CN113554611B - Insulator self-explosion defect detection method and device, terminal and storage medium - Google Patents
Insulator self-explosion defect detection method and device, terminal and storage medium Download PDFInfo
- Publication number
- CN113554611B CN113554611B CN202110816193.8A CN202110816193A CN113554611B CN 113554611 B CN113554611 B CN 113554611B CN 202110816193 A CN202110816193 A CN 202110816193A CN 113554611 B CN113554611 B CN 113554611B
- Authority
- CN
- China
- Prior art keywords
- self
- remote sensing
- image data
- insulator
- convolution
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 93
- 230000007547 defect Effects 0.000 title claims abstract description 55
- 239000012212 insulator Substances 0.000 title claims abstract description 49
- 238000004880 explosion Methods 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 claims abstract description 66
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 230000008569 process Effects 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 14
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000001629 suppression Effects 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000012217 deletion Methods 0.000 description 3
- 230000037430 deletion Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012805 post-processing Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009194 climbing Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000010339 dilation Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000004804 winding Methods 0.000 description 1
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
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- 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
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method, a device, a terminal and a storage medium for detecting self-explosion defects of insulators, wherein the method comprises the following steps: acquiring insulator remote sensing image data in a target area, and preprocessing the remote sensing image data to acquire preprocessed remote sensing image data; constructing a self-explosion defect detection model, wherein the detection model comprises the following steps: the system comprises a basic network unit, a mixed cavity convolution unit and a detection unit; inputting the preprocessed remote sensing image data into the self-explosion defect detection model to obtain a defect detection result. The invention can improve the detection efficiency of the insulator self-explosion defect and the detection accuracy.
Description
Technical Field
The present invention relates to the field of image recognition technologies, and in particular, to a method, an apparatus, a terminal, and a storage medium for detecting an insulator self-explosion defect.
Background
The overhead transmission line connects the power station, the transformer substation and the user to form a transmission and distribution network. In transmission lines, insulators are one of the necessary devices, which play an important role in electrical isolation and mechanical support. The failure of the insulator directly threatens the stability and safety of the transmission line. It is counted that the accident caused by the insulator fault is the highest in proportion to the power system fault. Therefore, it is important to detect the defects of the insulator timely and intelligently. In recent years, transmission line inspection has generally used conventional methods such as manual patrol, manned helicopter patrol, and climbing robot patrol. However, the traditional inspection method is low in efficiency due to the wide distribution of the power transmission lines and the complex terrain of the area. Unmanned Aerial Vehicle (UAV) inspection has advantages such as with low costs, consuming less than above traditional method. With the development of image processing technology, insulator defect detection based on unmanned aerial vehicle remote sensing images is increasingly valued by power enterprises. However, since the insulator self-explosion defect area in the image is small and the background is complex, the rapid and intelligent detection of the insulator self-explosion defect is a challenging task.
At present, the existing insulator detection method has the following problems:
1. the existing target detection algorithm can be divided into a first-order detection method and a second-order detection method. The first-order detection method has high instantaneity, but has low detection precision; the second order detection method has higher detection accuracy than the first order detection method, but is difficult to train and has low detection speed. Both methods cannot balance the detection speed and accuracy.
2. The occupied area proportion of the insulator defect area in the remote sensing image is small, and the existing method is easy to cause the loss of the characteristic information of a small target in the characteristic extraction process.
3. The existing target detection algorithm is easy to cause the reduction of the target detection accuracy when insulators are overlapped and shielded in the remote sensing image.
Disclosure of Invention
The purpose of the invention is that: the method, the device, the terminal and the storage medium for detecting the self-explosion defects of the insulator are provided, so that the detection efficiency of the self-explosion defects of the insulator can be improved, and the detection accuracy can be improved.
In order to achieve the above object, the present invention provides a method for detecting an insulator self-explosion defect, comprising:
acquiring insulator remote sensing image data in a target area, and preprocessing the remote sensing image data to acquire preprocessed remote sensing image data;
constructing a self-explosion defect detection model, wherein the detection model comprises the following steps: the system comprises a basic network unit, a mixed cavity convolution unit and a detection unit;
inputting the preprocessed remote sensing image data into the self-explosion defect detection model to obtain a defect detection result.
