CN114529537A - Abnormal target detection method, system, equipment and medium for photovoltaic panel - Google Patents
Abnormal target detection method, system, equipment and medium for photovoltaic panel Download PDFInfo
- Publication number
- CN114529537A CN114529537A CN202210173145.6A CN202210173145A CN114529537A CN 114529537 A CN114529537 A CN 114529537A CN 202210173145 A CN202210173145 A CN 202210173145A CN 114529537 A CN114529537 A CN 114529537A
- Authority
- CN
- China
- Prior art keywords
- photovoltaic panel
- image
- abnormal
- photovoltaic
- position information
- 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
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 138
- 238000001514 detection method Methods 0.000 title claims abstract description 75
- 230000011218 segmentation Effects 0.000 claims abstract description 48
- 238000000034 method Methods 0.000 claims abstract description 26
- 238000012549 training Methods 0.000 claims description 32
- 238000004590 computer program Methods 0.000 claims description 9
- 238000002372 labelling Methods 0.000 claims description 8
- 230000002547 anomalous effect Effects 0.000 claims description 7
- 238000010248 power generation Methods 0.000 abstract description 7
- 238000007689 inspection Methods 0.000 abstract description 5
- 238000009825 accumulation Methods 0.000 abstract description 4
- 239000002699 waste material Substances 0.000 abstract description 4
- 230000005684 electric field Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 3
- 239000000428 dust Substances 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 210000003608 fece Anatomy 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
Images
Classifications
-
- 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
- G06T7/0008—Industrial image inspection checking presence/absence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/50—Photovoltaic [PV] energy
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Photovoltaic Devices (AREA)
Abstract
The invention discloses a method, a system, equipment and a medium for detecting abnormal targets of photovoltaic panels, wherein the method for detecting the abnormal targets of the photovoltaic panels obtains images of the photovoltaic panels in a photovoltaic power station, inputs the images of the photovoltaic panels into a segmentation model to obtain segmentation data of the photovoltaic panels, the segmentation data of the photovoltaic panels comprise background-removed image information of the photovoltaic panels and first position information corresponding to each photovoltaic panel, utilizes a target detection model to identify the abnormal targets in the segmentation data of the photovoltaic panels and output second position information of the abnormal targets, determines the abnormal photovoltaic panels where the abnormal targets are located based on the first position information and the second position information, can quickly and accurately identify the abnormal targets after background interference is removed, avoids loss caused by accumulation of foreign matters of the photovoltaic panels, improves the power generation capacity of a photovoltaic electric field, and further can enhance the safety of photovoltaic power station power generation, meanwhile, the waste of manpower resources caused by manual inspection can be avoided.
Description
Technical Field
The invention relates to the technical field of detection, in particular to a method, a system, equipment and a medium for detecting an abnormal target of a photovoltaic panel.
Background
With the development of national economy of China, the great investment and construction of a novel energy industry, particularly photovoltaic power generation, are used as novel, clean and renewable energy sources in the 21 st century. In a photovoltaic station which is built and put into operation, a photovoltaic panel is placed outdoors, and foreign matters such as dust, bird droppings, a covering and the like are easily accumulated on the surface of the photovoltaic panel, which directly affects the power generation efficiency and the service life of the photovoltaic panel.
For the detection of the abnormal target on the surface of the photovoltaic panel, there are three main ways at present: the method comprises the following steps of artificial subjective judgment, sensor pre-installation and computer vision-based detection. The defects of the traditional artificial subjective judgment method include low detection efficiency, high labor cost, high risk and strong subjectivity. The dust sensor has the defects of high installation cost, low sensitivity to some types of foreign matters, high operation and maintenance cost and the like. The detection based on computer vision is mainly based on threshold binarization by a traditional image processing method at present, or dust/foreign matters in a photovoltaic panel in an aerial image of an unmanned aerial vehicle are detected by using a single neural network, but the foreign matters are difficult to accurately detect by using the computer vision method, on one hand, the background of the picture shot by the unmanned aerial vehicle is very complex, on the other hand, the size of the foreign matters is uncertain, most of the foreign matters are small in size, the occupation ratio in the photovoltaic panel is very small, and therefore, the accuracy of a single end-to-end detection model is very low.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a method, a system, equipment and a medium for detecting an abnormal target of a photovoltaic panel.
The invention solves the technical problems through the following technical scheme:
the invention provides an abnormal target detection method of a photovoltaic panel, which comprises the following steps:
acquiring a photovoltaic panel image in a photovoltaic power station;
inputting the photovoltaic panel image into a segmentation model to obtain photovoltaic panel segmentation data; the photovoltaic panel segmentation data comprise photovoltaic panel image information with background removed and first position information corresponding to each photovoltaic panel;
identifying an abnormal target in the photovoltaic panel segmentation data by using a target detection model and outputting second position information of the abnormal target; the detection network of the target detection model comprises at least two levels of anchor frames, each level comprises a plurality of different anchor frames, and different levels correspond to different anchor frame size ranges;
and determining the abnormal photovoltaic panel where the abnormal target is located based on the first position information and the second position information.
Preferably, the abnormal target detection method further includes:
acquiring the number information of each photovoltaic panel in the photovoltaic panel image and the positioning information corresponding to the photovoltaic panel image shot by shooting equipment in the photovoltaic power station;
determining coordinate information of the abnormal photovoltaic panel based on the number information, the first position information, the second position information, and the positioning information.
Preferably, the segmentation model is obtained by inputting historical image training data into a Mask-RCNN model (an image segmentation model) for training; the historical image training data is a historical image subjected to outline labeling of the photovoltaic panel region.
Preferably, the target detection model is obtained by inputting historical abnormal image training data into a RetinaNet model (a target detection model) for training; and the historical abnormal image training data is a historical abnormal image which is subjected to abnormal target labeling.
Preferably, the step of acquiring the positioning information corresponding to the photovoltaic panel image shot by the shooting device in the photovoltaic power station includes:
acquiring equipment positioning information when shooting equipment shoots the photovoltaic panel image in the photovoltaic power station, and taking the equipment positioning information as the positioning information of the photovoltaic panel image.
Preferably, the step of acquiring an image of a photovoltaic panel in the photovoltaic power plant comprises:
and acquiring a high-altitude image of the photovoltaic panel in the photovoltaic power station as the photovoltaic panel image.
Preferably, before the step of acquiring an image of a photovoltaic panel in a photovoltaic power plant, the abnormal target detection method further includes:
a shooting route is made according to the distribution condition of the photovoltaic panels in the photovoltaic power station; wherein the distribution condition comprises at least one of quantity distribution and area distribution of the photovoltaic panels;
and remotely controlling the shooting equipment to shoot along the shooting route.
The present invention also provides an abnormal target detection system of a photovoltaic panel, the abnormal target detection system including:
the image acquisition module is used for acquiring a photovoltaic panel image in the photovoltaic power station;
the segmentation module is used for inputting the photovoltaic panel image into a segmentation model so as to obtain photovoltaic panel segmentation data; the photovoltaic panel segmentation data comprise photovoltaic panel image information with background removed and first position information corresponding to each photovoltaic panel;
the identification module is used for identifying an abnormal target in the photovoltaic panel segmentation data by using a target detection model and outputting second position information of the abnormal target; the detection network of the target detection model comprises at least two levels of anchor frames, each level comprises a plurality of different anchor frames, and different levels correspond to different anchor frame size ranges;
and the first determining module is used for determining the abnormal photovoltaic panel where the abnormal target is located based on the first position information and the second position information.
Preferably, the abnormal target detecting system further includes:
the information acquisition module is used for acquiring the number information of each photovoltaic panel in the photovoltaic panel image and the positioning information corresponding to the photovoltaic panel image shot by the shooting equipment in the photovoltaic power station;
and the second determining module is used for determining the coordinate information of the abnormal photovoltaic panel based on the number information, the first position information, the second position information and the positioning information.
Preferably, the segmentation model is obtained by inputting historical image training data into a Mask-RCNN model for training; the historical image training data is a historical image subjected to outline labeling of the photovoltaic panel region.
Preferably, the target detection model is obtained by inputting historical abnormal image training data into a RetinaNet model for training; and the historical abnormal image training data is a historical abnormal image which is subjected to abnormal target labeling.
Preferably, the information obtaining module is further configured to obtain device positioning information when a shooting device shoots the photovoltaic panel image in the photovoltaic power station, and use the device positioning information as positioning information of the photovoltaic panel image.
Preferably, the image acquisition module is further configured to acquire an aerial image of the photovoltaic panel in the photovoltaic power station as the photovoltaic panel image.
Preferably, the abnormal target detecting system further includes:
the shooting module is used for making a shooting route according to the distribution condition of the photovoltaic panels in the photovoltaic power station; wherein the distribution condition comprises at least one of quantity distribution and area distribution of the photovoltaic panels;
and remotely controlling the shooting equipment to shoot along the shooting route.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the computer program to realize the abnormal target detection method of the photovoltaic panel.
The present invention also provides a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the abnormal target detection method of a photovoltaic panel as described above.
The positive progress effects of the invention are as follows: according to the method, after background interference is removed by using the segmentation model, the abnormal target is quickly and accurately identified by using the target detection model, so that the detection efficiency of the abnormal target is improved by using the two models, the loss caused by accumulation of foreign matters on the photovoltaic panel is avoided, the generated energy of a photovoltaic electric field is improved, the power generation safety of the photovoltaic power station can be enhanced, and meanwhile, the waste of manpower resources caused by manual inspection can be avoided.
Drawings
Fig. 1 is a flowchart of an abnormal target detection method of a photovoltaic panel according to embodiment 1 of the present invention.
Fig. 2 is a schematic view of first position information of a photovoltaic panel according to embodiment 1 of the present invention.
Fig. 3 is a schematic diagram of an abnormal target detection effect of the photovoltaic panel according to embodiment 1 of the present invention.
Fig. 4 is a flowchart of an abnormal target detection method of a photovoltaic panel according to embodiment 2 of the present invention.
