CN111079752A - Method and device for identifying circuit breaker in infrared image and readable storage medium - Google Patents
Method and device for identifying circuit breaker in infrared image and readable storage medium Download PDFInfo
- Publication number
- CN111079752A CN111079752A CN201911320194.2A CN201911320194A CN111079752A CN 111079752 A CN111079752 A CN 111079752A CN 201911320194 A CN201911320194 A CN 201911320194A CN 111079752 A CN111079752 A CN 111079752A
- Authority
- CN
- China
- Prior art keywords
- circuit breaker
- image
- infrared
- infrared image
- target
- 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
- 238000000034 method Methods 0.000 title claims abstract description 67
- 238000001514 detection method Methods 0.000 claims abstract description 74
- 238000012549 training Methods 0.000 claims abstract description 71
- 238000012545 processing Methods 0.000 claims abstract description 33
- 238000007781 pre-processing Methods 0.000 claims abstract description 17
- 238000003062 neural network model Methods 0.000 claims abstract description 7
- 230000003321 amplification Effects 0.000 claims description 15
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000002372 labelling Methods 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000002329 infrared spectrum Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
本申请公开了一种识别红外图像中的断路器的方法、装置及计算机可读存储介质。其中，方法包括对初始训练样本集进行图像预处理，使得每张红外样本图像上标注的断路器的检测框与预设参考面之间的夹角和参考面与相应断路器所成夹角相同；并进行样本总数扩增处理，最后利用处理得到的训练样本集训练神经网络模型得到目标识别模型。将待处理红外图像输入至目标识别模型中，得到包含标注断路器检测框的红外图像，该检测框与参考面之间的第一夹角与断路器和参考面所成的第二夹角的差值不大于预设角度误差值。本申请无需收集大量训练样本便可训练得到精度高的模型，且可提高断路器在红外图像中检测位置和实际所处位置的贴合度。
The present application discloses a method, device and computer-readable storage medium for identifying a circuit breaker in an infrared image. The method includes image preprocessing on the initial training sample set, so that the angle between the detection frame of the circuit breaker marked on each infrared sample image and the preset reference plane and the angle formed by the reference plane and the corresponding circuit breaker are the same ; And the total number of samples is amplified, and finally the neural network model is trained by using the training sample set obtained by processing to obtain the target recognition model. Input the infrared image to be processed into the target recognition model, and obtain the infrared image including the detection frame of the marked circuit breaker, the first angle between the detection frame and the reference plane and the second angle formed by the circuit breaker and the reference plane The difference is not greater than the preset angle error value. The present application can train a model with high precision without collecting a large number of training samples, and can improve the fit between the detected position of the circuit breaker in the infrared image and the actual position.
Description
Technical Field
The present application relates to the field of image processing technologies, and in particular, to a method and an apparatus for identifying a circuit breaker in an infrared image, and a computer-readable storage medium.
Background
Along with infrared equipment's wide application in the electrified detection of circuit breaker, power supply company has accumulated a large amount of fortune dimension infrared picture data, whether have the circuit breaker to sieve by the manual work in the infrared picture at present to by the position that artifical mark circuit breaker is located in infrared image, so that be used for later stage storage and statistics etc..
However, the manual detection efficiency is low, the outdoor environment is complex, the detection result cannot be given in real time, in addition, the manual labeling format is not standard, and the labeled picture data does not help to improve the detection efficiency and the identification rate in the future, so that the use and analysis efficiency of the historical data is low, and the inspection and identification work in the future cannot be helped powerfully.
In order to solve the problems of manual detection and labeling, related technologies perform screening through an image recognition technology, but due to the problem of the infrared picture shooting angle, a large-scale picture is needed to be used as a training sample set, the recognition result is not always labeled according to the angle of the breaker in the picture, and most useless background information exists in the region obtained through labeling.
Disclosure of Invention
The method, the device and the computer readable storage medium for identifying the circuit breaker in the infrared image can be used for training to obtain a high-precision model without collecting a large number of training sample sets, so that the fitting degree of the detection position and the actual position of the circuit breaker in the infrared image is improved, and useless background information contained in the circuit breaker area is greatly reduced.
In order to solve the above technical problems, embodiments of the present invention provide the following technical solutions:
an embodiment of the present invention provides a method for identifying a circuit breaker in an infrared image, including:
acquiring an infrared image to be processed containing a target circuit breaker;
inputting the infrared image to be processed into a pre-constructed target identification model to obtain an infrared image containing a target circuit breaker detection frame marked by the data marking format, wherein the difference value between a first included angle between the target circuit breaker detection frame and the reference surface and a second included angle formed by the target circuit breaker and the reference surface is not more than a preset angle error value;
the target identification model is obtained by training a neural network model by using a training sample set obtained by carrying out image preprocessing and sample total number amplification processing on an initial training sample set, and an included angle between a detection frame of a circuit breaker and a preset reference surface, which is marked on each infrared sample image in the training sample set according to a preset data marking format, is the same as an included angle formed by the reference surface and the corresponding circuit breaker.
Optionally, each infrared sample image in the initial training sample set includes a corresponding label, where the label is used to indicate whether a circuit breaker in the graph is tilted and a tilt direction; the initial training sample set is subjected to sample total number amplification processing through image angle rotation, and the process of the initial training sample set for sample total number amplification processing through image angle rotation comprises the following steps:
respectively rotating the infrared sample image with the label being inclined according to a plurality of angle values in a first preset angle range, taking the image obtained by rotation as a new sample image, and simultaneously generating a corresponding label;
respectively rotating the infrared sample image with the label being not inclined according to a plurality of angle values in a second preset angle range, taking the image obtained by rotation as a new sample image, and simultaneously generating a corresponding label;
wherein a value in the second preset angle range is smaller than a value in the first preset angle range.
Optionally, after the to-be-processed infrared image including the target circuit breaker is acquired, the method further includes:
generating identification information of the infrared image to be processed so as to establish a corresponding relation between the infrared image to be processed and the target circuit breaker;
and the identification information is the number information of the target circuit breaker.
Optionally, the target circuit breaker detection frame is a rectangular frame, and after the to-be-processed infrared image is input into the target identification model and an infrared image including the target circuit breaker detection frame marked by the data marking format is obtained, the method further includes:
and outputting the coordinate information of the four vertexes of the target circuit breaker detection frame and the equipment parameter information of the target circuit breaker to be used as the attribute information of the infrared image for storage.
Optionally, the pre-image-preprocessing the initial training sample set includes:
dividing the infrared sample images in the initial sample training set into two types according to whether the circuit breaker in the image is inclined or not, and generating an image group to be processed and a positive target image group; the circuit breaker of the infrared sample image in the positive target image group is parallel to the central axis of the image;
calculating the deflection angle of a current circuit breaker in the current infrared sample image deviating from the reference surface for each infrared sample image in the image group to be processed, rotating the current infrared sample image according to the deflection angle until the current circuit breaker is parallel to the central axis of the image, marking a detection frame of the circuit breaker in the current infrared sample image by using a data marking method, and rotating the current infrared sample image until the angle formed by the current circuit breaker and the reference surface is the deflection angle after the detection frame is generated.
Optionally, the target identification model identifies the circuit breaker in the infrared image based on an SSD image identification method, and marks the circuit breaker detection frame in a VOC format.
Optionally, after acquiring the to-be-processed infrared image including the circuit breaker, the method further includes:
and carrying out gray level processing on the infrared image to be processed to obtain a binary infrared image, so as to input the binary infrared image into the target identification model.
Another aspect of an embodiment of the present invention provides an apparatus for identifying a circuit breaker in an infrared image, including:
the image acquisition module is used for acquiring an infrared image to be processed containing a target circuit breaker;
the target identification module is used for inputting the infrared image to be processed into a pre-constructed target identification model to obtain an infrared image which comprises a target circuit breaker detection frame marked by the data marking format, and the difference value between a first included angle between the target circuit breaker detection frame and the reference surface and a second included angle formed by the target circuit breaker and the reference surface is not more than a preset angle error value; the target recognition model is obtained by training a neural network model by using a training sample set obtained by carrying out image preprocessing and sample total number amplification processing on an initial training sample set, and an included angle between a detection frame of a circuit breaker and a preset reference surface, which is marked on each infrared sample image in the training sample set according to a preset data marking format, is the same as an included angle formed by the reference surface and the corresponding circuit breaker.
An embodiment of the present invention further provides an apparatus for identifying a circuit breaker in an infrared image, including a processor, where the processor is configured to implement the steps of the method for identifying a circuit breaker in an infrared image according to any one of the preceding items when executing a computer program stored in a memory.
Finally, an embodiment of the present invention provides a computer-readable storage medium, where a program for identifying a circuit breaker in an infrared image is stored, and when the program for identifying a circuit breaker in an infrared image is executed by a processor, the steps of the method for identifying a circuit breaker in an infrared image are implemented as in any one of the previous embodiments.
The technical scheme provided by the application has the advantages that the included angle between the detection frame of the circuit breaker marked on the sample image according to the preset data marking format and the preset reference surface is the same as the included angle between the reference surface and the corresponding circuit breaker by image preprocessing the sample image, so that invalid background information contained in the detection frame is reduced; the total number of training samples is increased through the total number of samples amplification processing, so that the angles in the images where various circuit breakers are located can be covered in a training sample set, and a model with high identification precision can be obtained without collecting a large number of training sample sets; the circuit breaker in the infrared image can be effectively marked in real time by utilizing a target recognition model obtained based on the training sample set through deep learning, so that the fitting degree of the detection position and the actual position of the circuit breaker in the infrared image is improved, and useless background information contained in the circuit breaker area in the detection frame is greatly reduced; in addition, the image data with uniform formatting is generated, and the storage and the later use of a back-end server are facilitated.
In addition, the embodiment of the invention also provides a corresponding implementation device and a computer readable storage medium for the method for identifying the circuit breaker in the infrared image, so that the method has higher practicability, and the device, the equipment and the computer readable storage medium have corresponding advantages.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the related art, the drawings required to be used in the description of the embodiments or the related art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a schematic flowchart of a method for identifying a circuit breaker in an infrared image according to an embodiment of the present invention;
fig. 2 is a structural diagram of an embodiment of an apparatus for identifying a circuit breaker in an infrared image according to an embodiment of the present invention;
fig. 3 is a block diagram of another specific embodiment of an apparatus for identifying a circuit breaker in an infrared image according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the disclosure, the invention will be described in further detail with reference to the accompanying drawings and specific embodiments. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms "first," "second," "third," "fourth," and the like in the description and claims of this application and in the above-described drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "comprising" and "having," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements but may include other steps or elements not expressly listed.
Having described the technical solutions of the embodiments of the present invention, various non-limiting embodiments of the present application are described in detail below.
Referring to fig. 1, fig. 1 is a schematic flow chart of a method for identifying a circuit breaker in an infrared image according to an embodiment of the present invention, where the embodiment of the present invention may include the following:
s101: and carrying out image preprocessing and total sample number amplification processing on the initial training sample set in advance, and training the neural network model by using the training sample set obtained by processing to obtain a target recognition model.
In the application, the target identification model can be divided into two steps in the execution process, wherein the first step is to identify the circuit breaker in the image in the infrared image, and the second step is to set a detection frame for the circuit breaker in the image after identifying the circuit breaker so as to frame the area position of the infrared image where the circuit breaker is located. For example, the target identification model is to identify the circuit breaker in the infrared image based on the SSD image identification method, and mark the circuit breaker detection frame in the VOC format. Of course, any image recognition algorithm and any labeling algorithm in any format can be used, which does not affect the implementation of the present application.
It can be understood that, the infrared image obtained by collecting the breaker is not straight in the picture due to the shooting angle problem, where straight means parallel to the central axis of the image but having various angles, such as 45 degrees in the left oblique direction and 180 degrees in the horizontal direction, but all the detection algorithms draw a straight frame at present, the detection frame cannot be attached to the target, the area of all pixel points in the frame is relatively large in order to include the breaker, correspondingly, the included useless background information is much, the size of all the detection frames is the same and is the same as that of the infrared image without an inclination angle, so that most of the breaker area of the detection frame of the infrared image with a large inclination angle is not framed, and subsequent temperature detection cannot be performed at all. Based on this, image preprocessing needs to be performed on the sample image in the initial training sample set, an included angle between the detection frame of the circuit breaker marked on the processed infrared sample image according to the preset data marking format and the preset reference surface is the same as an included angle formed by the reference surface and the corresponding circuit breaker, and the reference surface can be, for example, a plane where the central axis of the image is located, that is, the inclination angle of the detection frame and the clear angle of the circuit breaker are kept consistent.
In addition, because fewer samples are collected in an actual scene, various target angles cannot be covered, and the number of samples and the types of samples of a training model determine the training precision of the model within a certain range. Based on this, the method and the device also perform sample total number amplification processing on the sample images in the initial training sample set, for example, the sample total number can be increased through image angle rotation, and the rotated sample images are stored as new training samples, so that the number and diversity of the training sample set are increased, and the model training precision is improved.
S102: and acquiring the infrared image to be processed containing the target circuit breaker.
The technical scheme provided by the application can be directly integrated into the equipment for collecting the infrared image, so that after the infrared image collecting equipment obtains the infrared image with the definition meeting the requirement, the infrared image collecting equipment can obtain a target recognition model by utilizing an S101 training and mark the circuit breaker by utilizing a detection frame; the technical scheme provided by the application can be integrated on a handheld terminal or any intelligent terminal device or a server, the infrared image acquisition device sends the acquired infrared image containing the target circuit breaker to the handheld terminal or the intelligent terminal or the server, and the terminal or the server marks the target area where the voltage transformer to be detected in the image is located by using the model obtained through S101 training after receiving the original image. In order to facilitate storage and tracing of subsequent infrared images and solve the problem of false detection caused by non-standard picture storage, index information can be added to the images after the infrared images are received, the index information is used for uniquely identifying information of the infrared images by establishing a corresponding relation between the infrared images to be processed and a target circuit breaker, namely identification information of the infrared images to be processed is generated, and the identification information can be the serial number of the circuit breaker.
S103: and inputting the infrared image to be processed into the target identification model to obtain the infrared image containing the target circuit breaker detection frame marked by the data marking format.
It can be understood that the angle value formed between the labeled detection frame and the reference surface is not completely the same as the angle value formed between the target circuit breaker and the reference surface, and both the error value and the accuracy requirement are considered, and the difference value between the first included angle between the target circuit breaker detection frame and the reference surface and the second included angle formed between the target circuit breaker and the reference surface is not greater than the preset angle error value. The angle error value may be, for example, 5 °, for example, the angle between the target circuit breaker and the plane of the central axis is 45 °, as long as the angle between the target circuit breaker detection frame and the plane of the central axis is [40 °, 50 ° ] is satisfactory. The preset angle error value can be valued according to the actual application scene and the precision requirement, and the application does not limit the value at all.
In the technical scheme provided by the embodiment of the invention, the sample image is subjected to image preprocessing, so that the included angle between the detection frame of the circuit breaker marked on the sample image according to the preset data marking format and the preset reference surface is the same as the included angle between the reference surface and the corresponding circuit breaker, and the invalid background information contained in the detection frame is reduced; the total number of training samples is increased through the total number of samples amplification processing, so that the angles in the images where various circuit breakers are located can be covered in a training sample set, and a model with high identification precision can be obtained without collecting a large number of training sample sets; the circuit breaker in the infrared image can be effectively marked in real time by utilizing a target recognition model obtained based on the training sample set through deep learning, so that the fitting degree of the detection position and the actual position of the circuit breaker in the infrared image is improved, and useless background information contained in the circuit breaker area in the detection frame is greatly reduced; in addition, the image data with uniform formatting is generated, and the storage and the later use of a back-end server are facilitated.
In the foregoing embodiment, how to perform the total sample number amplification processing on the initial training sample set by rotating the image angle is not limited, and a specific implementation method is provided in this embodiment, and may include the following steps:
each infrared sample image in the initial training sample set comprises a corresponding label, and the label is used for indicating whether the circuit breaker in the graph is inclined or not and the inclination direction. For example, may include class 3 tags, respectively c representing a positive target, i.e., the breaker has not been tilted; cR represents that the breaker is tilted and the tilted direction of the breaker is from top right to bottom left in the drawing, and cL represents that the breaker is tilted and the tilted direction of the breaker is from top left to bottom right in the drawing.
Respectively rotating the infrared sample image with the label being inclined according to a plurality of angle values in a first preset angle range, taking the image obtained by rotation as a new sample image, and simultaneously generating a corresponding label; for example, for an infrared sample image of a cR, cL label, 2 sets of two angles of [0 °, 90 ° ], [ -90 °, 0 ° ] are randomly rotated for the circuit breaker between [ -90 °, 90 ° ], and then the original image is rotated and saved.
Respectively rotating the infrared sample image with the label being not inclined according to a plurality of angle values in a second preset angle range, taking the image obtained by rotation as a new sample image, and simultaneously generating a corresponding label; wherein the value in the second preset angle range is smaller than the value in the first preset angle range. For the infrared sample image marked c, in order to enhance the anti-interference capability, a certain perturbation can be added in a small angle range, for example, a random number can be taken at each of [ -4 °, 0 ° ], [0 °, 4 ° ], and then the image is rotated and saved.
Therefore, the embodiment of the invention realizes the total sample number amplification processing by rotating the image angle, not only increases the total number of training samples, but also enriches the types of the training samples, and is beneficial to improving the precision of the training model.
The foregoing embodiment is not limited to how to perform image preprocessing on the initial training sample set to achieve the desired effect, and a specific implementation method provided in this embodiment may include the following steps:
and dividing the infrared sample images in the initial sample training set into two types according to whether the breaker in the image is inclined or not, and generating an image group to be processed and a positive target image group. The breaker of the infrared sample image in the positive target image group is parallel to the central axis of the image, namely the infrared sample image in the positive target image group is not inclined; and all the infrared sample images in the image group to be processed are inclined.
Calculating the deflection angle of a current circuit breaker deviating from a reference surface in a current infrared sample image of each infrared sample image in the image group to be processed, rotating the current infrared sample image according to the deflection angle until the current circuit breaker is parallel to the central axis of the image, simultaneously marking a detection frame of the circuit breaker in the current infrared sample image by using a data marking method, and rotating the current infrared sample image to the deflection angle formed by the current circuit breaker and the reference surface after the detection frame is generated.
Optionally, a two-dimensional coordinate system may be established in the image, and the deflection angle is calculated by combining the coordinate value of each pixel point in the coordinate system with the basic geometric knowledge. The method comprises the steps of rotating the circuit breaker, namely enabling the circuit breaker to be parallel to the central axis of an image, setting a detection frame for the image by adopting any image identification method and marking method in the related technology, and rotating the detection frame with the image to the original image angle after setting the detection frame, so that the inclination angle of the detection frame is consistent with that of the circuit breaker.
Therefore, the embodiment of the invention improves the fitting degree of the detection frame and the circuit breaker by preprocessing the sample image, and is beneficial to improving the precision of the training model.
As an optional implementation manner, in order to further improve the image recognition effect and reduce the interference of useless background information, the to-be-processed infrared image may be subjected to gray scale processing to obtain a binarized infrared image, which is used for inputting the binarized infrared image into the target recognition model. Correspondingly, gray processing can be performed on each infrared sample image in the training sample set, and model training can be performed by using the image subjected to gray processing.
In addition, it should be noted that the system may further include an infrared spectrum database, where the infrared image of the circuit breaker is stored in the infrared spectrum database, and in order to improve richness of information stored in the database and facilitate operations such as daily maintenance and overhaul of the subsequent voltage transformer, after S103, if the target circuit breaker detection frame is a rectangular frame, the coordinate information of four vertices of the target circuit breaker detection frame and the device parameter information of the target circuit breaker are further output to be stored as attribute information of the infrared image.
It should be noted that, in the present application, there is no strict sequential execution order among the steps, and as long as the logical order is met, the steps may be executed simultaneously or according to a certain preset order, and fig. 1 is only an exemplary manner, and does not represent that only the execution order is the order.
The embodiment of the invention also provides a corresponding device for the method for identifying the circuit breaker in the infrared image, so that the method has higher practicability. Wherein the means can be described separately from the functional module point of view and the hardware point of view. The following describes an apparatus for identifying a circuit breaker in an infrared image according to an embodiment of the present invention, and the apparatus for identifying a circuit breaker in an infrared image described below and the method for identifying a circuit breaker in an infrared image described above may be referred to correspondingly.
Based on the angle of the functional module, referring to fig. 2, fig. 2 is a structural diagram of an apparatus for identifying a circuit breaker in an infrared image according to an embodiment of the present invention, in a specific implementation manner, the apparatus may include:
the model training module 201 is configured to perform image preprocessing and sample total number amplification processing on an initial training sample set in advance, train a neural network model with the training sample set obtained through the processing to obtain a target identification model, and obtain an included angle between a detection frame of a circuit breaker and a preset reference surface, which is marked on each infrared sample image in the training sample set according to a preset data marking format, and the included angle between the reference surface and the corresponding circuit breaker is the same.
And the image acquisition module 202 is configured to acquire an infrared image to be processed, which includes the target circuit breaker.
And the target identification module 203 is used for inputting the infrared image to be processed into the target identification model to obtain the infrared image containing the target circuit breaker detection frame marked by using the data marking format, and the difference value between a first included angle between the target circuit breaker detection frame and the reference surface and a second included angle formed by the target circuit breaker and the reference surface is not more than a preset angle error value.
Optionally, in some embodiments of this embodiment, the model training module 201 may include a sample amplification sub-module, where the sample amplification sub-module is configured to perform total sample number amplification processing by rotating an image angle, and specifically may include:
the first rotating unit is used for respectively rotating the infrared sample image with the label inclined according to a plurality of angle values in a first preset angle range, taking the image obtained by rotation as a newly added sample image and simultaneously generating a corresponding label;
the second rotating unit is used for respectively rotating the infrared sample image of which the label is not inclined according to a plurality of angle values in a second preset angle range, taking the image obtained by rotation as a newly added sample image and simultaneously generating a corresponding label; wherein the value in the second preset angle range is smaller than the value in the first preset angle range.
Optionally, in other embodiments of this embodiment, the model training module 201 may include an image preprocessing sub-module, where the image preprocessing sub-module includes:
the image classification unit is used for classifying the infrared sample images in the initial sample training set into two types according to whether the circuit breaker in the image is inclined or not, and generating an image group to be processed and a positive target image group; the circuit breaker of the infrared sample image in the positive target image group is parallel to the central axis of the image;
the image processing unit is used for calculating a deflection angle of a current circuit breaker in a current infrared sample image deviating from a reference surface for each infrared sample image in the image group to be processed, rotating the current infrared sample image according to the deflection angle until the current circuit breaker is parallel to a central axis of the image, meanwhile, marking a detection frame of the circuit breaker in the current infrared sample image by using a data marking method, and rotating the current infrared sample image to a deflection angle formed by the current circuit breaker and the reference surface after the detection frame is generated.
As an optional implementation manner, the apparatus may further include, for example, a grayscale processing module, an identification information generating module, and a storage information generating module; the gray processing module is used for carrying out gray processing on the infrared image to be processed to obtain a binary infrared image so as to input the binary infrared image into the target identification model; the identification information generation module is used for generating identification information of the infrared image to be processed so as to establish a corresponding relation between the infrared image to be processed and the target circuit breaker; the identification information is the number information of the target circuit breaker; the storage information generation module is used for outputting the coordinate information of the four vertexes of the target circuit breaker detection frame and the equipment parameter information of the target circuit breaker to be stored as the attribute information of the infrared image.
The functions of the functional modules of the device for identifying a circuit breaker in an infrared image according to the embodiments of the present invention may be specifically implemented according to the method in the embodiments of the method, and the specific implementation process may refer to the description related to the embodiments of the method, which is not described herein again.
Therefore, the embodiment of the invention can train to obtain the model with high precision without collecting a large amount of training sample sets, improves the fitting degree of the detection position of the circuit breaker in the infrared image and the actual position, and greatly reduces useless background information contained in the circuit breaker area.
The above-mentioned device for identifying a circuit breaker in an infrared image is described from the perspective of a functional module, and further, the present application also provides a device for identifying a circuit breaker in an infrared image, which is described from the perspective of hardware. Fig. 3 is a block diagram of another example of identifying a circuit breaker in an infrared image according to an embodiment of the present disclosure. As shown in fig. 3, the apparatus comprises a memory 30 for storing a computer program;
the processor 31 is configured to implement the steps of the method for detecting the abnormal temperature point of the voltage transformer according to the above embodiment when executing the computer program.
The processor 31 may include one or more processing cores, such as a 4-core processor, an 8-core processor, and the like. The processor 31 may be implemented in at least one hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 31 may also include a main processor and a coprocessor, where the main processor is a processor for processing data in an awake state, and is also called a Central Processing Unit (CPU); a coprocessor is a low power processor for processing data in a standby state. In some embodiments, the processor 31 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content required to be displayed on the display screen. In some embodiments, the processor 31 may further include an AI (Artificial Intelligence) processor for processing a calculation operation related to machine learning.
Memory 30 may include one or more computer-readable storage media, which may be non-transitory. Memory 30 may also include high speed random access memory, as well as non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. In this embodiment, the memory 30 is at least used for storing the following computer program 301, wherein after being loaded and executed by the processor 31, the computer program can implement the relevant steps of the testing method disclosed in any of the foregoing embodiments. In addition, the resources stored by the memory 30 may also include an operating system 302, data 303, and the like, and the storage may be transient storage or permanent storage. Operating system 302 may include Windows, Unix, Linux, etc. Data 303 may include, but is not limited to, data corresponding to test results, and the like.
In some embodiments, the testing device may further include a display 32, an input/output interface 33, a communication interface 34, a power source 35, a communication bus 36, and a sensor 37.
Those skilled in the art will appreciate that the configuration shown in FIG. 3 is not intended to be limiting of testing devices and may include more or fewer components than those shown.
The function of each functional module of the device for identifying a circuit breaker in an infrared image according to the embodiment of the present invention can be specifically implemented according to the method in the embodiment of the method, and the specific implementation process may refer to the description related to the embodiment of the method, which is not described herein again.
Therefore, the embodiment of the invention can train to obtain the model with high precision without collecting a large amount of training sample sets, improves the fitting degree of the detection position of the circuit breaker in the infrared image and the actual position, and greatly reduces useless background information contained in the circuit breaker area.
It is to be understood that, if the method for identifying the circuit breaker in the infrared image in the above embodiment is implemented in the form of a software functional unit and sold or used as a separate product, it may be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application may be substantially or partially implemented in the form of a software product, which is stored in a storage medium and executes all or part of the steps of the methods of the embodiments of the present application, or all or part of the technical solutions. And the aforementioned storage medium includes: a U disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), an electrically erasable programmable ROM, a register, a hard disk, a removable magnetic disk, a CD-ROM, a magnetic or optical disk, and other various media capable of storing program codes.
Based on this, the embodiment of the present invention further provides a computer readable storage medium, which stores a program for identifying a circuit breaker in an infrared image, and the program for identifying a circuit breaker in an infrared image is executed by a processor, and the steps of the method for identifying a circuit breaker in an infrared image are as described in any one of the above embodiments.
The functions of the functional modules of the computer-readable storage medium according to the embodiment of the present invention may be specifically implemented according to the method in the foregoing method embodiment, and the specific implementation process may refer to the related description of the foregoing method embodiment, which is not described herein again.
Therefore, the embodiment of the invention can train to obtain the model with high precision without collecting a large amount of training sample sets, improves the fitting degree of the detection position of the circuit breaker in the infrared image and the actual position, and greatly reduces useless background information contained in the circuit breaker area.
The embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same or similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
A method, an apparatus, and a computer-readable storage medium for identifying a circuit breaker in an infrared image provided by the present application are described in detail above. The principles and embodiments of the present invention are explained herein using specific examples, which are presented only to assist in understanding the method and its core concepts. It should be noted that, for those skilled in the art, it is possible to make various improvements and modifications to the present invention without departing from the principle of the present invention, and those improvements and modifications also fall within the scope of the claims of the present application.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911320194.2A CN111079752A (en) | 2019-12-19 | 2019-12-19 | Method and device for identifying circuit breaker in infrared image and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911320194.2A CN111079752A (en) | 2019-12-19 | 2019-12-19 | Method and device for identifying circuit breaker in infrared image and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111079752A true CN111079752A (en) | 2020-04-28 |
Family
ID=70315938
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911320194.2A Pending CN111079752A (en) | 2019-12-19 | 2019-12-19 | Method and device for identifying circuit breaker in infrared image and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111079752A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113838046A (en) * | 2021-10-08 | 2021-12-24 | 南方电网数字电网研究院有限公司 | Drainage plate overheating identification method and device, computer equipment and storage medium |
| WO2022142783A1 (en) * | 2020-12-29 | 2022-07-07 | 华为云计算技术有限公司 | Image processing method and related device |
| CN114821241A (en) * | 2022-05-23 | 2022-07-29 | 广东电网有限责任公司 | Expansion method of circuit breaker preventive test data sample and related device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101636827B1 (en) * | 2015-04-30 | 2016-07-07 | 주식회사 대영산전 | Vehicle number recognition systems |
| CN107798336A (en) * | 2017-09-18 | 2018-03-13 | 广东电网有限责任公司东莞供电局 | Infrared temperature measurement image component identification method |
| CN109101906A (en) * | 2018-07-27 | 2018-12-28 | 中国南方电网有限责任公司超高压输电公司贵阳局 | A kind of converting station electric power equipment infrared image exception real-time detection method and device |
| CN109948613A (en) * | 2019-03-22 | 2019-06-28 | 国网重庆市电力公司电力科学研究院 | Infrared image recognition method and device for arrester |
-
2019
- 2019-12-19 CN CN201911320194.2A patent/CN111079752A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101636827B1 (en) * | 2015-04-30 | 2016-07-07 | 주식회사 대영산전 | Vehicle number recognition systems |
| CN107798336A (en) * | 2017-09-18 | 2018-03-13 | 广东电网有限责任公司东莞供电局 | Infrared temperature measurement image component identification method |
| CN109101906A (en) * | 2018-07-27 | 2018-12-28 | 中国南方电网有限责任公司超高压输电公司贵阳局 | A kind of converting station electric power equipment infrared image exception real-time detection method and device |
| CN109948613A (en) * | 2019-03-22 | 2019-06-28 | 国网重庆市电力公司电力科学研究院 | Infrared image recognition method and device for arrester |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022142783A1 (en) * | 2020-12-29 | 2022-07-07 | 华为云计算技术有限公司 | Image processing method and related device |
| CN113838046A (en) * | 2021-10-08 | 2021-12-24 | 南方电网数字电网研究院有限公司 | Drainage plate overheating identification method and device, computer equipment and storage medium |
| CN114821241A (en) * | 2022-05-23 | 2022-07-29 | 广东电网有限责任公司 | Expansion method of circuit breaker preventive test data sample and related device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109615611B (en) | Inspection image-based insulator self-explosion defect detection method | |
| CN108648169B (en) | Method and device for automatic identification of insulator defects in high-voltage transmission towers | |
| CN109886928B (en) | Target cell marking method, device, storage medium and terminal equipment | |
| CN112990086B (en) | Remote sensing image building detection method and device and computer readable storage medium | |
| CN111080693A (en) | Robot autonomous classification grabbing method based on YOLOv3 | |
| US20190156202A1 (en) | Model construction in a neural network for object detection | |
| US20170308736A1 (en) | Three dimensional object recognition | |
| CN108228798A (en) | The method and apparatus for determining the matching relationship between point cloud data | |
| CN111144325A (en) | Fault identification and positioning method, device and equipment for power equipment of transformer substation | |
| CN112200081A (en) | Abnormal behavior identification method, device, electronic device and storage medium | |
| CN112348765A (en) | Data enhancement method and device, computer readable storage medium and terminal equipment | |
| CN116168351B (en) | Inspection method and device for power equipment | |
| CN111079752A (en) | Method and device for identifying circuit breaker in infrared image and readable storage medium | |
| Li et al. | Automatic comic page segmentation based on polygon detection | |
| CN115909059A (en) | Natural resource sample library establishing method and device | |
| CN111680750A (en) | Image recognition method, device and equipment | |
| CN111738252B (en) | Text line detection method, device and computer system in image | |
| CN112446231A (en) | Pedestrian crossing detection method and device, computer equipment and storage medium | |
| CN115908988A (en) | Defect detection model generation method, device, equipment and storage medium | |
| CN107767366B (en) | A transmission line fitting method and device | |
| CN114120086A (en) | Pavement disease recognition method, image processing model training method, device and electronic equipment | |
| CN110276260B (en) | A product detection method based on depth camera | |
| CN117788444A (en) | SMT patch offset detection method, SMT patch offset detection device and SMT patch offset detection system | |
| CN111723688A (en) | Evaluation method, device and electronic device for human action recognition results | |
| CN116681792A (en) | Subgraph generation method, device and equipment of vector drawing and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200428 |