Further, the preprocessing includes: graying processing, geometric transformation processing, and image enhancement processing.
Further, the mixed cavity convolution unit adopts the following calculation formula:
f n =f k +(f k -1)*(D r -1)
wherein f k The convolution kernel size is the cavity convolution; f (f) n The equivalent convolution kernel size is the cavity convolution; d (D) r Is the expansion rate; l (L) m-1 Representing the receptive field of the upper layer; l (L) m Representing the receptive field of the current layer; s is(s) i Indicating the i-th layer convolution step before the current layer.
Further, the detection unit adopts the following calculation formula:
wherein S is i For confidence score, DIOU represents distance intersection ratio, M represents prediction frame with highest confidence score, b i For the predicted frame to be processed, D is a predicted frame set, IOU represents the merging ratio, b represents the predicted frame center point, b gt Representing the center point of a real frame, ρ 2 (b,b gt ) Representing a predicted frame center point b and a true frame center point b gt And c represents the diagonal distance of the minimum closure area that can contain both the predicted and real frames.
Further, the graying process includes: component method, maximum value method, average value method and weighted average method; the geometric transformation process includes: nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation; the image enhancement process includes: spatial domain methods and frequency domain methods.
The invention also provides an insulator self-explosion defect detection device, which comprises: the device comprises an acquisition module, a construction module and a detection module, wherein,
the acquisition module is used for acquiring insulator remote sensing image data in a target area, preprocessing the remote sensing image data and acquiring preprocessed remote sensing image data;
the construction module is configured to construct a self-explosion defect detection model, where the detection model includes: the system comprises a basic network unit, a mixed cavity convolution unit and a detection unit;
the detection module is used for inputting the preprocessed remote sensing image data into the self-explosion defect detection model to obtain a defect detection result.
Further, the mixed cavity convolution unit adopts the following calculation formula:
f n =f k +(f k -1)*(D r -1)
wherein f k The convolution kernel size is the cavity convolution; f (f) n The equivalent convolution kernel size is the cavity convolution; d (D) r Is the expansion rate; l (L) m-1 Representing the receptive field of the upper layer; l (L) m Representing the receptive field of the current layer; s is(s) i Indicating the i-th layer convolution step before the current layer.
Further, the detection unit adopts the following calculation formula:
wherein S is i For confidence score, DIOU represents distance intersection ratio, M represents prediction frame with highest confidence score, b i For the predicted frame to be processed, D is a predicted frame set, IOU represents the merging ratio, b represents the predicted frame center point, b gt Representing the center point of a real frame, ρ 2 (b,b gt ) Representing a predicted frame center point b and a true frame center point b gt And c represents the diagonal distance of the minimum closure area that can contain both the predicted and real frames.
The invention also provides a computer terminal device, comprising: one or more processors; a memory coupled to the processor for storing one or more programs; the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method for detecting an insulator self-explosion defect as set forth in any one of the preceding claims.
The invention also provides a computer readable storage medium having stored thereon a computer program, characterized in that the computer program, when executed by a processor, implements the method for detecting an insulator self-explosion defect as described in any one of the above.
Compared with the prior art, the insulator self-explosion defect detection method, the device, the terminal equipment and the computer-readable storage medium have the beneficial effects that:
the invention takes the YOLO v4 network as a basic network model to consider the detection precision and the speed. Meanwhile, in order to solve the problem that the occupied area proportion of the insulator defects in the remote sensing image is small and the detection capability of a network to small targets is low, the method proposes to introduce a mixed cavity convolution module to improve a network model. And the shallow layer features are processed by the mixed cavity convolution module and then fused with the deep layer features to obtain a new feature map. Compared with the original feature map, the new feature map contains more small target feature information, so that the detection capability of the network on the small targets is further improved. And finally, screening the predicted frame by adopting an improved Gaussian weighted non-extremum suppression method in the post-processing process of the detection result, so as to avoid the false deletion of the shielded insulator predicted frame caused by insulator superposition.
Drawings
FIG. 1 is a schematic flow chart of an insulator self-explosion defect detection method provided by the invention;
FIG. 2 is a schematic diagram of the structure of the improved YOLOv4 model network provided by the present invention;
FIG. 3 is a schematic illustration of three different expansion rate hole convolutions provided by the present invention;
FIG. 4 is a schematic diagram of the structure of a hybrid hole convolution provided by the present invention;
fig. 5 is a schematic structural diagram of an insulator self-explosion defect detection device provided by the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be understood that the step numbers used herein are for convenience of description only and are not limiting as to the order in which the steps are performed.
It is to be understood that the terminology used in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
The terms "comprises" and "comprising" indicate the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The term "and/or" refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
As shown in fig. 1, a method for detecting an insulator self-explosion defect according to an embodiment of the present invention includes:
s1, acquiring insulator remote sensing image data in a target area, and preprocessing the remote sensing image data to obtain preprocessed remote sensing image data;
specifically, insulator remote sensing image data in a target area are obtained, and the remote sensing image data are preprocessed to obtain preprocessed remote sensing image data; wherein the preprocessing comprises: graying processing, geometric transformation processing, and image enhancement processing.
The graying process includes: component method, maximum value method, average value method and weighted average method; the geometric transformation process includes: nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation; the image enhancement process includes: spatial domain methods and frequency domain methods.
S2, constructing a self-explosion defect detection model, wherein the detection model comprises the following steps: the system comprises a basic network unit, a mixed cavity convolution unit and a detection unit;
specifically, the method is based on the YOLO v4 as a basic network, a mixed cavity convolution module is added to the basic network to improve the network structure, the detection capability of a small target is further improved, and an improved Gaussian weighted non-extremum suppression method is adopted to screen a detected prediction frame to improve the target detection accuracy. The improved network structure is shown in fig. 2. The basic YOLO v4 consists of the backbone network CSPDarknet53, plus SPP, PANet, and three heads. As can be seen from the network model diagram, the PANet module performs tensor splicing after upsampling the three feature diagrams with different depths extracted by the dark net-53, so that multi-scale feature fusion is realized, and the detection capability of the network is improved. In order to obtain more characteristic information of a small target, the invention fuses a shallow characteristic image output by a second residual block in a main network with a characteristic image obtained by 8 times downsampling, and obtains more target characteristic information by the fusion mode of the deep and shallow characteristic images, thereby further improving the detection capability of the network, in particular to the small target.
It should be noted that, because two feature images with different receptive fields are directly fused and are unfavorable for detecting target information with the same size, the spatial resolution of the feature images is easily reduced by using convolution and pooling operations in the traditional method, so that the invention introduces a hybrid cavity convolution module to process shallow feature images, and the module systematically aggregates deep and shallow network target feature information by utilizing cavity convolution under the condition of not losing resolution, thereby retaining more small target feature information.
The hole convolution introduces a "dilation rate" parameter, which defines the distance between the convolution kernels, as compared to the normal convolution. In the hole convolution, the size of the receptive field can be changed by setting different expansion rates. FIG. 3 shows three kinds of hole convolutions with different expansion rates, black dots represent hole convolution kernels, the three hole convolution kernels are 3×3, but the expansion rates are 1,2 and 3 respectively, and gray areas represent receptive fields after different winding machine operations.
The cavity convolution receptive field is calculated by the following steps:
f n =f k +(f k -1)*(D r -1)
wherein f k The convolution kernel size is the cavity convolution; f (f) n The equivalent convolution kernel size is the cavity convolution; d (D) r Is the expansion rate; l (L) m-1 Representing the receptive field of the upper layer; l (L) m Representing the receptive field of the current layer; s is(s) i Indicating the i-th layer convolution step before the current layer.
However, the above-described hole convolution framework has a theoretical problem, which we call "checkerboard problem". The cavity convolution result obtained by a certain layer of the network is from an independent set of the upper layer, and is not interdependent, so that the convolution results of the layer are not related; secondly, because the cavity convolves the sparse sampling input signal, the information obtained by the remote convolution has no correlation. It is therefore proposed herein to build a hole convolution module in series using multiple hole convolutions (HDCs) of different expansion rates to alleviate checkerboard problems instead of a single hole convolution. The hybrid hole convolution module is shown in fig. 4, and is formed by connecting three hole convolution blocks with different expansion rates in series, wherein the expansion rates are sequentially increased, and the expansion rates correspond to three subgraphs from left to right in fig. 3. Each cavity convolution block consists of a cavity convolution layer, a batch normalization layer and a Mish activation function. The input of the HDC is a feature map output by a second residual block of the main network CSPDarknet53, and the output is fused with the feature map obtained by 8 times of downsampling of the original network, so that the fusion of shallow features and deep features is realized, and the target detection capability of the model is improved.
It should be noted that, in the target detection network, a non-extremum suppression method (NMS) is often used to remove duplicate prediction frames, and the most accurate frames are screened out as prediction results, so that error suppression often occurs in the conventional NMS method when the targets are blocked from each other, so that the prediction frames are deleted by mistake. Therefore, the invention adopts the improved Gaussian weighting NMS algorithm to reset the confidence coefficient of the predicted frame and then complete the screening of the predicted frame, thereby avoiding the direct false deletion when the overlap and overlap ratio of the predicted frame is higher than the IOU, and the Gaussian weighting is defined as follows:
firstly, a prediction frame M with the highest confidence score is found out from a preselect frame set B, and M is added to a final set D. Calculating the distance intersection ratio DIOU of the preselected frames in M and B, wherein the DIOU not only considers the IOU, but also considers the predicted frame center point B and the real frame center point B gt Distance ρ between 2 (b,b gt ). The gaussian weighted confidence scores for the other pre-selected boxes are then recalculated. The above steps are repeated until the pre-selected box in B is recalculated. And finally deleting the prediction frames with scores smaller than the threshold value in the set D, wherein the reserved pre-selected frames are the final detection result.
S3, inputting the preprocessed remote sensing image data into the self-explosion defect detection model to obtain a defect detection result.
Specifically, the preprocessed remote sensing image data is input into the self-explosion defect detection model, and a defect detection result is obtained.
In one embodiment of the present invention, the preprocessing includes: graying processing, geometric transformation processing, and image enhancement processing.
In one embodiment of the present invention, the hybrid hole convolution unit adopts the following calculation formula:
f n =f k +(f k -1)*(D r -1)
wherein f k The convolution kernel size is the cavity convolution; f (f) n The equivalent convolution kernel size is the cavity convolution; d (D) r Is the expansion rate; l (L) m-1 Representing the receptive field of the upper layer; l (L) m Representing the receptive field of the current layer; s is(s) i Indicating the i-th layer convolution step before the current layer.
In one embodiment of the present invention, the detection unit uses the following calculation formula:
wherein S is i For confidence score, DIOU represents distance intersection ratio, M represents prediction frame with highest confidence score, b i For the predicted frame to be processed, D is a predicted frame set, IOU represents the merging ratio, b represents the predicted frame center point, b gt Representing the center point of a real frame, ρ 2 (b,b gt ) Representing a predicted frame center point b and a true frame center point b gt And c represents the diagonal distance of the minimum closure area that can contain both the predicted and real frames.
In one embodiment of the present invention, the graying process includes: component method, maximum value method, average value method and weighted average method; the geometric transformation process includes: nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation; the image enhancement process includes: spatial domain methods and frequency domain methods.
Compared with the prior art, the insulator self-explosion defect detection method provided by the embodiment of the invention has the beneficial effects that:
the invention takes the YOLO v4 network as a basic network model to consider the detection precision and the speed. Meanwhile, in order to solve the problem that the occupied area proportion of the insulator defects in the remote sensing image is small and the detection capability of a network to small targets is low, the method proposes to introduce a mixed cavity convolution module to improve a network model. And the shallow layer features are processed by the mixed cavity convolution module and then fused with the deep layer features to obtain a new feature map. Compared with the original feature map, the new feature map contains more small target feature information, so that the detection capability of the network on the small targets is further improved. And finally, screening the predicted frame by adopting an improved Gaussian weighted non-extremum suppression method in the post-processing process of the detection result, so as to avoid the false deletion of the shielded insulator predicted frame caused by insulator superposition.
As shown in fig. 5, the present invention further provides an apparatus 200 for detecting an insulator self-explosion defect, including: an acquisition module 201, a construction module 202 and a detection module 203, wherein,
the acquiring module 201 is configured to acquire remote sensing image data of an insulator in a target area, and perform preprocessing on the remote sensing image data to obtain preprocessed remote sensing image data;
the construction module 202 is configured to construct a self-explosion defect detection model, where the detection model includes: the system comprises a basic network unit, a mixed cavity convolution unit and a detection unit;
the detection module 203 is configured to input the preprocessed remote sensing image data to the self-explosion defect detection model, and obtain a defect detection result.
In one embodiment of the present invention, the hybrid hole convolution unit adopts the following calculation formula:
f n =f k +(f k -1)*(D r -1)
wherein f k The convolution kernel size is the cavity convolution; f (f) n The equivalent convolution kernel size is the cavity convolution; d (D) r Is the expansion rate; l (L) m-1 Representing the receptive field of the upper layer; l (L) m Representing the receptive field of the current layer; s is(s) i Indicating the i-th layer convolution step before the current layer.
In one embodiment of the present invention, the detection unit uses the following calculation formula:
wherein S is i For confidence score, DIOU represents distance intersection ratio, M represents prediction frame with highest confidence score, b i For the predicted frame to be processed, D is a predicted frame set, IOU represents the merging ratio, b represents the predicted frame center point, b gt Representing the center point of a real frame, ρ 2 (b,b gt ) Representing a predicted frame center point b and a true frame center point b gt And c represents the diagonal distance of the minimum closure area that can contain both the predicted and real frames.
The invention also provides a computer terminal device, comprising: one or more processors; a memory coupled to the processor for storing one or more programs; the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method for detecting an insulator self-explosion defect as set forth in any one of the preceding claims.
It should be noted that the processor may be a central processing unit (CentralProcessingUnit, CPU), other general purpose processors, digital signal processors (DigitalSignalProcessor, DSP), application specific integrated circuits (ApplicationSpecificIntegratedCircuit, ASIC), off-the-shelf programmable gate arrays (Field-ProgrammableGateArray, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., or any conventional processor that is a control center of the terminal device and that connects various parts of the terminal device using various interfaces and lines.
The memory mainly includes a program storage area, which may store an operating system, an application program required for at least one function, and the like, and a data storage area, which may store related data and the like. In addition, the memory may be a high-speed random access memory, a nonvolatile memory such as a plug-in hard disk, a smart memory card (SmartMediaCard, SMC), a secure digital (SecureDigital, SD) card, a flash memory card (FlashCard), etc., or other volatile solid state memory devices.
It should be noted that the above-mentioned terminal device may include, but is not limited to, a processor, a memory, and those skilled in the art will understand that the above-mentioned terminal device is merely an example, and does not constitute limitation of the terminal device, and may include more or fewer components, or may combine some components, or different components.
The present invention also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method for detecting an insulator self-explosion defect as described in any one of the above.
It should be noted that the computer program may be divided into one or more modules/units (e.g., computer program), which are stored in the memory and executed by the processor to accomplish the present invention. The one or more modules/units may be a series of computer program instruction segments capable of performing the specified functions, which instruction segments are used for describing the execution of the computer program in the terminal device.
The foregoing embodiments have been provided for the purpose of illustrating the general principles of the present invention, and are not to be construed as limiting the scope of the invention. It should be noted that any modifications, equivalent substitutions, improvements, etc. made by those skilled in the art without departing from the spirit and principles of the present invention are intended to be included in the scope of the present invention.
Claims (6)
1. The method for detecting the self-explosion defect of the insulator is characterized by comprising the following steps of:
acquiring insulator remote sensing image data in a target area, and preprocessing the remote sensing image data to acquire preprocessed remote sensing image data;
constructing a self-explosion defect detection model, wherein the detection model comprises the following steps: the system comprises a basic network unit, a mixed cavity convolution unit and a detection unit;
inputting the preprocessed remote sensing image data into the self-explosion defect detection model to obtain a defect detection result;
the mixed cavity convolution unit adopts the following calculation formula:
wherein,the convolution kernel size is the cavity convolution; />Equivalent convolution kernel size for hole convolution;/>Is the expansion rate; />Representing the receptive field of the upper layer; />Representing the receptive field of the current layer; />Indicating the convolution step length of the ith layer before the current layer;
the detection unit adopts the following calculation formula:
wherein,for confidence score, DIOU represents distance cross ratio, M represents prediction frame with highest confidence score,/for confidence score>For a predicted frame to be processed, D is a predicted frame set, IOU represents the cross ratio, b represents the predicted frame center point, and +.>Representing the center point of the real frame,/-, and>representing the predicted frame center b and the true frame center +.>The distance between c represents the diagonal distance of the minimum closure area that can contain both the predicted and real frames;
the base network unit takes a YOLO v4 network as a base network.
2. The method for detecting an insulator self-explosion defect according to claim 1, wherein the preprocessing comprises: graying processing, geometric transformation processing, and image enhancement processing.
3. The method for detecting an insulator self-explosion defect according to claim 2, wherein the graying process includes: component method, maximum value method, average value method and weighted average method; the geometric transformation process includes: nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation; the image enhancement process includes: spatial domain methods and frequency domain methods.
4. An insulator self-explosion defect detection device, which is characterized by comprising: the device comprises an acquisition module, a construction module and a detection module, wherein,
the acquisition module is used for acquiring insulator remote sensing image data in a target area, preprocessing the remote sensing image data and acquiring preprocessed remote sensing image data;
the construction module is configured to construct a self-explosion defect detection model, where the detection model includes: the system comprises a basic network unit, a mixed cavity convolution unit and a detection unit;
the detection module is used for inputting the preprocessed remote sensing image data into the self-explosion defect detection model to obtain a defect detection result;
the mixed cavity convolution unit adopts the following calculation formula:
wherein,the convolution kernel size is the cavity convolution; />The equivalent convolution kernel size is the cavity convolution; />Is the expansion rate; />Representing the receptive field of the upper layer; />Representing the receptive field of the current layer; />Indicating the convolution step length of the ith layer before the current layer;
the detection unit adopts the following calculation formula:
wherein,for confidence score, DIOU represents distance cross ratio, M represents prediction frame with highest confidence score,/for confidence score>For a predicted frame to be processed, D is a predicted frame set, IOU represents the cross ratio, b represents the predicted frame center point, and +.>Representing the center point of the real frame,/-, and>representing the predicted frame center b and the true frame center +.>The distance between c represents the diagonal distance of the minimum closure area that can contain both the predicted and real frames;
the base network unit takes a YOLO v4 network as a base network.
5. A computer terminal device, comprising:
one or more processors;
a memory coupled to the processor for storing one or more programs;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of insulator self-explosion defect detection of any one of claims 1 to 3.
6. A computer-readable storage medium, on which a computer program is stored, characterized in that the computer program, when executed by a processor, implements the insulator self-explosion defect detection method as claimed in any one of claims 1 to 3.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110816193.8A CN113554611B (en) | 2021-07-19 | 2021-07-19 | Insulator self-explosion defect detection method and device, terminal and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110816193.8A CN113554611B (en) | 2021-07-19 | 2021-07-19 | Insulator self-explosion defect detection method and device, terminal and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113554611A CN113554611A (en) | 2021-10-26 |
CN113554611B true CN113554611B (en) | 2024-02-02 |
Family
ID=78103453
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110816193.8A Active CN113554611B (en) | 2021-07-19 | 2021-07-19 | Insulator self-explosion defect detection method and device, terminal and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113554611B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115641330B (en) * | 2022-11-17 | 2023-06-30 | 东莞市众嘉印刷有限公司 | Flexible circuit board defect detection method and system based on image processing |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
BRPI0700515A (en) * | 2007-02-14 | 2008-09-30 | Inst De Tecnologia Para O Dese | breakthrough transformer isolator breakthrough sensor |
WO2018086299A1 (en) * | 2016-11-11 | 2018-05-17 | 广东电网有限责任公司清远供电局 | Image processing-based insulator defect detection method and system |
CN110176001A (en) * | 2019-06-03 | 2019-08-27 | 浙江大学 | A kind of high iron catenary insulator breakage accurate positioning method based on Grad-CAM algorithm |
CN112184746A (en) * | 2020-08-27 | 2021-01-05 | 西北工业大学 | Transmission line insulator defect analysis method |
-
2021
- 2021-07-19 CN CN202110816193.8A patent/CN113554611B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
BRPI0700515A (en) * | 2007-02-14 | 2008-09-30 | Inst De Tecnologia Para O Dese | breakthrough transformer isolator breakthrough sensor |
WO2018086299A1 (en) * | 2016-11-11 | 2018-05-17 | 广东电网有限责任公司清远供电局 | Image processing-based insulator defect detection method and system |
CN110176001A (en) * | 2019-06-03 | 2019-08-27 | 浙江大学 | A kind of high iron catenary insulator breakage accurate positioning method based on Grad-CAM algorithm |
CN112184746A (en) * | 2020-08-27 | 2021-01-05 | 西北工业大学 | Transmission line insulator defect analysis method |
Also Published As
Publication number | Publication date |
---|---|
CN113554611A (en) | 2021-10-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110633661A (en) | Semantic segmentation fused remote sensing image target detection method | |
CN111627015B (en) | Small sample defect identification method, device, equipment and storage medium | |
CN108073898B (en) | Method, device and equipment for identifying human head area | |
CN113936256A (en) | Image target detection method, device, equipment and storage medium | |
CN110674688A (en) | Face recognition model acquisition method, system and medium for video monitoring scene | |
CN109712071B (en) | Unmanned aerial vehicle image splicing and positioning method based on track constraint | |
CN112348921A (en) | Mapping method and system based on visual semantic point cloud | |
CN113011338B (en) | Lane line detection method and system | |
WO2023082453A1 (en) | Image processing method and device | |
CN112257669A (en) | Pedestrian re-identification method and device and electronic equipment | |
CN112529827A (en) | Training method and device for remote sensing image fusion model | |
CN114758337A (en) | Semantic instance reconstruction method, device, equipment and medium | |
CN110852327A (en) | Image processing method, image processing device, electronic equipment and storage medium | |
CN114120272A (en) | Multi-supervision intelligent lane line semantic segmentation method fusing edge detection | |
CN113554611B (en) | Insulator self-explosion defect detection method and device, terminal and storage medium | |
CN116363037A (en) | Multi-mode image fusion method, device and equipment | |
CN115512258A (en) | Desensitization method and device for video image, terminal equipment and storage medium | |
CN113807184B (en) | Obstacle detection method and device, electronic equipment and automatic driving vehicle | |
CN113012107B (en) | Power grid defect detection method and system | |
CN114220087A (en) | License plate detection method, license plate detector and related equipment | |
CN115170807B (en) | Image segmentation and model training method, device, equipment and medium | |
CN116229419A (en) | Pedestrian detection method and device | |
CN113947567B (en) | Defect detection method based on multitask learning | |
CN114648736A (en) | Robust engineering vehicle identification method and system based on target detection | |
CN114359948A (en) | Power grid wiring diagram primitive identification method based on overlapping sliding window mechanism and YOLOV4 |
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 |