Fig. 5 is a flowchart of steps S10a to S10b of the abnormal target detection method of the photovoltaic panel according to embodiment 2 of the present invention.
Fig. 6 is a schematic block diagram of an abnormal object detection system of a photovoltaic panel according to embodiment 3 of the present invention.
Fig. 7 is a block schematic diagram of an abnormal object detection system of a photovoltaic panel according to embodiment 4 of the present invention.
Fig. 8 is a schematic diagram of a hardware structure of an electronic device according to embodiment 5 of the present invention.
Detailed Description
The invention is further illustrated by the following examples, which are not intended to limit the scope of the invention.
Example 1
The present embodiment provides a method for detecting an abnormal target of a photovoltaic panel, as shown in fig. 1, the method for detecting an abnormal target of the present embodiment includes:
and S10, acquiring a photovoltaic panel image in the photovoltaic power station.
In another embodiment, the photovoltaic panel image may adopt both the visible light image and the infrared light image, so that this embodiment is not limited by the type of the image, and has a wide application range. Of course, other types of images can also be used as the photovoltaic panel image, and are not described herein. In addition, after the photovoltaic panel image is acquired, the image can be preprocessed, and the preprocessing method includes but is not limited to a series of preprocessing methods such as deformity correction and contrast enhancement.
S20, inputting the photovoltaic panel image into the segmentation model to obtain photovoltaic panel segmentation data; the photovoltaic panel segmentation data comprise photovoltaic panel image information with background removed and first position information corresponding to each photovoltaic panel.
It should be noted that, for the establishment of the coordinates, the central point of the image of the photovoltaic panel may be selected as the origin of coordinates, the horizontal direction is the horizontal axis, and the vertical direction is the vertical axis to establish the coordinate system, or any one of the corners of the image of the photovoltaic panel may be selected as the origin of coordinates, the horizontal direction is the horizontal axis, and the vertical direction is the vertical axis to establish the coordinate system.
Specifically, the first position information is corresponding position information of each photovoltaic panel in the photovoltaic panel image, as shown in fig. 2, the first position information may be a coordinate position a of a center point of the photovoltaic panel on the photovoltaic panel image, and the first position information may also be coordinate positions of four corners of the photovoltaic panel on the photovoltaic panel image (not shown). In addition, some external environment interferences such as a background can be removed by inputting the photovoltaic panel image into the segmentation model, namely, the same-size image only containing the photovoltaic panel area is obtained by setting the pixels belonging to the background in the photovoltaic panel image to be black and keeping the pixel values of the part of the pixels belonging to the photovoltaic panel area unchanged, so that the interference is avoided to improve the detection accuracy.
S30, identifying an abnormal target in the photovoltaic panel segmentation data by using the target detection model and outputting second position information of the abnormal target; the detection network of the target detection model comprises at least two levels of anchor frames, each level comprises a plurality of different anchor frames, and different levels correspond to different anchor frame size ranges. Alternatively, the identified abnormal target may be identified on the image information of the photovoltaic panel by using a rectangular frame or a circular frame, and thus, the second position information may be the coordinates of the center point of the rectangular or circular frame.
Specifically, the detection network of the target detection model can set anchor frames in multiple levels according to detection requirements, in the implementation process, if abnormal target conditions of the photovoltaic power station need to be detected quickly, the number of the levels of the anchor frames can be reduced, large-size anchor frames are preferentially selected to detect large targets on a photovoltaic panel in the photovoltaic power station, if abnormal target conditions of the photovoltaic power station need to be detected accurately, the number of the levels of the anchor frames can be increased, and the anchor frames in multiple size ranges are set in sequence to detect each target on the photovoltaic panel of the photovoltaic power station. For example, the anchor frame has a basic size of 8, three aspect ratios of 0.5, 1.0, 2.0 can be used, the number of feature scales for detection is 5, and the step sizes are 4, 8, 16, 32, 64, respectively.
And S40, determining the abnormal photovoltaic panel where the abnormal target is located based on the first position information and the second position information.
In this embodiment, as shown in fig. 3, it may be determined whether a distance between the second position information of the abnormal target and the first position information of the photovoltaic panel is smaller than a preset range, and if the distance is smaller than the preset range, it may be determined that the photovoltaic panel where the abnormal target is located is the abnormal photovoltaic panel, specifically, it may be determined whether a distance between a center point of the abnormal target and a center point of a certain photovoltaic panel is smaller than the preset range, it may also be determined whether the center point of the abnormal target is within an area formed by coordinates of four corners of a certain photovoltaic panel, it may also be determined whether an identification rectangular frame or a circular frame of the abnormal target is within an area formed by coordinates of four corners of a certain photovoltaic panel, it may be determined whether the photovoltaic panel has the abnormal target by various determination methods, and thus accuracy of detection may be improved.
In the embodiment, the photovoltaic panel image in the photovoltaic power station is acquired, the photovoltaic panel image is input into the segmentation model to acquire the photovoltaic panel segmentation data, wherein the photovoltaic panel segmentation data comprises the photovoltaic panel image information for removing the background and the first position information corresponding to each photovoltaic panel, the abnormal target in the photovoltaic panel segmentation data is identified by using the target detection model, the second position information of the abnormal target is output, and the abnormal photovoltaic panel where the abnormal target is located is determined based on the first position information and the second position information, so that the abnormal target can be identified quickly and accurately after background interference is removed, loss caused by accumulation of foreign matters in the photovoltaic panel is avoided, the power generation capacity of the photovoltaic power station is improved, the safety of power generation of the photovoltaic power station can be enhanced, and meanwhile, waste of manpower resources caused by manual inspection can be avoided.
Example 2
As shown in fig. 4, the method for detecting an abnormal object of a photovoltaic panel of the present embodiment is a further improvement of embodiment 1, and specifically:
the abnormal target detection method of the present embodiment further includes:
and S50, acquiring the number information of each photovoltaic panel in the photovoltaic panel image and the positioning information corresponding to the photovoltaic panel image shot by the shooting equipment in the photovoltaic power station.
Specifically, after the photovoltaic panel image is obtained, the photovoltaic panels may be numbered according to actual distribution of the photovoltaic panels in the photovoltaic power station, and the photovoltaic panels at the upper left corner of the photovoltaic panel image are sequentially numbered, where the numbering may be a 1-An, or roman numerals, and of course, the above numbering is only An example, and other numbering rules may also be used.
And S60, determining coordinate information of the abnormal photovoltaic panel based on the number information, the first position information, the second position information and the positioning information.
Specifically, after the abnormal photovoltaic panel is determined, the number information of the abnormal photovoltaic panel on the photovoltaic panel image and the positioning information corresponding to the shooting equipment when the photovoltaic panel image is shot can be combined, the actual coordinate information of the abnormal photovoltaic panel on the photovoltaic power station is further determined, the abnormal photovoltaic panel can be conveniently and quickly positioned by workers so as to be convenient for subsequent disposal, meanwhile, the abnormal target distribution report of the whole station can be generated by combining the modeling information of the photovoltaic power station in advance, and the health state of the photovoltaic panel can be conveniently evaluated and the evaluation result can be conveniently determined.
In an alternative embodiment, step S50 includes:
s51, acquiring device positioning information when the shooting device shoots the photovoltaic panel image in the photovoltaic power station, and taking the device positioning information as the positioning information of the photovoltaic panel image.
In this embodiment, the shooting device may include, but is not limited to, at least one or more of an aerial unmanned aerial vehicle and an aerial hot air balloon, and then the shooting device may carry an infrared camera and/or a visible light camera, and record positioning information of the device when shooting a photovoltaic panel image in a photovoltaic power station, where the positioning information may be corresponding longitude and latitude information of the device.
In an alternative embodiment, an aerial image of a photovoltaic panel within a photovoltaic power plant may be acquired as the photovoltaic panel image.
In an optional implementation mode, the segmentation model is obtained by inputting historical image training data into a Mask-RCNN model for training; the historical image training data is a historical image subjected to outline labeling of the photovoltaic panel region.
In an optional implementation mode, the target detection model is obtained by inputting historical abnormal image training data into a RetinaNet model for training; the historical abnormal image training data are historical abnormal images which are marked with abnormal targets. In order to prevent the result output by the model from being over-fitted, the training optimizer adopts a Stochastic Gradient Descent (SGD), and may further set a hyper-parameter, and may set a learning rate lr equal to 0.002, an momentum coefficient momentum equal to 0.9, and a weight attenuation coefficient of 0.0001, for example only.
In an alternative embodiment, as shown in fig. 5, before step S10, the abnormal target detecting method further includes:
s10a, making a shooting route according to the distribution condition of photovoltaic panels in the photovoltaic power station; the distribution condition comprises at least one of number distribution and area distribution of the photovoltaic panels;
and S10b, shooting along the shooting route by the remote control shooting device.
In particular, the photographing apparatus may be remotely controlled by means of a ground station.
In the embodiment, the number information of each photovoltaic panel in the photovoltaic panel image and the positioning information corresponding to the photovoltaic panel image in the photovoltaic power station are shot by the shooting device, and then the coordinate information of the abnormal photovoltaic panel is determined based on the number information, the first position information, the second position information and the positioning information, the coordinate information of the abnormal photovoltaic panel can be quickly positioned, a worker can conveniently and quickly position the abnormal photovoltaic panel so as to facilitate subsequent treatment, the inspection and maintenance efficiency of the photovoltaic power station is improved, and the performance and safety of the photovoltaic power station are ensured.
Example 3
The present embodiment provides an abnormal target detection system of a photovoltaic panel, as shown in fig. 6, the abnormal target detection system includes:
the image acquisition module 1 is used for acquiring photovoltaic panel images in the photovoltaic power station.
In one embodiment, the photovoltaic panel image may be a visible light image or an infrared light image, and in another embodiment, the photovoltaic panel image may be a visible light image and an infrared light image at the same time. Of course, other types of images can also be used as the photovoltaic panel image, and are not described herein. In addition, after the photovoltaic panel image is acquired, the image can be preprocessed, including but not limited to a series of preprocessing methods such as deformity correction and contrast enhancement.
The segmentation module 2 is used for inputting the photovoltaic panel image into a segmentation model to obtain photovoltaic panel segmentation data; the photovoltaic panel segmentation data comprise photovoltaic panel image information with background removed and first position information corresponding to each photovoltaic panel.
It should be noted that, for the establishment of the coordinates, the central point of the image of the photovoltaic panel may be selected as the origin of coordinates, the horizontal direction is the horizontal axis, and the vertical direction is the vertical axis to establish the coordinate system, or any one of the corners of the image of the photovoltaic panel may be selected as the origin of coordinates, the horizontal direction is the horizontal axis, and the vertical direction is the vertical axis to establish the coordinate system.
Specifically, the first position information is corresponding position information of each photovoltaic panel in the photovoltaic panel image, and as shown in fig. 2, the first position information may be a coordinate position (not shown) of a center point of the photovoltaic panel on the photovoltaic panel image, and the first position information may also be coordinate positions of four corners of the photovoltaic panel on the photovoltaic panel image. In addition, some external environment interferences such as a background can be removed by inputting the photovoltaic panel image into the segmentation model, namely, the same-size image only containing the photovoltaic panel area is obtained by setting the pixels belonging to the background in the photovoltaic panel image to be black and keeping the pixel values of the part of the pixels belonging to the photovoltaic panel area unchanged, so that the interference is avoided to improve the detection accuracy.
The identification module 3 is used for identifying an abnormal target in the photovoltaic panel segmentation data by using the target detection model and outputting second position information of the abnormal target; the detection network of the target detection model comprises at least two levels of anchor frames, each level comprises a plurality of different anchor frames, and different levels correspond to different anchor frame size ranges. Alternatively, the identified abnormal target may be identified on the image information of the photovoltaic panel by using a rectangular frame or a circular frame, and thus, the second position information may be the coordinates of the center point of the rectangular or circular frame.
Specifically, the detection network of the target detection model can set anchor frames in multiple levels according to detection requirements, in the implementation process, if abnormal target conditions of the photovoltaic power station need to be detected quickly, the number of the levels of the anchor frames can be reduced, large-size anchor frames are preferentially selected to detect large targets on a photovoltaic panel in the photovoltaic power station, if abnormal target conditions of the photovoltaic power station need to be detected accurately, the number of the levels of the anchor frames can be increased, and the anchor frames in multiple size ranges are set in sequence to detect each target on the photovoltaic panel of the photovoltaic power station. For example, the anchor frame has a basic size of 8, three aspect ratios of 0.5, 1.0, 2.0 can be used, the number of feature scales for detection is 5, and the step sizes are 4, 8, 16, 32, 64, respectively.
And the first determining module 4 is used for determining the abnormal photovoltaic panel where the abnormal target is located based on the first position information and the second position information.
In this embodiment, whether the distance between the second position information of the abnormal target and the first position information of the photovoltaic panel is smaller than a preset range or not can be determined, and if the distance is smaller than the preset range, the photovoltaic panel where the abnormal target is located can be determined to be the abnormal photovoltaic panel.
In the embodiment, the image acquisition module acquires the photovoltaic panel image in the photovoltaic power station, the segmentation module inputs the photovoltaic panel image into the segmentation model to acquire the photovoltaic panel segmentation data, wherein the photovoltaic panel segmentation data comprises photovoltaic panel image information of the removed background and first position information corresponding to each photovoltaic panel, identifying an abnormal target in the photovoltaic panel segmentation data by using the target detection model through the identification module and outputting second position information of the abnormal target, the abnormal photovoltaic panel where the abnormal target is located is determined through the first determining module based on the first position information and the second position information, so that the abnormal target can be quickly and accurately identified after background interference is removed in advance, and then can strengthen the security of photovoltaic power plant electricity generation, avoid because of the photovoltaic board because of the accumulation of foreign matter problem, the loss that causes improves the generated energy of photovoltaic electric field, can also avoid the artifical waste of patrolling and examining the manpower resources and causing simultaneously.
Example 4
As shown in fig. 7, the abnormal object detection system of the photovoltaic panel of the present embodiment is a further improvement of embodiment 3, specifically:
the abnormal object detection system of the present embodiment further includes:
and the information acquisition module 5 is used for acquiring the number information of each photovoltaic panel in the photovoltaic panel image and the positioning information corresponding to the photovoltaic panel image shot by the shooting equipment in the photovoltaic power station.
Specifically, after the photovoltaic panel image is obtained, the photovoltaic panels may be numbered according to actual distribution of the photovoltaic panels in the photovoltaic power station, and the photovoltaic panels at the upper left corner of the photovoltaic panel image are sequentially numbered, where the numbering may be a 1-An, or roman numerals, and of course, the above numbering is only An example, and other numbering rules may also be used.
And the second determining module 6 is used for determining the coordinate information of the abnormal photovoltaic panel based on the number information, the first position information, the second position information and the positioning information.
Specifically, after the abnormal photovoltaic panel is determined, the number information of the abnormal photovoltaic panel on the photovoltaic panel image and the positioning information corresponding to the shooting equipment when the photovoltaic panel image is shot can be combined, the actual coordinate information of the abnormal photovoltaic panel on the photovoltaic power station is further determined, workers can conveniently and quickly position the abnormal photovoltaic panel for subsequent treatment, meanwhile, the abnormal target distribution report of the whole station can be generated by combining the modeling information of the photovoltaic power station in advance, and the health state of the photovoltaic panel can be conveniently evaluated and the evaluation result can be determined.
In an optional implementation manner, the information obtaining module 5 is further configured to obtain device positioning information when the shooting device shoots a photovoltaic panel image in the photovoltaic power station, and use the device positioning information as positioning information of the photovoltaic panel image.
In this embodiment, the shooting device may include, but is not limited to, at least one or more of an aerial unmanned aerial vehicle and an aerial hot air balloon, and the shooting device may further carry an infrared camera and/or a visible light camera, and record positioning information of the device when shooting an image of a photovoltaic panel in a photovoltaic power station, where the positioning information may be corresponding longitude and latitude information of the device.
In an alternative embodiment, the image acquisition module 1 is further configured to acquire an aerial image of a photovoltaic panel within the photovoltaic power plant as the photovoltaic panel image.
In an optional implementation mode, the segmentation model is obtained by inputting historical image training data into a Mask-RCNN model for training; the historical image training data are historical images subjected to outline labeling of the photovoltaic panel area.
In an optional implementation mode, the target detection model is obtained by inputting historical abnormal image training data into a RetinaNet model for training; the historical abnormal image training data is a historical abnormal image which is marked with an abnormal target. In order to prevent the result output by the model from being over-fitted, the training optimizer adopts a Stochastic Gradient Descent (SGD), and may further set a hyper-parameter, and may set a learning rate lr equal to 0.002, an momentum coefficient momentum equal to 0.9, and a weight attenuation coefficient of 0.0001, for example only.
In an alternative embodiment, the system for detecting an abnormal object further comprises:
the shooting module 7 is used for making a shooting route according to the distribution condition of photovoltaic panels in the photovoltaic power station; the distribution condition comprises at least one of number distribution and area distribution of the photovoltaic panels;
the remote control shooting device shoots along the shooting route.
In particular, the photographing apparatus may be remotely controlled by means of a ground station.
In the embodiment, the number information of each photovoltaic panel in the photovoltaic panel image and the positioning information corresponding to the photovoltaic panel image in the photovoltaic power station are shot by the shooting equipment are obtained through the information acquisition module, the second determination module determines the coordinate information of the abnormal photovoltaic panel based on the number information, the first position information, the second position information and the positioning information, the coordinate information of the abnormal photovoltaic panel can be rapidly positioned, workers can rapidly position the abnormal photovoltaic panel to facilitate subsequent treatment, the inspection and maintenance efficiency of the photovoltaic power station is improved, and the performance and safety of the photovoltaic power station are guaranteed.
Example 5
Fig. 8 is a schematic structural diagram of an electronic device according to embodiment 5 of the present invention. The electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the abnormal object detection method of the photovoltaic panel of embodiment 1 or embodiment 2 when executing the program. The electronic device 30 shown in fig. 8 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiment of the present invention.
As shown in fig. 8, the electronic device 30 may be embodied in the form of a general purpose computing device, which may be, for example, a server device. The components of the electronic device 30 may include, but are not limited to: the at least one processor 31, the at least one memory 32, and a bus 33 connecting the various system components (including the memory 32 and the processor 31).
The bus 33 includes a data bus, an address bus, and a control bus.
The memory 32 may include volatile memory, such as Random Access Memory (RAM)321 and/or cache memory 322, and may further include Read Only Memory (ROM) 323.
The processor 31 executes various functional applications and data processing, such as the abnormal object detection method of the photovoltaic panel of embodiment 1 or embodiment 2 of the present invention, by running the computer program stored in the memory 32.
The electronic device 30 may also communicate with one or more external devices 34 (e.g., a keyboard, a pointing device, etc.). Such communication may be through input/output (I/O) interfaces 35. Also, model-generating device 30 may also communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet) via network adapter 36. As shown, network adapter 36 communicates with the other modules of model-generating device 30 via bus 33. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the model-generating device 30, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems, etc.
It should be noted that although in the above detailed description several units/modules or sub-units/modules of the electronic device are mentioned, such a division is merely exemplary and not mandatory. Indeed, the features and functionality of two or more of the units/modules described above may be embodied in one unit/module according to embodiments of the invention. Conversely, the features and functions of one unit/module described above may be further divided into embodiments by a plurality of units/modules.
Example 6
The present embodiment provides a computer-readable storage medium on which a computer program is stored, the program, when executed by a processor, implementing the abnormal object detection method of the photovoltaic panel of embodiment 1 or embodiment 2.
More specific examples, among others, that the readable storage medium may employ may include, but are not limited to: a portable disk, a hard disk, random access memory, read only memory, erasable programmable read only memory, optical storage device, magnetic storage device, or any suitable combination of the foregoing.
In a possible implementation, the invention can also be implemented in the form of a program product comprising program code for causing a terminal device to carry out a method of detecting an anomalous target of a photovoltaic panel implementing embodiment 1 or embodiment 2, when said program product is run on said terminal device.
Where program code for carrying out the invention is written in any combination of one or more programming languages, the program code may be executed entirely on the user device, partly on the user device, as a stand-alone software package, partly on the user device and partly on a remote device or entirely on the remote device.
While specific embodiments of the invention have been described above, it will be appreciated by those skilled in the art that this is by way of example only, and that the scope of the invention is defined by the appended claims. Various changes and modifications to these embodiments may be made by those skilled in the art without departing from the spirit and scope of the invention, and these changes and modifications are within the scope of the invention.
Claims (10)
1. An abnormal target detection method of a photovoltaic panel, characterized by comprising:
acquiring a photovoltaic panel image in a photovoltaic power station;
inputting the photovoltaic panel image into a segmentation model to obtain photovoltaic panel segmentation data; the photovoltaic panel segmentation data comprise photovoltaic panel image information with a background removed and first position information corresponding to each photovoltaic panel;
identifying an abnormal target in the photovoltaic panel segmentation data by using a target detection model and outputting second position information of the abnormal target; the detection network of the target detection model comprises at least two levels of anchor frames, each level comprises a plurality of different anchor frames, and different levels correspond to different anchor frame size ranges;
and determining an abnormal photovoltaic panel where the abnormal target is located based on the first position information and the second position information.
2. The abnormal object detection method of a photovoltaic panel according to claim 1, further comprising:
acquiring the number information of each photovoltaic panel in the photovoltaic panel image and the positioning information corresponding to the photovoltaic panel image shot by shooting equipment in the photovoltaic power station;
determining coordinate information of the abnormal photovoltaic panel based on the number information, the first position information, the second position information, and the positioning information.
3. The method for detecting the abnormal target of the photovoltaic panel as claimed in claim 1, wherein the segmentation model is obtained by inputting historical image training data into a Mask-RCNN model for training; the historical image training data is a historical image subjected to outline labeling of the photovoltaic panel region.
4. The abnormal target detection method of a photovoltaic panel according to claim 1, wherein the target detection model is obtained by inputting historical abnormal image training data into a RetinaNet model for training; and the historical abnormal image training data is a historical abnormal image which is subjected to abnormal target labeling.
5. The method for detecting the abnormal target of the photovoltaic panel as claimed in claim 2, wherein the step of obtaining the positioning information corresponding to the image of the photovoltaic panel shot by the shooting device in the photovoltaic power station comprises:
acquiring equipment positioning information when shooting equipment shoots the photovoltaic panel image in the photovoltaic power station, and taking the equipment positioning information as the positioning information of the photovoltaic panel image.
6. The method of claim 1, wherein the step of obtaining an image of a photovoltaic panel within a photovoltaic power plant comprises:
and acquiring a high-altitude image of the photovoltaic panel in the photovoltaic power station as the photovoltaic panel image.
7. The method of detecting an anomalous target in a photovoltaic panel as in claim 1, wherein prior to said step of obtaining an image of a photovoltaic panel within a photovoltaic power plant, the method of detecting an anomalous target further comprises:
a shooting route is formulated according to the distribution condition of the photovoltaic panels in the photovoltaic power station; wherein the distribution condition comprises at least one of quantity distribution and area distribution of the photovoltaic panels;
and remotely controlling the shooting equipment to shoot along the shooting route.
8. An anomalous target detection system for a photovoltaic panel, said anomalous target detection system comprising:
the image acquisition module is used for acquiring a photovoltaic panel image in the photovoltaic power station;
the segmentation module is used for inputting the photovoltaic panel image into a segmentation model so as to obtain photovoltaic panel segmentation data; the photovoltaic panel segmentation data comprise photovoltaic panel image information with background removed and first position information corresponding to each photovoltaic panel;
the identification module is used for identifying an abnormal target in the photovoltaic panel segmentation data by using a target detection model and outputting second position information of the abnormal target; the detection network of the target detection model comprises at least two levels of anchor frames, each level comprises a plurality of different anchor frames, and different levels correspond to different anchor frame size ranges;
and the first determining module is used for determining the abnormal photovoltaic panel where the abnormal target is located based on the first position information and the second position information.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method for detecting an anomalous target in a photovoltaic panel as claimed in any one of claims 1 to 7 when the computer program is executed by the processor.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out a method for anomalous target detection of a photovoltaic panel as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210173145.6A CN114529537A (en) | 2022-02-24 | 2022-02-24 | Abnormal target detection method, system, equipment and medium for photovoltaic panel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210173145.6A CN114529537A (en) | 2022-02-24 | 2022-02-24 | Abnormal target detection method, system, equipment and medium for photovoltaic panel |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114529537A true CN114529537A (en) | 2022-05-24 |
Family
ID=81624272
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210173145.6A Pending CN114529537A (en) | 2022-02-24 | 2022-02-24 | Abnormal target detection method, system, equipment and medium for photovoltaic panel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114529537A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116205892A (en) * | 2023-03-13 | 2023-06-02 | 北京智盟信通科技有限公司 | Method and device for evaluating working condition of photovoltaic string |
| CN118115821A (en) * | 2024-04-28 | 2024-05-31 | 国网辽宁省电力有限公司辽阳供电公司 | Photovoltaic panel fault identification method based on image feature extraction |
| CN118379291A (en) * | 2024-06-24 | 2024-07-23 | 尚特杰电力科技有限公司 | Photovoltaic panel image detection method, storage medium and electronic device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111144324A (en) * | 2019-12-28 | 2020-05-12 | 西安因诺航空科技有限公司 | System and method for analyzing and managing abnormal target of photovoltaic panel during unmanned aerial vehicle inspection |
| CN111339882A (en) * | 2020-02-19 | 2020-06-26 | 山东大学 | Power transmission line hidden danger detection method based on example segmentation |
| CN113393459A (en) * | 2021-08-09 | 2021-09-14 | 旻投电力发展有限公司 | Infrared image photovoltaic module visual identification method based on example segmentation |
-
2022
- 2022-02-24 CN CN202210173145.6A patent/CN114529537A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111144324A (en) * | 2019-12-28 | 2020-05-12 | 西安因诺航空科技有限公司 | System and method for analyzing and managing abnormal target of photovoltaic panel during unmanned aerial vehicle inspection |
| CN111339882A (en) * | 2020-02-19 | 2020-06-26 | 山东大学 | Power transmission line hidden danger detection method based on example segmentation |
| CN113393459A (en) * | 2021-08-09 | 2021-09-14 | 旻投电力发展有限公司 | Infrared image photovoltaic module visual identification method based on example segmentation |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116205892A (en) * | 2023-03-13 | 2023-06-02 | 北京智盟信通科技有限公司 | Method and device for evaluating working condition of photovoltaic string |
| CN116205892B (en) * | 2023-03-13 | 2024-04-05 | 北京智盟信通科技有限公司 | Method and device for evaluating working condition of photovoltaic string |
| CN118115821A (en) * | 2024-04-28 | 2024-05-31 | 国网辽宁省电力有限公司辽阳供电公司 | Photovoltaic panel fault identification method based on image feature extraction |
| CN118115821B (en) * | 2024-04-28 | 2024-08-06 | 国网辽宁省电力有限公司辽阳供电公司 | Photovoltaic panel fault identification method based on image feature extraction |
| CN118379291A (en) * | 2024-06-24 | 2024-07-23 | 尚特杰电力科技有限公司 | Photovoltaic panel image detection method, storage medium and electronic device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114529537A (en) | Abnormal target detection method, system, equipment and medium for photovoltaic panel | |
| CN112633535B (en) | Intelligent inspection method and system for photovoltaic power station based on unmanned aerial vehicle image | |
| CN108365557B (en) | Method and system for fine routing inspection of power transmission line by unmanned aerial vehicle | |
| WO2021027366A1 (en) | Method, device and system for detecting cell sheet of photovoltaic power station | |
| CN104077773A (en) | Image edge detection method, and image target identification method and device | |
| CN111446920A (en) | Photovoltaic power station monitoring method, device and system | |
| CN111740699A (en) | Photovoltaic panel fault detection and identification method and device and unmanned aerial vehicle | |
| CN113433971A (en) | Method, device, equipment and storage medium for acquiring data of high-rise building exterior wall | |
| CN111598942A (en) | Method and system for automatically positioning electric power facility instrument | |
| CN111539355A (en) | Photovoltaic panel foreign matter detection system and detection method based on deep neural network | |
| CN112449147B (en) | Video cluster monitoring system of photovoltaic power station and image processing method thereof | |
| CN115424141A (en) | Photovoltaic installed capacity calculation method and device, electronic equipment and medium | |
| CN116797944A (en) | Detection method and system for identifying cleanliness of photovoltaic panel based on unmanned aerial vehicle image | |
| CN116993665A (en) | Intelligent detection method for construction progress of construction engineering working face based on computer vision | |
| Özer et al. | An approach based on deep learning methods to detect the condition of solar panels in solar power plants | |
| CN112184903B (en) | Method, device, equipment and medium for detecting high-voltage line tree obstacle risk points | |
| CN116758060B (en) | Vertical basket of flowers visual detection system of battery piece | |
| CN116344378B (en) | Intelligent detection system for photovoltaic panel production and detection method thereof | |
| CN115311589B (en) | Hidden danger processing method and equipment for lighting building | |
| CN112001964A (en) | Flood evolution process inundation range measuring method based on deep learning | |
| CN114078234B (en) | Detection method, system, storage medium and equipment for power supply area construction process | |
| CN113284103B (en) | Substation equipment defect online detection method based on space transformation fast R-CNN model | |
| CN115082385A (en) | Photovoltaic power station defect early warning method and device based on inclination angle of photovoltaic module | |
| CN114841932A (en) | Foreign matter detection method, system, equipment and medium for photovoltaic panel of photovoltaic power station | |
| CN112927222A (en) | Method for realizing multi-type photovoltaic array hot spot detection based on hybrid improved Faster R-CNN |
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 |