tower through a quick positioning technology, then determines the position information of a target by using a target identification algorithm, and finally realizes the high-definition image information acquisition of the target by converting target pixel information into a servo control system quantity.
As shown in fig. 3, the method specifically includes:
the method comprises the following steps: and roughly positioning the position of the tower. When the unmanned aerial vehicle inspection system reaches the suspension point along the planned track route, the tower coarse positioning system is started to realize the automatic adjustment of the nacelle or the cradle head, and the data acquisition equipment is aligned to the tower target.
Step two: and automatically identifying the transmission line target. Acquiring image information of a frame of tower, and quickly identifying accessory typical components on the tower, such as typical components of insulators, vibration dampers, grading rings and the like, by using an improved convolutional neural network algorithm.
Step three: and calculating the rotation quantity. And automatically identifying and positioning the obtained target position by utilizing the second target, and determining the position of the target relative to the whole image. And if the target position is not at the center of the image, calculating the pixel-level offset of the target from the center, and converting the pixel value into the camera rotation amount to complete the position adjustment of the target in the image.
Step four: and rotating the holder and acquiring the image information at the new position again, and executing the second step and the third step to enable the image target to be in the center of the camera view field.
Step five: and collecting target component information. When the area ratio of the target in the image is less than 50%, zooming the camera or pushing the camera to the position of the unmanned aerial vehicle away from the target to enable the ratio of the target in the image to be not less than 50%; and starting a shutter to finish shooting.
The tower coarse position positioning technology based on the unmanned aerial vehicle inspection platform attitude and GPS information mainly uses inspection platform navigation information and tower GPS data, and according to conversion among several coordinate systems, the guiding azimuth and pitch angle values of the electric power inspection nacelle or the cradle head are obtained through calculation. The tower position coarse positioning in the first step comprises the following specific steps:
A. and converting a geodetic coordinate system and a geocentric coordinate system according to the GPS coordinates of the unmanned aerial vehicle platform and the tower to obtain the position information of the tower relative to the unmanned aerial vehicle platform.
B. When the unmanned aerial vehicle reaches the hovering position, the azimuth and the pitch angle of the airborne pod or the holder are calculated through the conversion of a ground center coordinate system and an inertia coordinate system, the distance D, the azimuth α and the pitch angle β of the target relative to the airborne coordinate system are calculated, and the formula is as follows:
and x, y and z are positions of the target in the airborne coordinate system after coordinate conversion.
C. Through servo control, the visual axis of the airborne nacelle or the cloud deck is aligned with the tower, so that the tower is in the large visual field range of the electric power line patrol nacelle.
The specific steps of target identification in the third step are as follows:
the identification and positioning of the component is accomplished based on a modified convolutional neural network (fast region conditional neural network).
A. Constructing power transmission line image feature extraction network
1. Based on the power transmission line inspection image database, the positions of components such as a tower, an insulator, a vibration damper, a grading ring and the like in the image are marked manually, and category attribute labels are added to the corresponding components.
2. The method comprises the steps of taking calibrated target component data (target area images and category labels) of the power transmission line as a training input data set, adopting a front 13-layer network in a VGG16 convolutional network, mainly comprising convolutional layers and pooling layers, realizing extraction of target image features of the power transmission line by optimizing parameters of each layer, and finally obtaining 512-dimensional high-level semantic features.
B. Constructing regional extraction networks and classification networks
1. The region extraction network is used for extracting candidate positions of the target component, defining three proportions and three scales of reference windows, forming nine region extraction kernels, and performing region acquisition on regions where targets may exist.
2. The classification network is used for classifying the extracted target area and determining the class attribute of the target. And the classification network adopts a softmax algorithm to realize the classification of various targets.
3. The area extraction network and the classification network are both connected behind the feature extraction network, and in order to obtain a better network, the losses of the area extraction network and the classification network need to be optimized, and the loss function is as follows:
wherein i represents the ith network region extraction core; λ represents a balance parameter; n is a radical ofclsRepresents the total category number; n is a radical ofregRepresenting the number of extracted target frames; p is a radical ofiRepresenting the prediction probability for the category of the ith target region,
representing the actual type probability of the object. t is tiInformation of four coordinates representing the extracted target frame,
representing actual coordinate information of the object. L isclsLog loss, L, for a classregRepresenting the loss of fit of the extraction box to the actual target box.
512-dimensional feature data obtained by the feature extraction network is used as input data for training the region extraction network and the classification network, and the combination relationship and the training process of the two networks are shown in fig. 2.
Performing combined tuning on the feature extraction network, the regional extraction network and the classification network, initializing the feature extraction network by using a VGG16 network, and extracting 512-dimensional feature data; respectively inputting the loss function into a regional extraction network and a classification network, optimizing the loss function by using a random gradient descent method, transmitting the loss to a feature extraction network through a feedback transmission network, and adjusting and optimizing internal parameters of the feature extraction network; and after multiple iterations, training the whole network is completed, and a power transmission line detection model is obtained and used for an actual detection process.
C. Object detection network
And completing the detection process of the power transmission line component by fusing the feature extraction network, the area extraction network and the classification network. In the detection process, feature extraction operation is performed only once to obtain 512-dimensional image abstract features, the regional extraction network quickly obtains possible position information of the target, and the classification network is used for classifying the target category, wherein the specific flow is as follows:
1. performing feature extraction convolution operation on the input image by using a feature extraction network to obtain a 512-dimensional feature map;
2. generating a large number of transmission line target candidate region frames on the feature map by using a region extraction network;
3. extracting features in the candidate region frames on the feature map and forming high-dimensional feature vectors, calculating the category score of the features of each target region by a classification network, and only keeping the first 300 target candidate frames with higher scores;
4. and carrying out non-maximum suppression on the target candidate area frame of the power transmission line, and predicting a more appropriate target peripheral frame position.
The vision servo system can be classified into an Eye-in-Hand system (Eye in Hand) and an Eye-to-Hand system (Eye to Hand) according to the installation position of the vision sensor. The eye-in-hand system is used for installing the sensor on the robot end effector, so that the vision range is large, the movement is flexible, and high control precision can be obtained. When the visual servo is carried out, the position and the orientation of the object are calculated in real time by using an image visual means, and the rotation angle and the rotation quantity of the nacelle or the holder are determined according to the information of the orientation and the position of the object. The servo control process of step four is image-based eye-on-hand vision servo control.
The camera and the video camera are simultaneously arranged on the cloud platform or the nacelle and are in coaxial positions. The visual servo is carried out by adopting an Eye-in-Hand mode, the holder has m degrees of freedom, and the angular speed of the rotation of the holder is q ═ q1,L,qp]The linear velocity of the tip is r ═ r1,L,rm]The two have the following relationship: r is JrQ, wherein
And converting the position transformation of the holder into the change of image parameters through a camera perspective projection matrix, thereby obtaining the transformation relation between the image characteristic space and the holder tail end position space. Assuming image coordinatesTo (x, y), the final transformation relation is:
wherein
f is the camera focal length. u ═ vx,νy,νz,ωx,ωy,ωz]TAnd ν and ω indicate linear and angular velocities of rotation of the camera. J. the design is a squareImageThe image feature change rate is the Jacobian matrix of the image, and the relation between the image feature change rate and the holder speed is represented by a composite Jacobian matrix J ═ JImage·JrAnd (4) showing.
According to the position and angle information of the transmission line target relative to the image center obtained in the third step, the pixel remembering offset (P) is obtainedx,Py) Because the pan-tilt control adopts a Look-then-move mode, the linear mapping relation between the target pixel level offset and the pan-tilt rotation amount is defined as follows:
(Px,Py)＝JImage*u
the Jacobian matrix of the image and the angular velocity and the linear velocity of the rotation of the camera can be solved by selecting a conversion equation. And adjusting the camera through the servo control system to place the target at the center position of the view field, thereby completing the automatic acquisition process.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.
Claims (12)
1. An intelligent image acquisition method for unmanned aerial vehicle inspection of a power transmission line is characterized by comprising the following steps: the method comprises the following steps:
(1) when the unmanned aerial vehicle inspection system reaches a suspension point along a planned track route, performing coarse positioning on a tower, realizing automatic adjustment of a nacelle or a cloud deck, and aligning data acquisition equipment to a tower target;
(2) acquiring image information of a frame of tower, optimizing and improving the loss of an area extraction network and a classification network of a convolutional neural network, and quickly identifying an attached typical component on the tower by using an improved convolutional neural network algorithm;
(3) determining the position of the target relative to the whole image according to the identified target position, calculating the pixel-level offset of the target from the center when the target position is not in the center of the image, and completing the position adjustment of the target in the image by converting the pixel value into the rotation quantity of the camera;
(4) judging the area occupation ratio of the target in the image, if the occupation ratio is smaller than a set value, zooming the camera or propelling the unmanned aerial vehicle to the position away from the target to enable the occupation ratio of the target in the image to meet the set condition, and shooting;
(5) acquiring image information of a new position, executing the step (2) to the step (4), and continuously acquiring the inspection image;
in the step (2), the feature extraction network, the area extraction network and the classification network are combined and optimized, the feature extraction network is initialized by using a convolution network, and multi-dimensional feature data are extracted; respectively inputting the loss function into a regional extraction network and a classification network, optimizing the loss function by using a random gradient descent method, transmitting the loss to a feature extraction network through a feedback transmission network, and adjusting and optimizing internal parameters of the feature extraction network; and after multiple iterations, training the whole network is completed, and a power transmission line detection model is obtained.
2. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (1), the tower position coarse positioning specifically comprises the following steps:
(1-1) converting a geodetic coordinate system and a geocentric coordinate system according to GPS coordinates of the unmanned aerial vehicle platform and the tower to obtain position information of the tower relative to the unmanned aerial vehicle platform;
(1-2) when the unmanned aerial vehicle reaches the hovering position, converting a ground center coordinate system and an inertia coordinate system, and calculating to obtain the azimuth and the pitch angle of the airborne nacelle or the cloud deck and the distance of the target relative to the airborne coordinate system;
and (1-3) aligning the visual axis of the airborne nacelle or the holder to the tower through servo control, so that the tower appears in the large visual field range of the electric power line patrol nacelle.
3. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 2, characterized in that: and (1-2) calculating the azimuth angle and the pitch angle of the target relative to the airborne coordinate system according to the position coordinate information of the target in the airborne coordinate system after coordinate conversion.
4. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (2), based on the power transmission line inspection image database, the positions of the tower, the insulator, the vibration damper and the voltage-sharing ring component in the image are marked, category attribute labels are added to corresponding components, the marked power transmission line target component data are used as a training input data set, a front 13-layer network in a VGG16 convolutional network is adopted, the extraction of the power transmission line target image features is achieved, and finally the multi-dimensional high-level semantic features are obtained.
5. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (2), a reference window is constructed, a plurality of region extraction kernels are formed, region acquisition is performed on regions where targets may exist, and a region extraction network is formed to extract candidate positions of the target components.
6. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (2), the softmax algorithm is adopted to classify the extracted target area, the category attribute of the target is determined, and classification of multiple types of targets is realized.
7. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (2), the multidimensional high-level semantic feature data obtained by the feature extraction network is used as input data for training the area extraction network and the classification network.
8. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (2), a power transmission line component detection process is completed by fusing a feature extraction network, a regional extraction network and a classification network, multi-dimensional image abstract features are obtained by performing feature extraction operation only once in the detection process, the regional extraction network quickly obtains possible position information of a target, and the classification network is used for classifying the target category.
9. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (2), a feature extraction convolution operation is carried out on the input image by using a feature extraction network to obtain a multi-dimensional feature map; generating a large number of transmission line target candidate region frames on the feature map by using a region extraction network; and taking out the features in the candidate region frames on the feature map and forming a high-dimensional feature vector, calculating the category score of the features of each target region by using a classification network, only keeping the target candidate frames with the scores higher than a set value, carrying out non-maximum value suppression on the target candidate region frames of the power transmission line, and predicting the more appropriate target peripheral frame position.
10. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (4), the servo control of the pan-tilt or the pod is performed based on the eye-in-hand vision servo control system.
11. The intelligent image acquisition method for the unmanned aerial vehicle inspection of the power transmission line according to claim 1, which is characterized in that: in the step (4), the camera and the video camera are simultaneously installed on the cloud deck or the nacelle and are located at coaxial positions, the position conversion of the cloud deck is converted into the change of image parameters through a camera perspective projection matrix, so that the conversion relation between an image characteristic space and a cloud deck tail end position space is obtained, the position and angle information of the electric transmission line target relative to the image center is obtained, the pixel recorded offset is obtained, the rotating angular speed and linear speed of the camera are determined according to the linear mapping relation between the target pixel level offset and the cloud deck rotation quantity, the camera is adjusted through the servo control system to place the target at the center position, and therefore the automatic acquisition process is completed.
12. The utility model provides an image intelligent acquisition system for transmission line unmanned aerial vehicle patrols and examines, characterized by: the method specifically comprises the following steps:
the rough positioning module is configured to perform rough positioning on the tower when the unmanned aerial vehicle inspection system reaches a suspension point along a planned track route, so that automatic adjustment of a nacelle or a holder is realized, and the data acquisition equipment is aligned to a tower target;
the target identification module is configured to acquire image information of a frame of tower, optimize and improve the loss of an area extraction network and a classification network of the convolutional neural network, and quickly identify an attached typical component on the tower by using an improved convolutional neural network algorithm;
the computing module is configured to determine the position of the target relative to the whole image according to the identified target position, calculate the pixel-level offset of the target from the center when the target position is not in the center of the image, and complete the position adjustment of the target in the image by converting the pixel value into the rotation amount of the camera;
the servo control module is configured to judge the area occupation ratio of the target in the image, and if the occupation ratio is smaller than a set value, the camera is zoomed in or the position of the unmanned aerial vehicle away from the target is pushed forward, so that the occupation ratio of the target in the image meets a set condition, and shooting is carried out;
optimizing and improving the loss of the area extraction network and the classification network of the convolutional neural network, wherein the optimizing and improving the loss of the area extraction network and the classification network of the convolutional neural network comprises the steps of carrying out joint tuning on the feature extraction network, the area extraction network and the classification network, initializing the feature extraction network by using the convolutional network, and extracting multi-dimensional feature data; respectively inputting the loss function into a regional extraction network and a classification network, optimizing the loss function by using a random gradient descent method, transmitting the loss to a feature extraction network through a feedback transmission network, and adjusting and optimizing internal parameters of the feature extraction network; and after multiple iterations, training the whole network is completed, and a power transmission line detection model is obtained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710805540.0A CN107729808B (en) | 2017-09-08 | 2017-09-08 | An image intelligent acquisition system and method for unmanned aerial vehicle inspection of transmission lines |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710805540.0A CN107729808B (en) | 2017-09-08 | 2017-09-08 | An image intelligent acquisition system and method for unmanned aerial vehicle inspection of transmission lines |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107729808A CN107729808A (en) | 2018-02-23 |
| CN107729808B true CN107729808B (en) | 2020-05-01 |
Family
ID=61205825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710805540.0A Active CN107729808B (en) | 2017-09-08 | 2017-09-08 | An image intelligent acquisition system and method for unmanned aerial vehicle inspection of transmission lines |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107729808B (en) |
Families Citing this family (80)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108416299A (en) * | 2018-03-09 | 2018-08-17 | 重庆览辉信息技术有限公司 | A kind of power circuit intelligent inspection system |
| CN108956640B (en) * | 2018-04-04 | 2020-12-18 | 国网智能科技股份有限公司 | Vehicle-mounted detection device and detection method suitable for power distribution line inspection |
| CN108961276B (en) * | 2018-04-04 | 2020-09-25 | 国网智能科技股份有限公司 | Method and system for automatic collection of inspection data of distribution lines based on visual servoing |
| CN108871409A (en) * | 2018-05-02 | 2018-11-23 | 北京畅东科技有限公司 | A kind of fault detection method and system |
| CN110543800B (en) * | 2018-05-29 | 2023-08-08 | 北京京东尚科信息技术有限公司 | Target recognition tracking method and device for pod and pod |
| CN110557604A (en) * | 2018-05-31 | 2019-12-10 | 北京星闪世图科技有限公司 | unmanned aerial vehicle image full-automatic shooting method device for intelligent inspection of electric power facilities |
| CN108829966B (en) * | 2018-06-07 | 2023-03-21 | 国网上海市电力公司 | Enhanced grid determination method for power line patrol system |
| CN108803668B (en) * | 2018-06-22 | 2021-08-24 | 中国南方电网有限责任公司超高压输电公司广州局 | Intelligent inspection unmanned aerial vehicle nacelle system for static target monitoring |
| CN108921218B (en) * | 2018-06-29 | 2022-06-24 | 炬大科技有限公司 | Target object detection method and device |
| CN109032175B (en) * | 2018-07-13 | 2021-08-06 | 辽宁东科电力有限公司 | Unmanned aerial vehicle power inspection path calculation method based on Archimedes spiral |
| CN108955645A (en) * | 2018-07-16 | 2018-12-07 | 福州日兆信息科技有限公司 | Three-dimensional modeling method and device applied to communication iron tower intelligent patrol detection |
| CN109003029A (en) * | 2018-07-20 | 2018-12-14 | 福州大学 | A kind of intelligent repository method for inspecting and system based on deep learning |
| CN108646786A (en) * | 2018-07-24 | 2018-10-12 | 上海伯镭智能科技有限公司 | A kind of mechanical equipment cruising inspection system and its method based on multiaxis unmanned plane |
| CN108919838B (en) * | 2018-08-27 | 2021-03-26 | 河海大学常州校区 | An automatic tracking method for unmanned aerial vehicle transmission line based on binocular vision |
| CN109116865A (en) * | 2018-09-19 | 2019-01-01 | 苏州傲特欣智能科技有限公司 | Large scale equipment unmanned plane cruising inspection system and its method based on machine vision |
| CN109459437A (en) * | 2018-11-07 | 2019-03-12 | 天津市普迅电力信息技术有限公司 | Multi-rotor unmanned aerial vehicle transmission tower defect identification method based on high accuracy positioning |
| CN109544544A (en) * | 2018-11-30 | 2019-03-29 | 长讯通信服务有限公司 | It is a kind of that Obj State recognition methods is safeguarded based on deep learning and the mobile communication of unmanned plane |
| CN109636848B (en) * | 2018-12-17 | 2020-12-25 | 武汉天乾科技有限责任公司 | Unmanned aerial vehicle-based oil and gas pipeline inspection method |
| CN109669474B (en) * | 2018-12-21 | 2022-02-15 | 国网安徽省电力有限公司淮南供电公司 | Adaptive hovering position optimization algorithm for multi-rotor UAV based on prior knowledge |
| CN109765939A (en) * | 2018-12-21 | 2019-05-17 | 中国科学院自动化研究所南京人工智能芯片创新研究院 | Cloud platform control method, device and the storage medium of unmanned plane |
| CN109754420B (en) | 2018-12-24 | 2021-11-12 | 深圳市道通智能航空技术股份有限公司 | Target distance estimation method and device and unmanned aerial vehicle |
| CN109961003A (en) * | 2018-12-26 | 2019-07-02 | 安徽继远软件有限公司 | An Airborne Auxiliary Patrol Device for Embedded Transmission Lines Based on FPGA |
| CN109976339B (en) * | 2019-03-15 | 2021-11-30 | 国网智能科技股份有限公司 | Vehicle-mounted distribution network inspection data acquisition method and inspection system |
| CN109933096B (en) * | 2019-03-15 | 2021-11-30 | 国网智能科技股份有限公司 | Cloud deck servo control method and system |
| CN109885091B (en) * | 2019-03-21 | 2021-08-24 | 华北电力大学（保定） | A method and system for autonomous flight control of an unmanned aerial vehicle |
| WO2020191668A1 (en) | 2019-03-27 | 2020-10-01 | Hangzhou Fabu Technology Co. Ltd | Proposal processing method and related products |
| CN110113570A (en) * | 2019-05-07 | 2019-08-09 | 江苏方天电力技术有限公司 | A kind of autonomous cruising inspection system of power transmission line unmanned machine and its working method |
| CN110351478A (en) * | 2019-05-21 | 2019-10-18 | 江苏看见云软件科技有限公司 | A kind of unmanned plane acquiring and transmission system handling violation information |
| CN110385324B (en) * | 2019-06-03 | 2021-07-27 | 浙江大华技术股份有限公司 | Cleaning method, system, readable storage medium and equipment for camera device |
| CN110142785A (en) * | 2019-06-25 | 2019-08-20 | 山东沐点智能科技有限公司 | A kind of crusing robot visual servo method based on target detection |
| CN110516551B (en) * | 2019-07-29 | 2023-04-07 | 上海交通大学烟台信息技术研究院 | Vision-based line patrol position deviation identification system and method and unmanned aerial vehicle |
| CN110580717B (en) * | 2019-08-15 | 2022-10-21 | 成都优艾维智能科技有限责任公司 | A method for generating autonomous inspection routes of unmanned aerial vehicles for power towers |
| CN110648319B (en) * | 2019-09-19 | 2022-03-18 | 国网智能科技股份有限公司 | Equipment image acquisition and diagnosis system and method based on double cameras |
| CN110618424B (en) * | 2019-09-27 | 2021-09-21 | 中科九度(北京)空间信息技术有限责任公司 | Remote high-voltage line discovery method based on multi-sensor fusion |
| CN110738179A (en) * | 2019-10-18 | 2020-01-31 | 国家电网有限公司 | A kind of identification method and related device of electric equipment |
| CN111178148B (en) * | 2019-12-06 | 2023-06-02 | 天津大学 | A ground target geographic coordinate positioning method based on UAV vision system |
| CN111047703B (en) * | 2019-12-23 | 2023-09-26 | 杭州电力设备制造有限公司 | User high-voltage distribution equipment identification and space reconstruction method |
| CN110986871A (en) * | 2019-12-24 | 2020-04-10 | 广西大学 | An Image Measurement Method of Crossover Line Spacing Based on RTK Precise Positioning |
| WO2021168707A1 (en) * | 2020-02-26 | 2021-09-02 | 深圳市大疆创新科技有限公司 | Focusing method, apparatus and device |
| CN111275015A (en) * | 2020-02-28 | 2020-06-12 | 广东电网有限责任公司 | Unmanned aerial vehicle-based power line inspection electric tower detection and identification method and system |
| CN112585554B (en) * | 2020-03-27 | 2025-03-04 | 深圳市大疆创新科技有限公司 | UAV inspection method, device and UAV |
| CN112198891B (en) * | 2020-04-22 | 2021-12-07 | 北京理工大学 | Multi-gyroplane autonomous recovery method |
| CN113645397A (en) * | 2020-04-27 | 2021-11-12 | 杭州海康机器人技术有限公司 | Tracking method, device and system for moving target object |
| CN113687601B (en) * | 2020-05-18 | 2022-11-29 | 联华电子股份有限公司 | Automatic adjustment method of equipment and intelligent adjustment device applying same |
| CN111586303A (en) * | 2020-05-22 | 2020-08-25 | 浩鲸云计算科技股份有限公司 | Control method and device for dynamically tracking road surface target by camera based on wireless positioning technology |
| CN111611989B (en) * | 2020-05-22 | 2023-08-01 | 四川智动木牛智能科技有限公司 | Multi-target accurate positioning identification method based on autonomous robot |
| CN111741216B (en) * | 2020-06-17 | 2021-08-03 | 中国人民解放军陆军工程大学 | A method for capturing images from unmanned aerial vehicles |
| CN111738189A (en) * | 2020-06-29 | 2020-10-02 | 广东电网有限责任公司 | A method, device, terminal and medium for inspection and control of crimping fittings of power transmission lines |
| CN111768378B (en) * | 2020-06-29 | 2025-01-14 | 广东电网有限责任公司 | A method and device for detecting live voltage of crimping hardware |
| CN112087598B (en) * | 2020-08-03 | 2023-04-07 | 精英数智科技股份有限公司 | Intelligent tripod head inspection method and device and tripod head camera inspection system |
| CN111985448A (en) * | 2020-09-02 | 2020-11-24 | 深圳壹账通智能科技有限公司 | Vehicle image recognition method and device, computer equipment and readable storage medium |
| CN112163055B (en) * | 2020-09-09 | 2023-11-21 | 成都深瑞同华科技有限公司 | A method for labeling poles and towers |
| CN112232132A (en) * | 2020-09-18 | 2021-01-15 | 北京理工大学 | A Target Recognition and Positioning Method Based on Navigation Information |
| CN112229845A (en) * | 2020-10-12 | 2021-01-15 | 国网河南省电力公司濮阳供电公司 | High-precision Rao Tower intelligent inspection method for UAV based on visual navigation technology |
| CN112330915B (en) * | 2020-10-29 | 2023-02-28 | 五邑大学 | Unmanned aerial vehicle forest fire prevention early warning method and system, electronic equipment and storage medium |
| CN112102395B (en) * | 2020-11-09 | 2022-05-20 | 广东科凯达智能机器人有限公司 | Autonomous inspection method based on machine vision |
| CN112367400B (en) * | 2020-11-12 | 2022-04-29 | 广东电网有限责任公司 | Intelligent inspection method and system for power internet of things with edge cloud coordination |
| CN112650285B (en) * | 2020-12-17 | 2025-02-07 | 内蒙古京隆发电有限责任公司 | Joint inspection method and system |
| CN112837273A (en) * | 2021-01-12 | 2021-05-25 | 云南电网有限责任公司电力科学研究院 | An image verification method, system, computer equipment and storage medium |
| CN113034501B (en) * | 2021-05-25 | 2021-10-15 | 国网江苏省电力有限公司泰州供电分公司 | A method and device for identifying the tilt fault of the pressure equalizing ring based on the detection of key points |
| CN113298035B (en) * | 2021-06-17 | 2024-12-03 | 上海红檀智能科技有限公司 | UAV power tower detection and autonomous cruising method based on image recognition |
| CN113626900B (en) * | 2021-08-05 | 2023-06-20 | 复旦大学 | Method and system for generating facility asset icon of road comprehensive pole system |
| CN113534845B (en) * | 2021-08-18 | 2022-06-21 | 国网湖南省电力有限公司 | Autonomous inspection method and system of UAV for overhead distribution line based on GNSS positioning |
| CN113408510B (en) * | 2021-08-23 | 2021-12-24 | 中科方寸知微(南京)科技有限公司 | Method and system for target deviation correction of transmission line based on deep learning and one-hot coding |
| CN113419564B (en) * | 2021-08-24 | 2021-12-03 | 天津市普迅电力信息技术有限公司 | Power channel inspection method based on fuzzy path |
| CN114281100B (en) * | 2021-12-03 | 2023-09-05 | 国网智能科技股份有限公司 | Unmanned aerial vehicle inspection system and method without hovering |
| CN114326771A (en) * | 2021-12-31 | 2022-04-12 | 国网湖北省电力有限公司超高压公司 | Unmanned aerial vehicle shooting route generation method and system based on image recognition |
| CN114492860B (en) * | 2021-12-31 | 2025-02-18 | 山东信通电子股份有限公司 | A method and device for automatic inspection of high voltage transmission lines |
| CN114462646B (en) * | 2022-03-15 | 2022-11-15 | 成都中轨轨道设备有限公司 | Pole number plate identification method and system based on contact network safety inspection |
| CN114900609B (en) * | 2022-04-29 | 2023-03-24 | 北京数字绿土科技股份有限公司 | Automatic shooting control method and system for unmanned aerial vehicle |
| CN115314609B (en) * | 2022-06-21 | 2023-11-28 | 中南大学 | An automated collection method and device for aluminum electrolytic cell fire eye video |
| CN115097836A (en) * | 2022-06-30 | 2022-09-23 | 国电南瑞科技股份有限公司 | Transmission line inspection method, system and storage medium based on image registration |
| CN115239997B (en) * | 2022-07-19 | 2023-06-09 | 哈尔滨市爱硕电力工程有限公司 | High and cold region transmission line tower state monitoring system based on 5G Internet of things |
| CN115825957B (en) * | 2022-11-23 | 2024-12-17 | 国网智能科技股份有限公司 | Unmanned aerial vehicle and method applied to power inspection and power inspection system |
| CN115578662A (en) * | 2022-11-23 | 2023-01-06 | 国网智能科技股份有限公司 | Unmanned aerial vehicle front-end image processing method, system, storage medium and equipment |
| CN115995043B (en) * | 2023-02-16 | 2024-12-03 | 深圳金三立视频科技股份有限公司 | Transmission line hidden danger target identification method and computer readable storage medium |
| CN116202489A (en) * | 2023-03-27 | 2023-06-02 | 国网浙江省电力有限公司电力科学研究院 | Method and system for co-locating power transmission line inspection machine and pole tower and storage medium |
| CN116907511B (en) * | 2023-09-12 | 2023-12-05 | 北京宝隆泓瑞科技有限公司 | Method for converting pipeline coordinates into image coordinates |
| CN117437360A (en) * | 2023-11-03 | 2024-01-23 | 石河子大学 | A fast and high-precision reconstruction method of power lines based on multi-source data fusion |
| CN119002522B (en) * | 2024-10-24 | 2024-12-13 | 北京尚优力达科技有限公司 | Multi-UAV trajectory planning method and device applicable to large-area power transmission lines |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104538899A (en) * | 2015-01-19 | 2015-04-22 | 中兴长天信息技术(北京)有限公司 | Wireless-transmission-based unmanned aerial vehicle platform for power line inspection |
| CN104978580A (en) * | 2015-06-15 | 2015-10-14 | 国网山东省电力公司电力科学研究院 | Insulator identification method for unmanned aerial vehicle polling electric transmission line |
| CN106092054A (en) * | 2016-05-30 | 2016-11-09 | 广东能飞航空科技发展有限公司 | A kind of power circuit identification precise positioning air navigation aid |
| KR20170086992A (en) * | 2016-01-19 | 2017-07-27 | 광주대학교산학협력단 | The System for obtaining three-dimensional spatial image using unmanned aerial vehicle |
| KR20170091352A (en) * | 2016-02-01 | 2017-08-09 | 주식회사에스에이티 | Method for detecting working area and performing continuous working in the detected working area and the unmanned air vehicle performing the same |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110137547A1 (en) * | 2009-12-03 | 2011-06-09 | Electronics And Telecommunications Research Institute | System and method for generating spatial information |
| US20140316614A1 (en) * | 2012-12-17 | 2014-10-23 | David L. Newman | Drone for collecting images and system for categorizing image data |
-
2017
- 2017-09-08 CN CN201710805540.0A patent/CN107729808B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104538899A (en) * | 2015-01-19 | 2015-04-22 | 中兴长天信息技术(北京)有限公司 | Wireless-transmission-based unmanned aerial vehicle platform for power line inspection |
| CN104978580A (en) * | 2015-06-15 | 2015-10-14 | 国网山东省电力公司电力科学研究院 | Insulator identification method for unmanned aerial vehicle polling electric transmission line |
| KR20170086992A (en) * | 2016-01-19 | 2017-07-27 | 광주대학교산학협력단 | The System for obtaining three-dimensional spatial image using unmanned aerial vehicle |
| KR20170091352A (en) * | 2016-02-01 | 2017-08-09 | 주식회사에스에이티 | Method for detecting working area and performing continuous working in the detected working area and the unmanned air vehicle performing the same |
| CN106092054A (en) * | 2016-05-30 | 2016-11-09 | 广东能飞航空科技发展有限公司 | A kind of power circuit identification precise positioning air navigation aid |
Non-Patent Citations (1)
| Title |
|---|
| "基于无人机的输电线路典型设备图像信息采集方法研究";冯智慧;《电瓷避雷器》;20160831(第4期);第22-26页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107729808A (en) | 2018-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107729808B (en) | An image intelligent acquisition system and method for unmanned aerial vehicle inspection of transmission lines | |
| CN113298035B (en) | UAV power tower detection and autonomous cruising method based on image recognition | |
| CN108803668B (en) | Intelligent inspection unmanned aerial vehicle nacelle system for static target monitoring | |
| CN109857144B (en) | Unmanned aerial vehicle, unmanned aerial vehicle control system and control method | |
| CN110889324A (en) | Thermal infrared image target identification method based on YOLO V3 terminal-oriented guidance | |
| CN108229587B (en) | Autonomous transmission tower scanning method based on hovering state of aircraft | |
| CN113408510B (en) | Method and system for target deviation correction of transmission line based on deep learning and one-hot coding | |
| CN115144879B (en) | Multi-machine multi-target dynamic positioning system and method | |
| CN112066977B (en) | Multi-target matching and cataloging method for photoelectric measurement network | |
| CN115100554A (en) | A UAV power inspection system based on intelligent vision and its detection method | |
| CN109669474B (en) | Adaptive hovering position optimization algorithm for multi-rotor UAV based on prior knowledge | |
| CN109976339B (en) | Vehicle-mounted distribution network inspection data acquisition method and inspection system | |
| CN116894936B (en) | Marine target recognition and positioning method and system based on UAV vision | |
| CN112862862A (en) | Airplane autonomous oil receiving device based on artificial intelligence visual tracking and application method | |
| CN116202489A (en) | Method and system for co-locating power transmission line inspection machine and pole tower and storage medium | |
| CN112947526A (en) | Unmanned aerial vehicle autonomous landing method and system | |
| CN110132049A (en) | A self-aiming sniper rifle based on an unmanned aerial vehicle platform | |
| Lin et al. | A multi-target detection framework for multirotor UAV | |
| CN112198884A (en) | Unmanned aerial vehicle mobile platform landing method based on visual guidance | |
| CN115471555A (en) | A Pose Determination Method for UAV Infrared Inspection Based on Image Feature Point Matching | |
| CN116243725A (en) | Substation drone inspection method and system based on visual navigation | |
| CN112008730B (en) | Sun position identification and avoidance method based on solid-state area array laser radar | |
| Li et al. | Prediction of wheat gains with imagery from four-rotor UAV | |
| CN106291169A (en) | High pressure overhead power line line walking UAV system photoelectric detection system | |
| CN119861739B (en) | Autonomous inspection system of underwater robot based on binocular vision |
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 | ||
| CB02 | Change of applicant information |
Address after: Ji'nan City, Shandong Province Wang Yue Road 250003 No. 2000 Applicant after: Electric Power Research Institute of State Grid Shandong Electric Power Company Applicant after: National Network Intelligent Technology Co., Ltd. Applicant after: State Grid Corporation of China Address before: Ji'nan City, Shandong Province Wang Yue Road 250003 No. 2000 Applicant before: Electric Power Research Institute of State Grid Shandong Electric Power Company Applicant before: Shandong Luneng Intelligent Technology Co., Ltd. Applicant before: State Grid Corporation of China |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20201028 Address after: 250101 Electric Power Intelligent Robot Production Project 101 in Jinan City, Shandong Province, South of Feiyue Avenue and East of No. 26 Road (ICT Industrial Park) Patentee after: National Network Intelligent Technology Co.,Ltd. Address before: Ji'nan City, Shandong Province Wang Yue Road 250003 No. 2000 Patentee before: ELECTRIC POWER RESEARCH INSTITUTE OF STATE GRID SHANDONG ELECTRIC POWER Co. Patentee before: National Network Intelligent Technology Co.,Ltd. Patentee before: STATE GRID CORPORATION OF CHINA |
|
| TR01 | Transfer of patent right |