CN113960068A - Wind power blade damage detection method - Google Patents
Wind power blade damage detection method Download PDFInfo
- Publication number
- CN113960068A CN113960068A CN202111392653.5A CN202111392653A CN113960068A CN 113960068 A CN113960068 A CN 113960068A CN 202111392653 A CN202111392653 A CN 202111392653A CN 113960068 A CN113960068 A CN 113960068A
- Authority
- CN
- China
- Prior art keywords
- image
- data acquisition
- area
- information
- wind turbine
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 15
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 230000007547 defect Effects 0.000 claims abstract description 12
- 238000013480 data collection Methods 0.000 claims abstract description 10
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 5
- 238000000034 method Methods 0.000 claims description 15
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 238000007635 classification algorithm Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 230000005611 electricity Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000004576 sand Substances 0.000 description 2
- 208000004350 Strabismus Diseases 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 238000011179 visual inspection Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Wind Motors (AREA)
Abstract
The utility model discloses a wind-powered electricity generation blade damage detection method, including: s1, setting M x N data acquisition regions aiming at the region of the wind turbine blade, wherein M and N are natural numbers; s2, sequentially collecting image information of the data collection area, wherein the image information at least comprises sequence number information of the data collection area; s3, extracting first features in the image information aiming at different data acquisition areas; s4, based on the first features, splicing the image information of the data acquisition regions based on a feature matching algorithm to obtain a spliced image frame; and S5, performing image recognition on the spliced image frame, and judging whether defect characteristics exist or not. According to the wind turbine blade damage detection method provided by the invention, the corresponding damage area on the wind turbine blade can be determined by carrying out multiple local data acquisition on the wind turbine blade, then splicing the acquired data into a whole image, and carrying out image recognition based on a convolutional neural network algorithm model on the whole image.
Description
Technical Field
The invention relates to an image identification method, in particular to a wind power blade damage detection method.
Background
Wind energy has been widely used in various fields as a clean and renewable environment-friendly energy source. For the purpose of maximizing wind energy utilization, wind farms are generally selected in open areas such as the sea, the gobi desert, and mountainous areas. Most of onshore wind power is built in areas with concentrated wind power resources, but the areas usually have the characteristic of frequent wind and sand weather, the surface of the blade of the wind driven generator is eroded by wind and sand all the year round to cause various defects such as surface pits, scratches, cracks and the like, and the defects are repaired in time to cause irreversible major safety accidents. At present, the traditional blade surface defect detection method mainly comprises knocking sound discrimination, ground telescope observation, manual visual inspection and the like, and the various methods listed above have the defects of large blind area, long time consumption, high labor intensity, high-altitude operation, low efficiency and the like.
Disclosure of Invention
In view of the foregoing problems in the prior art, an object of an aspect of the present invention is to provide a wind turbine blade damage detection method capable of detecting wind turbine blade damage by using a machine vision method.
In order to achieve the above object, a wind turbine blade damage detection method according to an aspect of the present invention includes:
s1, setting M x N data acquisition regions aiming at the region of the wind turbine blade, wherein M and N are natural numbers;
s2, sequentially collecting image information of the data collection area, wherein the image information at least comprises sequence number information of the data collection area;
s3, extracting first features in the image information aiming at different data acquisition areas;
s4, based on the first features, splicing the image information of the data acquisition regions based on a feature matching algorithm to obtain a spliced image frame;
and S5, performing image recognition on the spliced image frame, and judging whether defect characteristics exist or not.
Preferably, the image recognition is performed by any one or more of the following: an image retrieval algorithm, a matching algorithm for sample images of known classification, or an image classification algorithm.
Preferably, when the data acquisition region of the wind turbine blade is divided, the adjustable cradle head is used for adjusting the image acquisition unit to take the blade rotation center region as a first data acquisition region, and then other regions surrounding the first data acquisition region are sequentially formed through the adjustable cradle head; wherein M = N = 3.
Preferably, the extracting the first feature in the image information includes:
determining a first area of the picture information adjacent to the first area according to the serial number information of the picture information;
mapping the first areas of two adjacent images to the same coordinate system;
based on the same coordinate system, based on RGB pixel matching;
setting an area with a matching value higher than a threshold value as a confidence area based on the set threshold value;
storing the confidence region as a first feature.
Preferably, when image stitching is performed based on a feature matching algorithm, the method includes: and carrying out image splicing based on an OpenCV SURF algorithm.
Preferably, the image recognition includes:
extracting a second feature contained in the image aiming at the spliced image;
and inputting the second features into a convolutional neural network algorithm model for feature recognition.
Preferably, the extracting the second feature included in the image includes:
calling a wind power blade defect training library file, and extracting a key area in the training library file;
and when the key area is detected, recording the coordinates of the center point of the key area.
Preferably, before performing image stitching and/or image recognition, the method further includes preprocessing the image information, where the preprocessing includes:
converting the image data into a gray scale image;
histogram equalization processing and/or median filtering are performed to eliminate some of the noise in the image.
Preferably, the pretreatment further comprises:
setting an interested area for the image data after the noise is eliminated;
acquiring the geometric center of the region of interest;
the coordinates of the geometric center are saved.
According to the wind turbine blade damage detection method provided by the invention, the corresponding damage area on the wind turbine blade can be determined by carrying out multiple local data acquisition on the wind turbine blade, then splicing the acquired data into a whole image, and carrying out image recognition based on a convolutional neural network algorithm model on the whole image. Compared with manual inspection diagnosis through a telescope, the method has the advantages that the efficiency and the safety are improved.
Drawings
Fig. 1 is a flowchart of a wind turbine blade damage detection method according to the present invention.
Fig. 2 is a system structure block diagram applied to the wind turbine blade damage detection method of the present invention.
Fig. 3 is a schematic diagram of a data acquisition region divided for a wind turbine blade in the wind turbine blade damage detection method of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
Various aspects and features of the present invention are described herein with reference to the drawings.
These and other characteristics of the invention will become apparent from the following description of a preferred form of embodiment, given as a non-limiting example, with reference to the accompanying drawings.
It should also be understood that, although the invention has been described with reference to some specific examples, a person of skill in the art shall certainly be able to achieve many other equivalent forms of the invention, having the characteristics as set forth in the claims and hence all coming within the field of protection defined thereby.
The above and other aspects, features and advantages of the present invention will become more apparent in view of the following detailed description when taken in conjunction with the accompanying drawings.
As shown in fig. 1, a wind turbine blade damage detection method according to an aspect of the present invention includes:
s1, setting M x N data acquisition regions aiming at the region of the wind turbine blade, wherein M and N are natural numbers; fig. 3 illustrates an implementation manner, in which, as illustrated in fig. 3, when data acquisition regions are divided for blades of a wind turbine, an adjustable cradle head is used to adjust an image acquisition unit to use a blade rotation center region as a first data acquisition region, and then other regions surrounding the first data acquisition region are sequentially formed by the adjustable cradle head; wherein M = N = 3. The adjustable cloud platform, the adjustable cloud platform that is sold in the market at present all can, and the staff can set for initial position based on the blade center, then upper left respectively, the top, upper right, lower right, down, left side suitable position that squints in proper order to make wind-powered electricity generation blade all cover. And 3 × 3 data acquisition regions are finally formed.
S2, sequentially collecting image information of the data collection area, wherein the image information at least comprises sequence number information of the data collection area; in other words, after the data acquisition region is set, the image information adjacent to each acquired image in the upper, lower, left, and right directions thereof has been determined at the time of output, that is, each image has included the serial number information at the time of output.
S3, extracting first features in the image information aiming at different data acquisition areas; extracting a first feature in the image information specifically includes: determining a first area of the picture information adjacent to the first area according to the serial number information of the picture information; mapping the first areas of two adjacent images to the same coordinate system; based on the same coordinate system, based on RGB pixel matching; setting an area with a matching value higher than a threshold value as a confidence area based on the set threshold value; storing the confidence region as a first feature.
S4, based on the first features, splicing the image information of the data acquisition regions based on a feature matching algorithm to obtain a spliced image frame; when image splicing is carried out based on a feature matching algorithm, the method comprises the following steps: and carrying out image splicing based on an OpenCV SURF algorithm.
And S5, performing image recognition on the spliced image frame, and judging whether defect characteristics exist or not. In some embodiments, the image recognition comprises: extracting a second feature contained in the image aiming at the spliced image; and inputting the second features into a convolutional neural network algorithm model for feature recognition. Preferably, the extracting the second feature included in the image includes: calling a wind power blade defect training library file, and extracting a key area in the training library file; and when the key area is detected, recording the coordinates of the center point of the key area. Preferably, the image recognition is performed by any one or more of the following: an image retrieval algorithm, a matching algorithm for sample images of known classification, or an image classification algorithm.
In addition, before image splicing and/or image recognition, preprocessing is further included for the image information, and the preprocessing includes: converting the image data into a gray scale image; histogram equalization processing and/or median filtering are performed to eliminate some of the noise in the image. Or, the preprocessing further comprises: setting an interested area for the image data after the noise is eliminated; acquiring the geometric center of the region of interest; the coordinates of the geometric center are saved.
Fig. 2 is a schematic structural diagram of a system to which the wind turbine blade damage detection method of the present invention is applied. As shown in fig. 2, the system includes an adjustable pan-tilt, an image acquisition unit, an image preprocessing unit, an image stitching unit, and an image recognition unit, where the image acquisition unit is configured to set M × N data acquisition regions for regions of a wind turbine blade, where M and N are natural numbers; sequentially collecting image information of the data collection area, wherein the image information at least comprises serial number information of the data collection area; the image preprocessing unit is configured to convert the image data into a grayscale image; performing histogram equalization processing and/or performing median filtering to eliminate partial noise in the image; the image stitching unit is configured to extract a first feature in the image information for different data acquisition regions; based on the first characteristics, image information of the data acquisition regions is spliced based on a characteristic matching algorithm to obtain a spliced image frame; the image identification unit is configured to perform image identification on the spliced image frames and judge whether defect characteristics exist or not.
The above embodiments are only exemplary embodiments of the present invention, and are not intended to limit the present invention, and the scope of the present invention is defined by the claims. Various modifications and equivalents may be made by those skilled in the art within the spirit and scope of the present invention, and such modifications and equivalents should also be considered as falling within the scope of the present invention.
Claims (9)
1. The wind power blade damage detection method comprises the following steps:
s1, setting M x N data acquisition regions aiming at the region of the wind turbine blade, wherein M and N are natural numbers;
s2, sequentially collecting image information of the data collection area, wherein the image information at least comprises sequence number information of the data collection area;
s3, extracting first features in the image information aiming at different data acquisition areas;
s4, based on the first features, splicing the image information of the data acquisition regions based on a feature matching algorithm to obtain a spliced image frame;
and S5, performing image recognition on the spliced image frame, and judging whether defect characteristics exist or not.
2. The method of claim 1, wherein the image recognition is performed by any one or more of: an image retrieval algorithm, a matching algorithm for sample images of known classification, or an image classification algorithm.
3. The method according to claim 1, wherein when data acquisition regions are divided for the blades of the wind turbine generator, the adjustable cradle head is used for adjusting the image acquisition unit to take the rotation center region of the blade as a first data acquisition region, and then other regions surrounding the first data acquisition region are sequentially formed through the adjustable cradle head; wherein M = N = 3.
4. The method of claim 3, extracting a first feature in the image information, comprising:
determining a first area of the picture information adjacent to the first area according to the serial number information of the picture information;
mapping the first areas of two adjacent images to the same coordinate system;
based on the same coordinate system, based on RGB pixel matching;
setting an area with a matching value higher than a threshold value as a confidence area based on the set threshold value;
storing the confidence region as a first feature.
5. The method of claim 4, when performing image stitching based on the feature matching algorithm, comprising: and carrying out image splicing based on an OpenCV SURF algorithm.
6. The method of claim 1, when performing image recognition, comprising:
extracting a second feature contained in the image aiming at the spliced image;
and inputting the second features into a convolutional neural network algorithm model for feature recognition.
7. The method of claim 6, extracting a second feature contained in the image, comprising:
calling a wind power blade defect training library file, and extracting a key area in the training library file;
and when the key area is detected, recording the coordinates of the center point of the key area.
8. The method of claim 1, further comprising preprocessing the image information prior to image stitching and/or image recognition, the preprocessing comprising:
converting the image data into a gray scale image;
histogram equalization processing and/or median filtering are performed to eliminate some of the noise in the image.
9. The method of claim 8, the pre-processing, further comprising:
setting an interested area for the image data after the noise is eliminated;
acquiring the geometric center of the region of interest;
the coordinates of the geometric center are saved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111392653.5A CN113960068A (en) | 2021-11-23 | 2021-11-23 | Wind power blade damage detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111392653.5A CN113960068A (en) | 2021-11-23 | 2021-11-23 | Wind power blade damage detection method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113960068A true CN113960068A (en) | 2022-01-21 |
Family
ID=79471638
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111392653.5A Pending CN113960068A (en) | 2021-11-23 | 2021-11-23 | Wind power blade damage detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113960068A (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106841214A (en) * | 2017-01-21 | 2017-06-13 | 兰州理工大学 | A kind of non-contact wind power blade dust storm erosion degree detection method |
| CN108254077A (en) * | 2018-01-02 | 2018-07-06 | 国网上海市电力公司 | The GIS thermal fault detection methods merged based on part with global characteristics information |
| CN110136059A (en) * | 2019-04-04 | 2019-08-16 | 哈尔滨理工大学 | The wind electricity blade image split-joint method of linear feature |
| CN110378879A (en) * | 2019-06-26 | 2019-10-25 | 杭州电子科技大学 | A kind of Bridge Crack detection method |
| CN111539409A (en) * | 2020-04-09 | 2020-08-14 | 武汉大学 | Ancient tomb question and character recognition method based on hyperspectral remote sensing technology |
| CN112085694A (en) * | 2020-06-30 | 2020-12-15 | 数笙智能科技(上海)有限公司 | Artificial intelligence automatic inspection wind energy fan blade system |
| CN112200911A (en) * | 2020-11-06 | 2021-01-08 | 北京易达恩能科技有限公司 | Region overlapping type three-dimensional map construction method and device combined with markers |
| CN112360699A (en) * | 2020-10-22 | 2021-02-12 | 华能大理风力发电有限公司 | Intelligent inspection and diagnosis analysis method for blades of full-automatic wind generating set |
| CN112730422A (en) * | 2020-12-01 | 2021-04-30 | 中广核工程有限公司 | Nuclear power station containment vessel defect detection method and system |
| CN112881467A (en) * | 2021-03-15 | 2021-06-01 | 中国空气动力研究与发展中心超高速空气动力研究所 | Large-size composite material damage imaging and quantitative identification method |
| CN113176268A (en) * | 2021-05-18 | 2021-07-27 | 哈尔滨理工大学 | Wind power blade surface damage detection method based on cloud deck shooting image |
| CN113205458A (en) * | 2021-05-28 | 2021-08-03 | 上海扩博智能技术有限公司 | Weak texture blade splicing method, system, equipment and medium |
-
2021
- 2021-11-23 CN CN202111392653.5A patent/CN113960068A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106841214A (en) * | 2017-01-21 | 2017-06-13 | 兰州理工大学 | A kind of non-contact wind power blade dust storm erosion degree detection method |
| CN108254077A (en) * | 2018-01-02 | 2018-07-06 | 国网上海市电力公司 | The GIS thermal fault detection methods merged based on part with global characteristics information |
| CN110136059A (en) * | 2019-04-04 | 2019-08-16 | 哈尔滨理工大学 | The wind electricity blade image split-joint method of linear feature |
| CN110378879A (en) * | 2019-06-26 | 2019-10-25 | 杭州电子科技大学 | A kind of Bridge Crack detection method |
| CN111539409A (en) * | 2020-04-09 | 2020-08-14 | 武汉大学 | Ancient tomb question and character recognition method based on hyperspectral remote sensing technology |
| CN112085694A (en) * | 2020-06-30 | 2020-12-15 | 数笙智能科技(上海)有限公司 | Artificial intelligence automatic inspection wind energy fan blade system |
| CN112360699A (en) * | 2020-10-22 | 2021-02-12 | 华能大理风力发电有限公司 | Intelligent inspection and diagnosis analysis method for blades of full-automatic wind generating set |
| CN112200911A (en) * | 2020-11-06 | 2021-01-08 | 北京易达恩能科技有限公司 | Region overlapping type three-dimensional map construction method and device combined with markers |
| CN112730422A (en) * | 2020-12-01 | 2021-04-30 | 中广核工程有限公司 | Nuclear power station containment vessel defect detection method and system |
| CN112881467A (en) * | 2021-03-15 | 2021-06-01 | 中国空气动力研究与发展中心超高速空气动力研究所 | Large-size composite material damage imaging and quantitative identification method |
| CN113176268A (en) * | 2021-05-18 | 2021-07-27 | 哈尔滨理工大学 | Wind power blade surface damage detection method based on cloud deck shooting image |
| CN113205458A (en) * | 2021-05-28 | 2021-08-03 | 上海扩博智能技术有限公司 | Weak texture blade splicing method, system, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111275679B (en) | Image-based solar cell defect detection system and method | |
| CN109145872B (en) | CFAR and Fast-RCNN fusion-based SAR image ship target detection method | |
| CN111797712B (en) | Remote sensing image cloud and cloud shadow detection method based on multi-scale feature fusion network | |
| CN107895376A (en) | Based on the solar panel recognition methods for improving Canny operators and contour area threshold value | |
| CN111862013B (en) | Insulator detection method, device and equipment based on deep convolutional neural network | |
| CN103442209A (en) | Video monitoring method of electric transmission line | |
| CN104483326A (en) | High-voltage wire insulator defect detection method and high-voltage wire insulator defect detection system based on deep belief network | |
| CN108108736A (en) | A kind of solar energy photovoltaic panel spot identification method | |
| CN106841214A (en) | A kind of non-contact wind power blade dust storm erosion degree detection method | |
| CN114973032B (en) | Deep convolutional neural network-based photovoltaic panel hot spot detection method and device | |
| CN113487563B (en) | EL image-based self-adaptive detection method for hidden cracks of photovoltaic module | |
| CN108537170A (en) | A kind of power equipment firmware unmanned plane inspection pin missing detection method | |
| CN111178177A (en) | Cucumber disease identification method based on convolutional neural network | |
| CN114241364A (en) | Method for quickly calibrating foreign object target of overhead transmission line | |
| CN117314893A (en) | Quality detection method for photovoltaic steel structure component based on image processing | |
| CN111667473A (en) | Insulator hydrophobicity grade judging method based on improved Canny algorithm | |
| CN114842243A (en) | Fault detection method, system and equipment based on large-span inspection robot | |
| CN104573662A (en) | Cloud judgment method and system | |
| CN110136059A (en) | The wind electricity blade image split-joint method of linear feature | |
| CN112465776B (en) | Crack intelligent detection method based on wind turbine surface blurred image | |
| CN110082355A (en) | A kind of wind electricity blade detection system | |
| CN112381794B (en) | Printing defect detection method based on deep convolution generation network | |
| CN113960068A (en) | Wind power blade damage detection method | |
| Huang et al. | Mango surface defect detection based on HALCON | |
| Zhang et al. | Wind Turbine Blade Damage Detection Based on the Improved YOLOv5 Algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220121 |
|
| RJ01 | Rejection of invention patent application after publication |