CN113033446A - Transmission tower identification and positioning method based on high-resolution remote sensing image - Google Patents
Transmission tower identification and positioning method based on high-resolution remote sensing image Download PDFInfo
- Publication number
- CN113033446A CN113033446A CN202110355197.0A CN202110355197A CN113033446A CN 113033446 A CN113033446 A CN 113033446A CN 202110355197 A CN202110355197 A CN 202110355197A CN 113033446 A CN113033446 A CN 113033446A
- Authority
- CN
- China
- Prior art keywords
- transmission tower
- remote sensing
- positioning
- sensing image
- tower
- 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.)
- Granted
Links
- 230000005540 biological transmission Effects 0.000 title claims abstract description 73
- 238000000034 method Methods 0.000 title claims abstract description 22
- 238000001514 detection method Methods 0.000 claims abstract description 55
- 238000012360 testing method Methods 0.000 claims description 22
- 238000012549 training Methods 0.000 claims description 20
- 230000006870 function Effects 0.000 claims description 10
- 238000012795 verification Methods 0.000 claims description 10
- 230000004927 fusion Effects 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 7
- 238000013434 data augmentation Methods 0.000 claims description 5
- 230000001629 suppression Effects 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000012935 Averaging Methods 0.000 claims description 2
- 238000003384 imaging method Methods 0.000 claims description 2
- 238000012986 modification Methods 0.000 claims description 2
- 230000004048 modification Effects 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 abstract description 4
- 206010063385 Intellectualisation Diseases 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000007689 inspection Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 238000005452 bending Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Multimedia (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method for identifying and positioning a transmission tower based on a high-resolution remote sensing image, which can realize the detection of the transmission tower in a large range based on the high-resolution remote sensing image. Based on the improved YOLOv3 target rapid detection algorithm applied to the detection and positioning of the transmission tower of the high-resolution remote sensing image, aiming at the fact that the transmission tower presents different shapes in the remote sensing image, the size of a prior frame is reset through K-means, a CIoU loss function is introduced into frame regression, DIoU and NMS are combined, and the problem of missed detection of YOLOv3 on the small dense targets is solved. Aiming at the problem that the number of layers of feature extraction deepens the reduction of the details and the position information of the target, the expression capability of the final feature map is enriched by adding an SPP module. The method solves the problem that the YOLOv3 fails to detect the oversized image target, realizes automatic identification and positioning of the transmission tower, and provides guarantee for safe operation of a power grid.
Description
Technical Field
The invention belongs to the technical field of image processing, and particularly relates to a transmission tower identification and positioning method based on a high-resolution remote sensing image.
Background
The rapid development of strong intelligent power grids and energy Internet forms a power grid system taking an extra-high voltage transmission channel as a backbone in China. However, the extra-high voltage power grid has a long transmission distance, a large coverage area and complex environmental conditions, and is susceptible to various natural and artificial disasters such as strong wind, freezing, lightning strike, flood and external damage, which often causes serious tower structure failures such as tower collapse, large deformation of a tower head, tower inclination, tower material loss and main material bending. The high-voltage transmission tower is one of the most important infrastructures in the electrical facility, the operation state of the high-voltage transmission tower determines the operation stability and safety of the whole power grid, and the detection of the transmission tower is also extremely important.
At present, the detection of the transmission tower mainly comprises manual field investigation, unmanned aerial vehicle inspection (refer to Wangguan, Liu \ 20429, Wang Binghai and the like), a method for identifying the transmission tower inspected by the unmanned aerial vehicle, which is used for identifying the transmission tower inspected by the unmanned aerial vehicle, which is Chinese, 201510219238.8), helicopter inspection, aviation digital photography inspection, online devices (refer to Zhang Jian Yong, Zhonghai, Zhu Bao Zhong, Wujun and Wang Jing, a positioning method and a device of the position of the tower, which are Chinese, 201410855677.3) and the like, and has the defects of small monitoring range, limited work under severe environment and large-scale disaster conditions and the like, in order to ensure the safe and stable operation of the backbone power grids, a new monitoring means with wide monitoring range, short period, high efficiency and low cost is urgently needed, and on the other hand, remote sensing satellites are rapidly developed, the resolution of remote sensing images is higher and higher, and the intelligentization of the power grid industry is accelerated by utilizing the remote sensing images to detect the targets of large-area transmission towers. However, the existing remote sensing image-based tower detection method generally adopts manual interpretation or machine detection, the manual interpretation needs a large amount of manpower and material resources and is influenced by human subjectivity, and the machine detection method has poor generalization capability and cannot adapt to the diversity of the transmission tower. With the rapid development of artificial intelligence, the intellectualization of national power grids is a common development target of the power grid industry. The automatic detection of the high-voltage transmission tower is extremely important in the intellectualization of the national power grid.
In recent years, rapid development of deep learning has attracted much attention, and it can learn target features from mass image data, and provides an effective framework for automatically extracting target features (refer to: terrestrial peak, liuhua, yellow long tassel, yangbu, xixiong, liucaixi, summary of target detection technology based on deep learning: Computer Systems & Applications,2021,30(3):1-13[ doi:10.15888/j.cnki.csa.007839 ]). For example, YOLOv3, YOLOv3 uses Darknet-53 network, and adopts a multi-scale fusion mode for detection, which greatly improves the detection precision of small targets (refer to: Joseph Redmon, Ali Farhadi University of Washington. YOLOv3: An incorporated Improvement). Wherein YOLOv3 is the target detection network with the most balanced speed and precision so far, and the performance is particularly outstanding in the target detection task. As the transmission towers present different shapes in the remote sensing image and only one type of transmission tower is available in the detection task, the original prior frame is not applicable, and on the other hand, the number of layers of feature extraction deepens the details and the position information of the target and is gradually reduced due to the complex structure of the transmission towers (refer to Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun.
In conclusion, aiming at the problems of high cost, time consumption and labor consumption of a positioning method of a transmission tower, the problems of shape and complex structure of the transmission tower in a high-resolution remote sensing image and the problem that the number of layers of feature extraction deepens the details and the position information of a target are gradually reduced, the invention improves on the basis of the currently most popular YOLOv3 algorithm in the industry so as to realize the detection and positioning of the transmission tower with short construction period, high efficiency and large area and provide support for the intellectualization of a power grid.
Disclosure of Invention
Aiming at the fact that the transmission tower presents different shapes in a remote sensing image, the size of an anchor frame is reset through K-means, a CIoU loss function is introduced into frame regression, DIoU and NMS are combined, and the problem of missed detection of a dense small target by YOLOv3 is solved. Aiming at the problem that the number of layers of feature extraction deepens the details and the position information of a target and gradually decreases, the feature map fusion of local features and global features is realized by adding SPP modules, the expression capability of the final feature map is enriched, and therefore the accuracy of electric tower detection is improved. Finally, the method solves the problems that the YOLOv3 fails to detect the ultra-large image target and the high-precision positioning of the electric tower, and provides support for the intellectualization of the power industry.
In order to achieve the above object, the present invention comprises the steps of:
s1: the data processing mainly comprises geometric correction and image fusion of data, and transmission tower data sets with different backgrounds and different imaging shapes are obtained through manual screening. Dataset picture pixel size 416x 416; making tower pictures into a data set in a PASCAL VOC format, and dividing the data set into a training set, a verification set and a test set to perform data augmentation operation;
s2: calculating an anchor according to the data set in the S1, replacing a IoU loss function and improvement of non-maximum suppression, adding an SPP module, and constructing a transmission tower target detection network;
s3: adjusting network training parameters, and finally adjusting the network parameters of the target detection network obtained in the step S3 by using the training data obtained in the step S1 to obtain a better detection model;
s4: carrying out transmission tower model training according to the test set and the verification set obtained in the step S1 and the training network obtained in the step S3;
s5: testing the transmission tower detection according to the test set and the verification set obtained in the step S1 and the transmission tower detection model obtained in the step S4;
s6: judging whether the model is output or not according to the test result of S5, setting parameters, continuing training and testing, and circulating the operation until a detection model is obtained;
s7: and (5) obtaining a tower detection model according to S6 to detect and position the transmission tower.
Further, the data processing in step S1 mainly includes: the method for processing the original data comprises geometric correction and image fusion, and the used tool is ENVI; cutting the remote sensing image original image into 416x416 pixel values by using a Python program; the data augmentation mainly utilizes a Python program to rotate, translate, zoom and other operations on a data set so as to increase the diversity of data and avoid overfitting; the programming language is python 3.6.
Further, in the step S2, the anchor is modified, and the K-means algorithm is used for multiple calculations and averaging, so that a loss function of IoU and non-maximum suppression (DIoU-NMS) are replaced, so that the detection frame of the tower is more accurate, and the SPP network is added, so that the tower characteristics of different layers are richer; wherein the CIoU loss function and the DIoU-NMS formula are as follows:
CIoU loss function:
wherein:
DIoU-NMS：
wherein:
further, in step S3, the network parameters are adjusted: the maximum number of iterations is 20000, the batch size is 64 (set according to computer performance), the momentum is 0.9, the attenuation coefficient is 0.0005, and the learning rate is 0.001, so that the model converges as quickly as possible, reducing overfitting, the initial learning rate is 0.001, the batch size is 64, and the learning rate is adjusted when the number of iterations is 16000 and 18000.
Further, step S7, detecting and positioning a tower transmission tower, calculating the rough position and position of the tower, cutting a remote sensing image to be detected into the size required by a deep learning target detection network, storing the longitude and latitude coordinates of the upper left corner of the obtained cut image in a txt text, obtaining the pixel coordinates (x, y) of the upper left corner of the minimum circumscribed rectangle of the tower of a detection result in step S5, and calculating the rough coordinates of the tower in the cut tif image; and then, in step S5, obtaining a minimum external rectangular picture of the tower, further identifying four bases of the tower, and finally obtaining an accurate longitude and latitude coordinate of the tower, wherein the pixel coordinate (x, y) is restored to the tif image.
Therefore, the transmission tower identification and positioning method based on the high-resolution remote sensing image provides support for power grid intellectualization and has certain significance for the deep research of the subsequent transmission tower detection based on the remote sensing image, aiming at the problems that the transmission tower detection method based on the remote sensing image generally adopts manual interpretation or machine detection, the manual interpretation needs a large amount of manpower and material resources, is influenced by human subjectivity, and the machine detection method is poor in generalization capability and cannot adapt to the diversity of the transmission tower.
Drawings
The description of the present disclosure will become apparent and readily understood in conjunction with the following drawings, in which:
FIG. 1 is a flow chart of a method for identifying and positioning a transmission tower based on a high-resolution remote sensing image according to the present invention;
FIG. 2 is a diagram of towers in different areas;
FIG. 3 is a diagram of a YOLOv3 network architecture;
FIG. 4 is a block diagram of an SPP module;
FIG. 5 is a large graph detection flow chart;
FIG. 6 is a comparison of the algorithm of the present invention and the original algorithm.
Detailed Description
According to the steps shown in fig. 1, the method for identifying and positioning the transmission tower based on the high-resolution remote sensing image is explained in detail.
Step 1: the data processing comprises the steps of collecting a high-resolution remote sensing image as original data, and carrying out operations such as data annotation and data augmentation, wherein the detailed operations are as follows:
(1) the original image comprises a panchromatic image and a multispectral image, and the ENVI software is used for correcting and fusing the original data to obtain a high-resolution image;
(2) the obtained high-resolution remote sensing image is cut into sizes (416 × 416) required by network training by using a Python program, and the sizes are shown as samples of different regions in FIG. 2;
(3) converting the cut tif images into a jpg format, labeling a data set by using LabLeImg software according to a VOC data format, and randomly dividing the data set into a training set, a verification set and a test set;
(4) the data set is augmented and the Python program is used for rotation, translation, scaling and other operations.
Step 2: modifying the original network, wherein the detailed steps are as follows:
(1) the average clustering result obtained by clustering the data set of the transmission tower for multiple times by adopting a K-means clustering algorithm is as follows: (20,27), (29,35), (31,43), (57,34), (43,58), (59,59), (56,94), (104,47), (98,104) replacing the anchor of the original COCO dataset;
(2) replacing original IoU of YOLOv3 with a CIoU capable of better measuring the overlapping relation between a prediction box and a target box, replacing a non-maximum suppression (NMS) algorithm in YOLOv3 with a DIoU-NMS capable of better eliminating redundant borders, adding an spp module (shown in figure 4) under a network (shown in figure 3) of YOLOv3 to realize fusion of featherMap levels of local features and global features, enriching the expression capability of a final feature map, and accordingly improving the detection accuracy of a power transmission tower (shown in figure 6).
And step 3: setting training parameters: the maximum number of iterations is 20000, the batch size is 64 (set according to computer performance), the momentum is 0.9, the attenuation coefficient is 0.0005, and the learning rate is 0.001, so that the model converges as quickly as possible, reducing overfitting, the initial learning rate is 0.001, the batch size is 64, and the learning rate is adjusted when the number of iterations is 16000 and 18000. The training also converges smoothly after adjusting to the appropriate parameters.
And 4, step 4: and training the transmission tower model according to the test set and the verification set obtained in the step S1 and the training network obtained in the step S3.
And 5: and testing the transmission tower detection according to the test set and the verification set obtained in the step S1 and the transmission tower detection model obtained in the step S4.
Step 6: and judging whether the model is output or not according to the test result of the step S5, setting parameters, continuing training and testing, and circulating the operation until the target detection result is satisfied.
And 7: and (5) obtaining a tower detection model according to the step S6 to detect and position the transmission tower, wherein a large diagram detection flow is shown in FIG. 5, and the detailed steps are as follows:
(1) cutting an image to be identified, storing a corresponding longitude and latitude coordinate of the upper left corner, converting the tif image into a jpg format, selecting a trained target detection model to perform target detection on the transmission tower, obtaining a coordinate value of the upper left corner of a minimum external rectangular frame of the transmission tower, and restoring the coordinate value into the corresponding tif image to obtain a rough longitude and latitude coordinate of the transmission tower;
(2) and (5) cutting the transmission tower detection result obtained in the step (5) by using a rectangular frame, further identifying four bases of the tower, and finally obtaining the accurate longitude and latitude coordinates of the tower, wherein the pixel coordinates (x, y) are restored to the tif image.
The invention relates to a high-resolution remote sensing image-based transmission tower identification and positioning method, which aims at the problems that the traditional high-resolution remote sensing image application, especially the automatic identification and positioning field of targets, depends on manual means and visual interpretation for a long time, the targets are easy to be misjudged and misjudged, and the detection precision is low. Therefore, the intelligent level of the power system is improved by applying the deep learning to the automatic identification and positioning of the transmission tower of the high-resolution-ratio remote sensing image.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (8)
1. A transmission tower identification and positioning method based on a high-resolution remote sensing image is characterized by comprising the following steps:
s1: data processing, wherein a data set of a transmission tower is mainly acquired by a high-score series and a high-view first satellite and is obtained by geometric correction, image fusion and the like; manually screening to obtain data sets of the transmission towers with different backgrounds and different imaging shapes; dataset picture pixel size 416x 416; making tower pictures into a data set in a PASCAL VOC format, and dividing the data set into a training set, a verification set and a test set to perform data augmentation operation;
s2: calculating an anchor according to the data set in the S1, replacing a IoU loss function and improvement of non-maximum suppression, adding an SPP module, and constructing a transmission tower target detection network;
s3: adjusting network training parameters, and finally adjusting the network parameters of the target detection network obtained in the step S2 by using the training data obtained in the step S1;
s4: carrying out transmission tower model training according to the test set and the verification set obtained in the step S1 and the detection model obtained in the step S3;
s5: testing the transmission tower detection according to the test set and the verification set obtained in the step S1 and the transmission tower detection model obtained in the step S4;
s6: judging whether the model is output or not according to the test result of S5, setting parameters, continuing training and testing, and circulating the operation until a target detection model is obtained;
s7: and (5) obtaining a tower detection model according to S6 to detect and position the transmission tower.
2. The method for identifying and positioning the transmission tower based on the high-resolution remote sensing image according to claim 1, wherein the data processing in the step S1 mainly comprises geometric correction and image fusion of images, and the tool is ENVI; data clipping is to clip the remote sensing image original image into 416 × 416 pixel values by using a Python program; the data augmentation mainly utilizes Python programs to rotate, translate, zoom and the like on the data set to increase the diversity of data and avoid over-training, and the programming language is Python 3.6.
3. The method for identifying and positioning the transmission tower based on the high-resolution remote sensing image according to claim 1, wherein the algorithm modification in the step S2 comprises replacing an original anchor value, the anchor is obtained by using a K-means algorithm for multiple calculations and averaging, a IoU loss function is replaced, a non-maximum suppression function (DIoU-NMS) is modified, and tower characteristics of different layers fused with an SPP network are increased; wherein the CIoU loss function and the DIoU-NMS formula are as follows:
CIoU loss function:
wherein:
wherein:
4. the method for identifying and positioning the transmission tower based on the high-resolution remote sensing image according to claim 1, wherein in the step S3, network parameters are adjusted: the maximum number of iterations is 20000, the batch size is 64 (set according to computer performance), the momentum is 0.9, the attenuation coefficient is 0.0005, and the learning rate is 0.001, so that the model converges as quickly as possible, reducing overfitting, the initial learning rate is 0.001, the batch size is 64, and the learning rate is adjusted when the number of iterations is 16000 and 18000.
5. The method for identifying and positioning the transmission tower based on the high-resolution remote sensing image as claimed in claim 1, wherein the step S4 is performed to train the model of the transmission tower according to the network with the adjusted parameters obtained in the step S3.
6. The method for identifying and positioning the transmission tower based on the high-resolution remote sensing image according to claim 1, wherein the step S5 is used for testing the transmission tower according to the test set and the verification set obtained in the step S1 and the transmission tower detection model obtained in the step S4.
7. The method for identifying and positioning the transmission tower based on the high-resolution remote sensing image as claimed in claim 1, wherein the step S6 is performed according to the test result of the step S5 to judge whether the model is output or not, and to perform parameter setting, training and testing, and the operation is repeated until the target detection model is obtained.
8. The method for identifying and positioning the transmission tower based on the high-resolution remote sensing image according to claim 1, wherein the step S7 is that the target detection and positioning mainly comprises rough positioning and precise positioning of the transmission tower, and the calculation of the rough position and the position of the transmission tower is that the upper-left pixel coordinate (x, y) of the minimum circumscribed rectangle of the transmission tower of the detection result is obtained in the step S5, and the rough coordinate of the transmission tower is obtained in the cut tif image; and then, in the step S5, obtaining the minimum external rectangular picture of the tower, further identifying four bases of the tower, and finally obtaining the accurate longitude and latitude coordinates of the tower, wherein the pixel coordinates (x, y) are restored to the tif image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110355197.0A CN113033446B (en) | 2021-04-01 | 2021-04-01 | Transmission tower identification and positioning method based on high-resolution remote sensing image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110355197.0A CN113033446B (en) | 2021-04-01 | 2021-04-01 | Transmission tower identification and positioning method based on high-resolution remote sensing image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113033446A true CN113033446A (en) | 2021-06-25 |
| CN113033446B CN113033446B (en) | 2024-02-02 |
Family
ID=76453755
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110355197.0A Active CN113033446B (en) | 2021-04-01 | 2021-04-01 | Transmission tower identification and positioning method based on high-resolution remote sensing image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113033446B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114998242A (en) * | 2022-05-26 | 2022-09-02 | 电子科技大学 | Method for detecting power transmission line pole tower in satellite image |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2019101149A4 (en) * | 2019-09-30 | 2019-10-31 | Hu, Yaowen MR | An Image retrieval System for Brand Logos Based on Deep Learning |
| CN110728658A (en) * | 2019-09-16 | 2020-01-24 | 武汉大学 | High-resolution remote sensing image weak target detection method based on deep learning |
| CN111553204A (en) * | 2020-04-10 | 2020-08-18 | 国网内蒙古东部电力有限公司 | Transmission tower detection method based on remote sensing image |
| CN111950343A (en) * | 2020-06-24 | 2020-11-17 | 中国电力科学研究院有限公司 | Automatic transmission tower identification method and system based on satellite remote sensing technology |
| CN112287788A (en) * | 2020-10-20 | 2021-01-29 | 杭州电子科技大学 | Pedestrian detection method based on improved YOLOv3 and improved NMS |
-
2021
- 2021-04-01 CN CN202110355197.0A patent/CN113033446B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110728658A (en) * | 2019-09-16 | 2020-01-24 | 武汉大学 | High-resolution remote sensing image weak target detection method based on deep learning |
| AU2019101149A4 (en) * | 2019-09-30 | 2019-10-31 | Hu, Yaowen MR | An Image retrieval System for Brand Logos Based on Deep Learning |
| CN111553204A (en) * | 2020-04-10 | 2020-08-18 | 国网内蒙古东部电力有限公司 | Transmission tower detection method based on remote sensing image |
| CN111950343A (en) * | 2020-06-24 | 2020-11-17 | 中国电力科学研究院有限公司 | Automatic transmission tower identification method and system based on satellite remote sensing technology |
| CN112287788A (en) * | 2020-10-20 | 2021-01-29 | 杭州电子科技大学 | Pedestrian detection method based on improved YOLOv3 and improved NMS |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114998242A (en) * | 2022-05-26 | 2022-09-02 | 电子科技大学 | Method for detecting power transmission line pole tower in satellite image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113033446B (en) | 2024-02-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107742093B (en) | Real-time detection method, server and system for infrared image power equipment components | |
| CN109376605B (en) | Electric power inspection image bird-stab-prevention fault detection method | |
| US20210026377A1 (en) | Method and Device for Generating an Unmanned Aerial Vehicle Flight Trajectory, Computer Apparatus and Storage Medium | |
| CN110610483B (en) | Crack image acquisition and detection method, computer equipment and readable storage medium | |
| CN113286129B (en) | Inspection method and system for photovoltaic power station | |
| CN114373009B (en) | Building shadow height measurement intelligent calculation method based on high-resolution remote sensing image | |
| CN111862013A (en) | Insulator detection method, device and equipment based on deep convolutional neural network | |
| CN116503318A (en) | Aerial insulator multi-defect detection method, system and equipment integrating CAT-BiFPN and attention mechanism | |
| CN114596278A (en) | Method and device for detecting hot spot defects of photovoltaic panel of photovoltaic power station | |
| Sampath et al. | Estimation of rooftop solar energy generation using Satellite Image Segmentation | |
| CN113393459A (en) | Infrared image photovoltaic module visual identification method based on example segmentation | |
| CN116258980A (en) | Unmanned aerial vehicle distributed photovoltaic power station inspection method based on vision | |
| CN113033446A (en) | Transmission tower identification and positioning method based on high-resolution remote sensing image | |
| CN111753666A (en) | Method and system for detecting faults of small targets in power transmission line and storage medium | |
| CN103426153A (en) | Unmanned aerial vehicle remote sensing image quick splicing method | |
| Zhou et al. | Fault detection method of glass insulator aerial image based on the improved YOLOv5 | |
| CN117574317A (en) | Mountain fire monitoring method and device based on multi-mode data fusion of sky and ground | |
| CN116385465A (en) | Image segmentation model construction and image segmentation method, system, equipment and medium | |
| Wang et al. | Insulator defect detection based on improved YOLOv5 algorithm | |
| CN112465856A (en) | Unmanned aerial vehicle-based ship track correction method and device and electronic equipment | |
| CN116029440A (en) | Ultra-short-term power prediction method and device for photovoltaic power station | |
| KR20180082166A (en) | The Apparatus For Detecting Solar Panel | |
| CN115937492A (en) | Transformer equipment infrared image identification method based on feature identification | |
| CN113011295B (en) | Method, computer equipment and medium for identifying photovoltaic power station based on remote sensing image | |
| CN115482473A (en) | Graph convolution network model for extracting aerial image features and method for detecting abnormity |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |