CN116665077A

CN116665077A - Power transmission line detection shooting method and system based on AI (advanced identification) recognition technology

Info

Publication number: CN116665077A
Application number: CN202310661983.2A
Authority: CN
Inventors: 程永生; 谢蓓敏
Original assignee: Super High Voltage Co Of State Grid Jilin Electric Power Co ltd
Current assignee: Super High Voltage Co Of State Grid Jilin Electric Power Co ltd
Priority date: 2023-06-06
Filing date: 2023-06-06
Publication date: 2023-08-29

Abstract

The application provides a transmission line detection shooting method and a transmission line detection shooting system based on an AI (advanced technology) recognition technology, wherein the method comprises the following steps: acquiring image data transmitted by acquisition equipment in real time, and performing image preprocessing on the image data to obtain a preprocessed image; performing image edge detection on the preprocessed image to obtain a target image containing contour information of a target object; establishing an identification model of the power transmission line based on a deep neural network architecture, and carrying out target identification on the target image by utilizing the identification model so as to obtain the position information of the target object; and adjusting the position and the gesture of the acquisition equipment based on the position information so as to realize accurate acquisition of the image. The application realizes the identification and detection of the target object in the power transmission line based on the acquired image and picture combined with the identification technology, and automatically adjusts the position and the posture of the acquisition equipment, so that the acquisition equipment can automatically focus and quickly and accurately lock the target.

Description

Power transmission line detection shooting method and system based on AI (advanced identification) recognition technology

Technical Field

The application relates to the technical field of intelligent detection of unmanned aerial vehicles, in particular to a transmission line detection shooting method and a transmission line detection shooting system based on an AI identification technology.

Background

The electric line spacer is used as an important component for keeping the safety distance of the electric transmission line, and the equipment element is also used as a key inspection point for shooting in the inspection process. However, in the current three-dimensional route planning autonomous routing inspection operation based on the route point cloud, the position of the route spacer cannot be judged due to the point cloud density and the display effect, corresponding waypoint information is set, corresponding aerial photographing points can be set for only relatively obvious equipment elements such as an iron tower body, an insulator string, a wire and the like, and the unmanned aerial vehicle can fixedly execute the route waypoints to finish the route inspection operation. And through manual control unmanned aerial vehicle flies along the line, equipment components such as line wire, conductor spacer take a photograph and gather, fly and patrol in-process, the conductor receives factors such as windage yaw, can't ensure that the conductor spacer is in view central point position, and the manual work needs constantly to adjust unmanned aerial vehicle gesture and cloud platform angle, and is higher to manual work flight control level requirement, and very easily leads to the emergence of quick-witted accident because of artificial factor. Therefore, the recognition and detection scheme for the ground wire and the spacer of the power transmission line based on the machine vision of the unmanned aerial vehicle is provided, and the precise focusing shooting of the spacer in the inspection process is realized.

Disclosure of Invention

The embodiment of the application provides a transmission line detection shooting method and a transmission line detection shooting system based on an AI (advanced identification) recognition technology, which at least solve the defects in the related technologies.

In a first aspect, an embodiment of the present application provides a transmission line detection shooting method based on an AI identification technology, including the following steps:

step one: acquiring image data transmitted by acquisition equipment in real time, and performing image preprocessing on the image data to obtain a preprocessed image;

step two: performing image edge detection on the preprocessed image to obtain a target image containing contour information of a target object;

step three: establishing an identification model of the power transmission line based on a deep neural network architecture, and carrying out target identification on the target image by utilizing the identification model so as to obtain the position information of the target object;

step four: and adjusting the position and the gesture of the acquisition equipment based on the position information so as to realize accurate acquisition of the image.

Further, the first step includes:

image denoising is carried out on the image data, and the size and standard deviation of a Gaussian filter convolution kernel in the image data after image denoising are determined;

calculating the value of each element in the Gaussian filter convolution kernel according to the size of the Gaussian filter convolution kernel and the standard deviation, and applying the Gaussian filter convolution kernel to each pixel point in the image data after image denoising to obtain a corresponding output image;

and carrying out image enhancement on the output image to obtain the preprocessing image.

Further, the step of performing image enhancement on the output image to obtain the preprocessed image includes:

converting the output image into a gray level image, and counting the number of pixels of each gray level in the gray level image to generate a gray level histogram;

calculating a cumulative distribution function corresponding to each gray level according to the gray level histogram, and calculating a histogram equalization function by using the cumulative distribution function;

and mapping each gray level in the gray level image to a new gray level by using the histogram equalization function to obtain the preprocessing image.

Further, the second step includes:

and extracting separation information between key features and background features in the preprocessed image by using an edge detection algorithm to obtain the target image containing the contour information of the target object.

Further, the fourth step includes:

the laser sensor is used for ranging the target object to obtain ranging data of the target object;

and carrying out position and posture adjustment on the acquisition equipment by combining the position information with the ranging data and a PID control algorithm so as to realize accurate acquisition of images.

In a second aspect, an embodiment of the present application provides an AI-recognition technology-based power transmission line detection shooting system, including:

the preprocessing module is used for acquiring the image data transmitted by the acquisition equipment in real time and carrying out image preprocessing on the image data to obtain a preprocessed image;

the edge detection module is used for carrying out image edge detection on the preprocessed image so as to obtain a target image containing contour information of a target object;

the target recognition module is used for establishing a recognition model of the power transmission line based on the deep neural network architecture, and carrying out target recognition on the target image by utilizing the recognition model so as to obtain the position information of the target object;

and the gesture adjusting module is used for adjusting the position and gesture of the acquisition equipment based on the position information so as to realize accurate acquisition of the image.

Further, the preprocessing module includes:

the image denoising unit is used for performing image denoising on the image data and determining the size and standard deviation of a Gaussian filter convolution kernel in the image data after image denoising;

the image processing unit is used for calculating the value of each element in the Gaussian filter convolution kernel according to the size of the Gaussian filter convolution kernel and the standard deviation, and applying the Gaussian filter convolution kernel to each pixel point in the image data after image denoising so as to obtain a corresponding output image;

and the image enhancement unit is used for carrying out image enhancement on the output image so as to obtain the preprocessing image.

Further, the image enhancement unit is specifically configured to:

Further, the edge detection module includes:

and the edge detection unit is used for extracting separation information between key features and background features in the preprocessed image by utilizing an edge detection algorithm so as to obtain the target image containing the contour information of the target object.

Further, the posture adjustment module includes:

the laser ranging unit is used for ranging the target object by using a laser sensor so as to obtain ranging data of the target object;

and the gesture adjusting unit is used for adjusting the position and the gesture of the acquisition equipment by combining the position information with the ranging data and a PID control algorithm so as to realize accurate acquisition of images.

Compared with the related art, the power transmission line detection shooting method and the power transmission line detection shooting system based on the AI recognition technology provided by the embodiment of the application are based on the image pictures acquired by the acquisition equipment, the recognition detection of the target object in the power transmission line is realized by combining the recognition technology, the position and the gesture of the acquisition equipment are automatically adjusted by combining the spatial position relation of the target object and the acquisition equipment, so that the power transmission line can be automatically focused, the target can be quickly and accurately locked, the target object is shot, the image is clearer and more accurate, the image acquisition quality is improved, the inspection quality and the defect research and judgment efficiency of the power transmission line acquisition equipment are improved, and the operation and maintenance working quality effect of the power transmission line is comprehensively improved.

The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the other features, objects, and advantages of the application.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:

fig. 1 is a flowchart of a transmission line detection shooting method based on AI identification technology in a first embodiment of the application;

FIG. 2 is a detailed flowchart of step S101 in FIG. 1;

FIG. 3 is a detailed flowchart of step S104 in FIG. 1;

fig. 4 is a block diagram of a transmission line detection shooting system based on AI identification technology in a second embodiment of the present application.

The application will be further described in the following detailed description in conjunction with the above-described figures.

Detailed Description

The present application will be described and illustrated with reference to the accompanying drawings and examples in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application. All other embodiments, which can be made by a person of ordinary skill in the art based on the embodiments provided by the present application without making any inventive effort, are intended to fall within the scope of the present application.

It is apparent that the drawings in the following description are only some examples or embodiments of the present application, and it is possible for those of ordinary skill in the art to apply the present application to other similar situations according to these drawings without inventive effort. Moreover, it should be appreciated that while such a development effort might be complex and lengthy, it would nevertheless be a routine undertaking of design, fabrication, or manufacture for those of ordinary skill having the benefit of this disclosure, and thus should not be construed as having the benefit of this disclosure.

Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is to be expressly and implicitly understood by those of ordinary skill in the art that the described embodiments of the application can be combined with other embodiments without conflict.

Unless defined otherwise, technical or scientific terms used herein should be given the ordinary meaning as understood by one of ordinary skill in the art to which this application belongs. The terms "a," "an," "the," and similar referents in the context of the application are not to be construed as limiting the quantity, but rather as singular or plural. The terms "comprising," "including," "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, article, or apparatus that comprises a list of steps or modules (elements) is not limited to only those steps or elements but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus. The terms "connected," "coupled," and the like in connection with the present application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The term "plurality" as used herein means two or more. "and/or" describes an association relationship of an association object, meaning that there may be three relationships, e.g., "a and/or B" may mean: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship. The terms "first," "second," "third," and the like, as used herein, are merely distinguishing between similar objects and not representing a particular ordering of objects.

Example 1

Referring to fig. 1, a transmission line detection shooting method based on AI identification technology in a first embodiment of the application is shown, and the method specifically includes steps S101 to S104:

s101, acquiring image data transmitted by acquisition equipment in real time, and performing image preprocessing on the image data to obtain a preprocessed image;

further, referring to fig. 2, the step S101 specifically includes steps S1011 to S1013:

s1011, performing image denoising on the image data, and determining the convolution kernel size and standard deviation of a Gaussian filter in the image data after image denoising;

s1012, calculating the value of each element in the Gaussian filter convolution kernel according to the size of the Gaussian filter convolution kernel and the standard deviation, and applying the Gaussian filter convolution kernel to each pixel point in the image data after image denoising to obtain a corresponding output image;

and S1013, performing image enhancement on the output image to obtain the preprocessing image.

The current transmission line earth wire inspection process is influenced by human factors and climate factors, and a remote controller for manually and frequently operating an airplane and a load is required, so that the power line and a spacer are in a visual angle range, and the working efficiency of the mode is low. According to the embodiment, the unmanned aerial vehicle is provided with visible light and load equipment, the AI recognition technology is used for carrying out data preprocessing and feature extraction on real-time image transmission images of the unmanned aerial vehicle, recognition and detection of the spacer are finally realized, the position of a target can be combined, the posture of a cradle head of the unmanned aerial vehicle can be automatically adjusted, the shooting object is ensured to be positioned in the center of the image, and clear images are obtained;

in specific implementation, the image preprocessing is a key link for extracting a power transmission line, namely, each part of digital image is sorted out and transmitted to a corresponding processing module for analysis and processing, and mainly aims to eliminate irrelevant data in a picture, recover and highlight effective information which can help solve actual problems, enhance operability and reconnaissance of the information, greatly reduce data redundancy, achieve the effects of improving feature recognition and image matching, and mainly aim to perform operations such as denoising, smoothing and image enhancement on an original acquired image.

Specifically, when unmanned aerial vehicle electric power inspection aerial photography, the background of a power transmission line is changeable and is mostly influenced by various natural landscapes, and under the complex natural background, the characteristics are easy to be interfered and become blurred, and effective identification cannot be performed, so that a certain denoising treatment is performed on an image before the target characteristics are extracted, sharp noise caused by strong light interference in the image is filtered, the edge of the image is enhanced, and the characteristics of the target object are indirectly highlighted.

1) Image denoising

Based on the recognition characteristics of clear stick shape, obvious outline, large color contrast and the like of the transmission line spacer, the median filter is adopted to carry out denoising treatment, so that salt and pepper noise and other outliers are effectively removed, meanwhile, the image edge information is not lost, and the outline of the original line spacer can be maintained.

2) Smoothing process

For smoothing operation, since the contour information of the line spacer needs to be kept, a smoothing filter which is too strong is not suitable to be used, and therefore a milder Gaussian filter is selected to carry out smoothing processing, the smoothing filter can eliminate Gaussian noise and random noise in an image, smooth the image but not too fuzzy, and meanwhile edge information of the image can be kept.

And reading the power transmission line inspection picture to be subjected to smoothing, determining the size and standard deviation of a Gaussian filter convolution kernel, and calculating the value of each element in the Gaussian convolution kernel according to the determined size and standard deviation of the filter. The convolution kernel is applied to each pixel point on the input image, and corresponding pixel values in the output image are generated in a weighted average mode. The final output image is the result of the smoothing process by the Gaussian filter.

Further, the power transmission line inspection picture is enhanced by aiming at a histogram equalization algorithm after denoising and smoothing treatment, and the visual contrast and the definition of the image are added.

And converting the input image into a gray image, counting the number of pixels of each gray level, generating a gray histogram, and calculating a cumulative distribution function corresponding to each gray level according to the gray histogram. The cumulative distribution function reflects the proportion of pixels in the image having a gray value of a certain gray value or less. And calculating a histogram equalization function according to the cumulative distribution function. The histogram equalization function maps the original gray level to a new gray level, so that the gray level of the output image is more evenly distributed, thereby increasing the contrast and sharpness of the image. And applying the calculated histogram equalization function to each pixel point on the input image to obtain an output image. Specifically, for the gray value of each pixel point, a new gray value corresponding to the gray value is obtained by searching an equalization function, and the original gray value is replaced. The final output image is the result of the enhancement processing of the histogram equalization algorithm and is used for the subsequent feature detection and extraction.

S102, performing image edge detection on the preprocessed image to obtain a target image containing contour information of a target object;

in practice, in digital images, the brightness of the high frequency region varies greatly, with the most significant part of the image being the edge. Edges are mainly the separation between key features, between features and the background, characterizing image discontinuities such as abrupt brightness changes, texture marks, colored icons, etc. And extracting the profile information of the spacer in the power transmission line by using a Canny equal-edge detection algorithm.

S103, establishing an identification model of the power transmission line based on a deep neural network architecture, and carrying out target identification on the target image by utilizing the identification model so as to obtain the position information of the target object;

in specific implementation, the embodiment utilizes YOLOv3 for model construction, and YOLOv3 is a real-time target detection algorithm, has high processing speed and high accuracy, and has obvious advantages in processing large-scale picture data. Meanwhile, the technical algorithm has a simple training flow, can rapidly perform model training only by a small amount of sample data, and is suitable for detecting the target of the inspection scene of the transmission line with a small data volume. Therefore, a recognition model for the transmission line spacer is established and constructed based on the YOLOv3 by adopting a deep neural network architecture, and the extracted features are classified and recognized by calculating the shape, the size, the color and other features of the spacer in the image aiming at the unmanned aerial vehicle transmission line inspection picture, so that the detection and the positioning of the spacer are realized.

S104, adjusting the position and the gesture of the acquisition equipment based on the position information so as to realize accurate acquisition of the image.

Further, referring to fig. 3, the step S104 specifically includes steps S1041 to S1042:

s1041, ranging the target object by using a laser sensor to obtain ranging data of the target object;

s1042, the position information is combined with the distance measurement data and the PID control algorithm to adjust the position and the posture of the acquisition equipment so as to realize accurate acquisition of images.

In the implementation, aiming at image data provided by the unmanned aerial vehicle wire following inspection process, an edge detection algorithm is adopted to identify a spacer and a ground wire, the position of a picture transmission interface where the spacer of the ground wire is located is fed back, the unmanned aerial vehicle hovers first, then a laser radar ranging numerical value is combined, a large-scale SDK virtual rocker is called by utilizing a PID control algorithm, the position and the cradle head posture of the unmanned aerial vehicle are automatically adjusted, the ground wire is ensured to be located in the center of the picture transmission interface, and high-definition images of the ground wire are accurately acquired.

In summary, the power transmission line detection shooting method based on the AI identification technology in the above embodiment of the present application realizes identification detection of the target object in the power transmission line based on the image picture acquired by the acquisition device, combines the identification technology, combines the spatial position relationship between the target object and the acquisition device, automatically adjusts the position and the posture of the acquisition device, so that the power transmission line can automatically focus, quickly and accurately lock the target, shoots the target object, makes the image clearer and more accurate, improves the image acquisition quality, improves the inspection quality and defect research and judgment efficiency of the power transmission line acquisition device, and comprehensively improves the operation and maintenance working quality of the power transmission line.

Example two

In another aspect, referring to fig. 4, a transmission line detection shooting system based on AI identification technology in a second embodiment of the present application is shown, including:

the preprocessing module 11 is used for acquiring the image data transmitted by the acquisition equipment in real time and performing image preprocessing on the image data to obtain a preprocessed image;

further, the preprocessing module 11 includes:

Further, the image enhancement unit is specifically configured to:

An edge detection module 12, configured to perform image edge detection on the preprocessed image, so as to obtain a target image containing contour information of a target object;

further, the edge detection module 12 includes:

The target recognition module 13 is used for establishing a recognition model of the power transmission line based on a deep neural network architecture, and performing target recognition on the target image by utilizing the recognition model so as to obtain the position information of the target object;

and the gesture adjustment module 14 is used for adjusting the position and gesture of the acquisition device based on the position information so as to realize accurate acquisition of the image.

Further, the posture adjustment module 14 includes:

The functions or operation steps implemented when the above modules are executed are substantially the same as those in the above method embodiments, and are not described herein again.

The power transmission line detection shooting system based on the AI identification technology provided by the embodiment of the application has the same implementation principle and technical effects as those of the method embodiment, and for the sake of brief description, the corresponding contents in the method embodiment can be referred to where the system embodiment part is not mentioned.

The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The above examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of protection of the present application is to be determined by the appended claims.

Claims

1. The transmission line detection shooting method based on the AI recognition technology is characterized by comprising the following steps of:

2. The transmission line detection shooting method based on AI identification technology of claim 1, wherein the first step includes:

3. The AI-recognition-technology-based transmission line detection shooting method according to claim 2, wherein the step of image-enhancing the output image to obtain the preprocessed image includes:

4. The transmission line detection shooting method based on the AI-recognition technology of claim 1, wherein the step two includes:

5. The transmission line detection shooting method based on the AI-recognition technology of claim 1, wherein the fourth step includes:

6. An AI identification technology-based transmission line detection shooting system is characterized by comprising:

7. The AI-recognition-technology-based transmission line detection shooting system of claim 6, wherein the preprocessing module comprises:

8. The AI-recognition-technology-based transmission line detection shooting system of claim 7, wherein the image enhancement unit is specifically configured to:

9. The AI-recognition-technology-based transmission line detection shooting system of claim 6, wherein the edge detection module comprises:

10. The AI-recognition-technology-based transmission line detection shooting system of claim 6, wherein the gesture adjustment module comprises: