Unmanned aerial vehicle ranging method for power line inspection and related equipment

Abstract

The invention relates to the technical field of electrical automation, and provides an unmanned aerial vehicle ranging method for power line inspection and related equipment. The method comprises the following steps: acquiring unmanned aerial vehicle shot images of a plurality of target power lines; for each unmanned aerial vehicle shot image, segmenting the unmanned aerial vehicle shot image to obtain a power line image of the target power line corresponding to the unmanned aerial vehicle shot image; performing distance estimation on each power line image by using a power line depth estimation network to obtain a relative depth between each target power line and the unmanned aerial vehicle; obtaining an absolute depth between each target power line and the unmanned aerial vehicle, and constructing a fitting function according to all the relative depths and all the absolute depths; and measuring the distance between the to-be-inspected power line and the unmanned aerial vehicle by using the power line depth estimation network and the fitting function to obtain a final distance between the to-be-inspected power line and the unmanned aerial vehicle. The method can improve the ranging precision of the unmanned aerial vehicle.

Description

Technical Field

[0001] The present application relates to the field of electrical automation technology, and in particular to a drone ranging method and related equipment for power line inspection. Background Art

[0002] In recent years, drone inspections have become a primary means of daily power line maintenance, effectively improving both the efficiency and quality of inspections. However, despite the increasing maturity of drone technology and its industrial applications, existing drones for power line inspections face significant challenges in ranging applications due to the unique characteristics of power lines and their surrounding environments. This makes it difficult to quickly and accurately determine the safe flight range for drone inspections, leading to frequent drone collisions with power lines. Therefore, drone ranging systems are crucial for achieving distance control for power line inspection drones. Designing a fast, high-precision ranging system for power line inspection drones is of significant engineering significance.

[0003] The distance between inspection drones and power lines is mostly determined visually, posing a significant safety hazard. Consequently, domestic and international researchers have conducted research on ranging systems for drone inspections, primarily focusing on visible light imaging, infrared ranging, lidar ranging, and ultrasonic ranging. Infrared ranging systems are lightweight, highly directional, and capable of penetrating fog and clouds. They are also unaffected by illumination and shadows, making them suitable for measurements in fog and at night. However, infrared images suffer from low resolution, blurred edges, and a lack of texture detail, making it difficult to accurately identify the object being measured. Laser ranging is fast, but requires high roughness and tilt of the object's emitting surface, making it difficult to detect conductors with small cross-sections. Ultrasonic ranging is highly adaptable to various environments, but its measurement distance and range are significantly affected by factors such as the ultrasound's emission angle, power, and the distance to the charged object. Therefore, current drone ranging methods for power line inspection suffer from low accuracy. Summary of the Invention

[0004] The present application provides a drone ranging method and related equipment for power line inspection, which can solve the problem of low accuracy of drone ranging.

[0005] In a first aspect, an embodiment of the present application provides a drone ranging method for power line inspection, the drone ranging method comprising:

[0006] Acquire drone-photographed images of multiple target power lines; the drone-photographed images are images obtained by a drone used for power line inspection to photograph the target power lines;

[0007] Segmenting each drone-photographed image to obtain a power line image of the target power line corresponding to the drone-photographed image;

[0008] The power line depth estimation network is used to estimate the distance of each power line image and obtain the relative depth between each target power line and the UAV. The relative depth is the estimated distance between the target power line and the UAV.

[0009] Obtain the absolute depth between each target power line and the drone, and construct a fitting function based on all relative depths and all absolute depths. The absolute depth is the actual distance between the target power line and the drone, and the fitting function is used to describe the mapping relationship between relative depth and absolute depth.

[0010] The power line depth estimation network and fitting function are used to measure the distance between the power line to be inspected and the UAV, and the final distance between the power line to be inspected and the UAV is obtained.

[0011] Optionally, the power line depth estimation network includes a feature extraction module, a dynamic scale integration module, and a hierarchical recursive integration module connected in sequence;

[0012] The input end of the feature extraction module is the input end of the power line depth estimation network, and the output end of the hierarchical recursive integration module is the output end of the power line depth estimation network.

[0013] Optionally, the layered recursive integration module includes a first recurrent convolution layer, a second recurrent convolution layer, a third recurrent convolution layer, a first upsampling layer, a second upsampling layer, a third upsampling layer, a fourth upsampling layer, a first addition layer, a second addition layer, and a third addition layer;

[0014] The input end of the first recurrent convolutional layer and the input end of the first upsampling layer are both the input ends of the layered recursive integration module, and the output end of the third addition layer is the output end of the layered recursive integration module;

[0015] The output end of the first circular convolution layer is connected to the input end of the second circular convolution layer and the input end of the second upsampling layer respectively, the output end of the first upsampling layer is connected to the input end of the second circular convolution layer, the input end of the second upsampling layer, and the input end of the first addition layer respectively, the output end of the second circular convolution layer is connected to the input end of the third circular convolution layer and the input end of the third upsampling layer respectively, the output end of the second upsampling layer is connected to the input end of the third circular convolution layer, the input end of the third upsampling layer, and the input end of the first addition layer respectively, the output end of the third circular convolution layer is connected to the input end of the fourth upsampling layer, and the input end of the second addition layer respectively, the output end of the fourth upsampling layer is connected to the input end of the third addition layer, the output end of the first addition layer is connected to the input end of the second addition layer, and the output end of the second addition layer is connected to the input end of the third addition layer.

[0016] Optionally, construct a fitting function based on all relative depths and all absolute depths, including:

[0017] calculating a first parameter and a second parameter based on all relative depths and all absolute depths;

[0018] Construct a fitting function based on the first parameter and the second parameter.

[0019] Optionally, calculating the first parameter and the second parameter based on all relative depths and all absolute depths includes:

[0020] By formula:

[0021]

[0022] Calculate a first parameter a and a second parameter b;

[0023] Where n represents the number of target power lines, x represents the relative depth corresponding to the target power line, y represents the absolute depth corresponding to the target power line, ∑x represents the statistics of the relative depth corresponding to all target power lines, ∑y represents the statistics of the absolute depth corresponding to all target power lines, ∑xy represents the statistics of the product of the absolute depth and relative depth corresponding to all target power lines, ∑x 2 A statistic representing the square of the relative depths corresponding to all target power lines.

[0024] Optionally, the fitting function is:

[0025] y * =ax * +b

[0026] Among them, y * Indicates absolute depth, x * Indicates relative depth.

[0027] Optionally, the distance between the power line to be inspected and the UAV is measured using a power line depth estimation network and a fitting function to obtain a final distance between the power line to be inspected and the UAV, including:

[0028] Obtain the final drone-photographed image of the power line to be inspected;

[0029] Segment the final image without drones to obtain the final power line image of the power line to be inspected;

[0030] The power line depth estimation network is used to estimate the distance of the final power line image and obtain the final relative depth between the power line to be inspected and the drone;

[0031] The final relative depth is substituted into the fitting function to calculate the absolute depth corresponding to the power line to be inspected, and the absolute depth is used as the final distance between the power line to be inspected and the UAV.

[0032] In a second aspect, an embodiment of the present application provides a drone ranging device for power line inspection, comprising:

[0033] An acquisition module is used to acquire drone-photographed images of multiple target power lines; the drone-photographed images are images obtained by drones used for power line inspection of target power lines;

[0034] a segmentation module, for segmenting each drone-photographed image to obtain a power line image of a target power line corresponding to the drone-photographed image;

[0035] A distance estimation module is used to estimate the distance of each power line image using a power line depth estimation network to obtain the relative depth between each target power line and the UAV; the relative depth is the estimated distance between the target power line and the UAV;

[0036] A construction module is used to obtain the absolute depth between each target power line and the UAV and construct a fitting function based on all relative depths and all absolute depths. The absolute depth is the actual distance between the target power line and the UAV, and the fitting function is used to describe the mapping relationship between relative depth and absolute depth.

[0037] The measurement module is used to measure the distance between the power line to be inspected and the UAV by using the power line depth estimation network and the fitting function to obtain the final distance between the power line to be inspected and the UAV.

[0038] In a third aspect, an embodiment of the present application provides a terminal device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the drone ranging method for power line inspection is implemented.

[0039] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned drone ranging method for power line inspection.

[0040] The above solution of the present application has the following beneficial effects:

[0041] In an embodiment of the present application, by acquiring drone images of multiple target power lines, and then segmenting the drone images for each drone image, obtaining a power line image of the target power line corresponding to the drone image, and then using a power line depth estimation network to estimate the distance of each power line image, obtaining the relative depth between each target power line and the drone, and then obtaining the absolute depth between each target power line and the drone, constructing a fitting function based on all relative depths and all absolute depths, and finally using the power line depth estimation network and the fitting function to measure the distance between the power line to be inspected and the drone, obtaining the final distance between the power line to be inspected and the drone. Among them, the image captured by the drone has rich information such as color, shape, contrast, geometry and texture details. Distance measurement based on the image can improve the recognition accuracy of the power line, thereby improving the accuracy of the distance measurement. Constructing a fitting function can represent the mapping relationship between the relative depth obtained by the power line depth estimation network and the true distance. Distance measurement of the power line to be inspected based on the fitting function can further improve the accuracy of the drone distance measurement.

[0042] Other beneficial effects of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0044] Figure 1 A flowchart of a method for measuring distance between a drone and a power line for inspection according to an embodiment of the present application;

[0045] Figure 2A schematic diagram of the structure of a power line depth estimation network provided in one embodiment of the present application;

[0046] Figure 3 A schematic diagram of the structure of a drone ranging device for power line inspection provided in one embodiment of the present application;

[0047] Figure 4 A schematic diagram of the structure of a terminal device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0048] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0049] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0050] It will also be understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0051] As used in this specification and the appended claims, the term "if" can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0052] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0053] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0054] To address the low accuracy of existing drone ranging, an embodiment of the present application provides a drone ranging method for power line inspection. The drone ranging method obtains drone-captured images of multiple target power lines, then segments each drone-captured image to obtain a power line image of the target power line corresponding to the drone-captured image. A power line depth estimation network is then used to estimate the distance of each power line image to obtain the relative depth between each target power line and the drone. The absolute depth between each target power line and the drone is then obtained. A fitting function is constructed based on all relative depths and all absolute depths. Finally, the power line depth estimation network and the fitting function are used to measure the distance between the power line to be inspected and the drone to obtain the final distance between the power line to be inspected and the drone. The drone-captured images contain rich information such as color, shape, contrast, geometry, and texture details. Image-based ranging can improve the accuracy of power line recognition and thereby improve the accuracy of ranging. The fitting function is constructed to represent the mapping relationship between the relative depth obtained by the power line depth estimation network and the actual distance. Ranging the power line to be inspected based on the fitting function can further improve the accuracy of drone ranging.

[0055] Next, an example description is given of the drone ranging method for power line inspection provided in this application.

[0056] like Figure 1 As shown, the UAV ranging method for power line inspection provided by this application includes the following steps:

[0057] Step 11: Acquire drone-photographed images of multiple target power lines.

[0058] The above-mentioned drone-photographed image is an image obtained by a drone used for power line inspection, photographing a target power line. The above-mentioned target power line is the power line being inspected by the drone.

[0059] In some embodiments of the present application, drone-photographed images of target power lines can be obtained by using a camera or other equipment carried by the drone.

[0060] Step 12: segment each drone-photographed image to obtain a power line image of the target power line corresponding to the drone-photographed image.

[0061] Specifically, a model such as a convolutional neural network can be used to segment the image taken by the drone to obtain a power line image of the target power line corresponding to the image taken by the drone.

[0062] Step 13: Use the power line depth estimation network to estimate the distance of each power line image to obtain the relative depth between each target power line and the UAV.

[0063] The relative depth above is the estimated distance between the target power line and the UAV.

[0064] Specifically, the power line image is input into the input end of the power line depth estimation network. After the power line depth estimation network is operated, the relative depth between the target power line and the drone is obtained.

[0065] like Figure 2 As shown in FIG, the power line depth estimation network includes a feature extraction module, a dynamic scale integration module, and a hierarchical recursive integration module which are connected in sequence.

[0066] The input end of the feature extraction module is the input end of the power line depth estimation network, and the output end of the hierarchical recursive integration module is the output end of the power line depth estimation network.

[0067] Exemplarily, the feature extraction module is used to extract features from the input data, and may be a deep convolutional neural network architecture ResNeXt101. The dynamic scale integration module is used to perform multi-scale learning and integration on the input data, and may be a dynamic scale integration module (DSIM). In practical applications, the feature extraction module performs feature extraction on the input power line image to obtain feature maps of multiple scales, and inputs all feature maps into the dynamic scale integration module. The dynamic scale integration module is used to perform multi-scale learning and integration on all feature maps.

[0068] The above-mentioned layered recursive integration module includes a first recurrent convolution layer, a second recurrent convolution layer, a third recurrent convolution layer, a first upsampling layer, a second upsampling layer, a third upsampling layer, a fourth upsampling layer, a first addition layer, a second addition layer and a third addition layer.

[0069] The input end of the first recurrent convolutional layer and the input end of the first upsampling layer are both input ends of the layered recursive integration module, and the output end of the third addition layer is the output end of the layered recursive integration module.

[0070] The output end of the first circular convolution layer is connected to the input end of the second circular convolution layer and the input end of the second upsampling layer respectively, the output end of the first upsampling layer is connected to the input end of the second circular convolution layer, the input end of the second upsampling layer, and the input end of the first addition layer respectively, the output end of the second circular convolution layer is connected to the input end of the third circular convolution layer and the input end of the third upsampling layer respectively, the output end of the second upsampling layer is connected to the input end of the third circular convolution layer, the input end of the third upsampling layer, and the input end of the first addition layer respectively, the output end of the third circular convolution layer is connected to the input end of the fourth upsampling layer, and the input end of the second addition layer respectively, the output end of the fourth upsampling layer is connected to the input end of the third addition layer, the output end of the first addition layer is connected to the input end of the second addition layer, and the output end of the second addition layer is connected to the input end of the third addition layer.

[0071] It should be noted that the above-mentioned first circular convolution layer, second circular convolution layer, and third circular convolution layer are all used to perform circular convolution on all input data, the above-mentioned first upsampling layer, second upsampling layer, third upsampling layer, and fourth upsampling layer are all used to upsample all input data, and the above-mentioned first addition layer, second addition layer, and third addition layer are all used to perform addition calculation on all input data.

[0072] Figure 2 The feature map represents the data form of the output data of the module or layer, that is, for the dynamic scale integration module, the first recurrent convolution layer, the second recurrent convolution layer, the third recurrent convolution layer, the first upsampling layer, the second upsampling layer, and the third upsampling layer, the data form of the output data is a feature map. Input represents the input data, that is, the power line image, and Output represents the output data, that is, the relative depth.

[0073] Step 14: Obtain the absolute depth between each target power line and the UAV, and construct a fitting function based on all relative depths and all absolute depths.

[0074] The above absolute depth is the actual distance between the target power line and the UAV, and the fitting function is used to describe the mapping relationship between relative depth and absolute depth.

[0075] In some embodiments of the present application, the actual distance between the target power line and the drone can be obtained by other distance measurement methods (such as triangulation, depth measurement, etc.). The step of constructing a fitting function based on all relative depths and all absolute depths includes:

[0076] In the first step, the first parameter and the second parameter are calculated based on all relative depths and all absolute depths.

[0077] Specifically, through the formula:

[0078]

[0079] Calculate the first parameter a and the second parameter b.

[0080] Where n represents the number of target power lines, x represents the relative depth corresponding to the target power line, y represents the absolute depth corresponding to the target power line, ∑x represents the statistics of the relative depth corresponding to all target power lines, ∑y represents the statistics of the absolute depth corresponding to all target power lines, ∑xy represents the statistics of the product of the absolute depth and relative depth corresponding to all target power lines, ∑x 2 A statistic representing the square of the relative depths corresponding to all target power lines.

[0081] In the second step, a fitting function is constructed based on the first parameter and the second parameter.

[0082] Specifically, the fitting function is:

[0083] y * =ax * +b

[0084] Among them, y * Indicates absolute depth, x * Indicates relative depth.

[0085] For example, the data of relative depth and absolute depth are shown in Table 1.

[0086]

[0087] Table 1

[0088] Calculate key statistics: ∑ x =55,∑ y =66.3,∑ xy =456.8,∑ x 2 =385, n=10, the value of the first parameter calculated by the above formula is 1.117, the value of the second parameter is 0.4865, and the fitting function is: * =1.117x * +0.4865.

[0089] Step 15: Use the power line depth estimation network and the fitting function to measure the distance between the power line to be inspected and the UAV to obtain the final distance between the power line to be inspected and the UAV.

[0090] The above-mentioned power lines to be inspected are power lines that require drone inspection.

[0091] In some embodiments of the present application, the step of measuring the distance between the power line to be inspected and the drone using the power line depth estimation network and the fitting function to obtain the final distance between the power line to be inspected and the drone includes:

[0092] The first step is to obtain the final drone-photographed image of the power line to be inspected.

[0093] For example, the final drone-photographed image can be obtained using equipment such as a camera carried by the drone.

[0094] In the second step, the final image without drone photography is segmented to obtain the final power line image of the power line to be inspected.

[0095] For example, a model such as a convolutional neural network can be used to segment the final image taken without a drone to obtain a final power line image of the power line to be inspected.

[0096] In the third step, the power line depth estimation network is used to estimate the distance of the final power line image to obtain the final relative depth between the power line to be inspected and the drone.

[0097] Specifically, the final power line image is input into the power line depth estimation network. After the power line depth estimation network is operated, the final relative depth between the power line to be inspected and the drone is obtained.

[0098] In the fourth step, the final relative depth is substituted into the fitting function to calculate the absolute depth corresponding to the power line to be inspected, and the absolute depth is used as the final distance between the power line to be inspected and the drone.

[0099] For example, in order to implement the above-mentioned method of the present application, a monocular ranging system for a power line inspection drone is designed, including a hardware unit and a software algorithm unit.

[0100] The onboard hardware unit consists of an onboard edge computing device and a drone platform. The onboard edge computing device utilizes an ARM-based computing unit, a processor architecture based on the Reduced Instruction Set Computer (RISC) architecture, characterized by high efficiency, low power consumption, and high performance. It also features a graphics processing unit (GPU) and a neural network processing unit (NPU). The drone platform uses a DJI M30 series drone. The onboard edge computing device is wired to the drone's DJI IPSDK / OSDK interface for device power supply and data exchange with the drone system. The final distance measured between the drone and the power line is pushed to the drone's remote control display via the SDK. Table 2 lists the parameters of the onboard edge computing device.

[0101]

[0102]

[0103] Table 2

[0104] The parameters of the UAV platform are shown in Table 3.

[0105] Dimensions (folded) 365×215×195mm Diagonal motor wheelbase 668mm Weight (including two batteries) 3770±10g Maximum takeoff weight 4069g Structural design scope Pan: ±105°, Pitch: -135°-60°, Roll: ±45°

[0106] Table 3

[0107] The software algorithm unit includes the methods described in steps 11-15 above. The algorithm was developed using Python and PyCharm. It relies on the OpenCV and PyTorch libraries. The algorithm first uses the OpenCV library to preprocess the captured image, including image enhancement, denoising, and color correction, to improve image quality and clarity. The method is then implemented using the PyTorch library.

[0108] It is worth mentioning that the images taken by drones contain rich information such as color, shape, contrast, geometry and texture details. Image-based ranging can improve the accuracy of power line recognition and thus improve the accuracy of ranging. Constructing a fitting function can represent the mapping relationship between the relative depth obtained by the power line depth estimation network and the actual distance. Measuring the distance of the power lines to be inspected based on the fitting function can further improve the accuracy of drone ranging.

[0109] The following is an exemplary description of the drone ranging device for power line inspection provided in this application.

[0110] like Figure 3 As shown, an embodiment of the present application provides a UAV ranging device for power line inspection, and the UAV ranging device 300 for power line inspection includes:

[0111] An acquisition module 301 is configured to acquire drone-photographed images of a plurality of target power lines; the drone-photographed images are images obtained by a drone used for power line inspection to photograph the target power lines;

[0112] The segmentation module 302 is used to segment each drone-photographed image to obtain a power line image of the target power line corresponding to the drone-photographed image;

[0113] The distance estimation module 303 is used to use the power line depth estimation network to perform distance estimation on each power line image to obtain the relative depth between each target power line and the UAV; the relative depth is the estimated distance between the target power line and the UAV;

[0114] Construction module 304 is used to obtain the absolute depth between each target power line and the UAV, and construct a fitting function based on all relative depths and all absolute depths; the absolute depth is the actual distance between the target power line and the UAV, and the fitting function is used to describe the mapping relationship between the relative depth and the absolute depth;

[0115] The measurement module 305 is used to measure the distance between the power line to be inspected and the UAV using the power line depth estimation network and the fitting function to obtain the final distance between the power line to be inspected and the UAV.

[0116] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of this application. Their specific functions and technical effects can be found in the method embodiment section and will not be repeated here.

[0117] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0118] like Figure 4As shown, an embodiment of the present application provides a terminal device, and the terminal device D10 of this embodiment includes: at least one processor D100 ( Figure 4 Only one processor is shown in the figure), a memory D101, and a computer program D102 stored in the memory D101 and executable on the at least one processor D100, wherein the processor D100 implements the steps of any of the above method embodiments when executing the computer program D102.

[0119] Specifically, when the processor D100 executes the computer program D102, it obtains drone-photographed images of multiple target power lines, then segments the drone-photographed images for each drone-photographed image to obtain a power line image of the target power line corresponding to the drone-photographed image, then uses a power line depth estimation network to estimate the distance of each power line image to obtain the relative depth between each target power line and the drone, then obtains the absolute depth between each target power line and the drone, constructs a fitting function based on all relative depths and all absolute depths, and finally uses the power line depth estimation network and the fitting function to measure the distance between the power line to be inspected and the drone to obtain the final distance between the power line to be inspected and the drone. The images taken by the drone have rich information such as color, shape, contrast, geometry, and texture details. Image-based ranging can improve the accuracy of power line recognition and thereby improve the accuracy of ranging. Constructing a fitting function can represent the mapping relationship between the relative depth obtained by the power line depth estimation network and the true distance. Ranging the power line to be inspected based on the fitting function can further improve the accuracy of drone ranging.

[0120] The processor D100 may be a central processing unit (CPU), or may be another general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor may be a microprocessor or any conventional processor.

[0121] In some embodiments, the memory D101 may be an internal storage unit of the terminal device D10, such as a hard disk or memory of the terminal device D10. In other embodiments, the memory D101 may also be an external storage device of the terminal device D10, such as a plug-in hard disk, a smart memory card (SMC, SmartMedia Card), a secure digital (SD, Secure Digital) card, a flash card, etc. equipped on the terminal device D10. Furthermore, the memory D101 may also include both an internal storage unit of the terminal device D10 and an external storage device. The memory D101 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program. The memory D101 may also be used to temporarily store data that has been output or is to be output.

[0122] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned various method embodiments can be implemented.

[0123] An embodiment of the present application provides a computer program product. When the computer program product is run on a terminal device, the terminal device can implement the steps in the above-mentioned various method embodiments when executing the computer program product.

[0124] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of the above-mentioned various method embodiments. Among them, the computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the drone ranging method device / terminal device for power line inspection, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electric carrier signal, a telecommunication signal and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk.

[0125] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0126] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0127] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles described in the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A UAV ranging method for power line inspection, characterized in that: include: Acquire drone-captured images of multiple target power lines; The drone-photographed image is an image obtained by a drone used for power line inspection photographing a target power line; Segmenting each of the drone-photographed images to obtain a power line image of the target power line corresponding to the drone-photographed image; Using a power line depth estimation network to perform distance estimation on each of the power line images to obtain a relative depth between each target power line and the drone; the relative depth is the estimated distance between the target power line and the drone; Obtaining the absolute depth between each target power line and the UAV, and constructing a fitting function based on all relative depths and all absolute depths; the absolute depth is the actual distance between the target power line and the UAV, and the fitting function is used to describe the mapping relationship between the relative depth and the absolute depth; The power line depth estimation network and the fitting function are used to measure the distance between the power line to be inspected and the drone, so as to obtain the final distance between the power line to be inspected and the drone.

2. The UAV ranging method according to claim 1, characterized in that: The power line depth estimation network includes a feature extraction module, a dynamic scale integration module, and a hierarchical recursive integration module connected in sequence; The input end of the feature extraction module is the input end of the power line depth estimation network, and the output end of the hierarchical recursive integration module is the output end of the power line depth estimation network.

3. The UAV ranging method according to claim 2, characterized in that: The layered recursive integration module includes a first recurrent convolution layer, a second recurrent convolution layer, a third recurrent convolution layer, a first upsampling layer, a second upsampling layer, a third upsampling layer, a fourth upsampling layer, a first addition layer, a second addition layer and a third addition layer; The input end of the first recurrent convolution layer and the input end of the first upsampling layer are both input ends of the layered recursive integration module, and the output end of the third addition layer is the output end of the layered recursive integration module; The output end of the first circular convolution layer is respectively connected to the input end of the second circular convolution layer and the input end of the second upsampling layer, the output end of the first upsampling layer is respectively connected to the input end of the second circular convolution layer, the input end of the second upsampling layer, and the input end of the first addition layer, the output end of the second circular convolution layer is respectively connected to the input end of the third circular convolution layer and the input end of the third upsampling layer, the output end of the second upsampling layer is respectively connected to the input end of the third circular convolution layer, the input end of the third upsampling layer, and the input end of the first addition layer, the output end of the third circular convolution layer is connected to the input end of the fourth upsampling layer, the output end of the third upsampling layer is respectively connected to the input end of the fourth upsampling layer and the input end of the second addition layer, the output end of the fourth upsampling layer is connected to the input end of the third addition layer, the output end of the first addition layer is connected to the input end of the second addition layer, and the output end of the second addition layer is connected to the input end of the third addition layer.

4. The UAV ranging method according to claim 1, characterized in that: The fitting function is constructed according to all relative depths and all absolute depths, including: calculating a first parameter and a second parameter based on all relative depths and all absolute depths; A fitting function is constructed based on the first parameter and the second parameter.

5. The UAV ranging method according to claim 4, characterized in that: The calculating of the first parameter and the second parameter according to all relative depths and all absolute depths includes: By formula: Calculate a first parameter a and a second parameter b; Where n represents the number of target power lines, x represents the relative depth corresponding to the target power line, y represents the absolute depth corresponding to the target power line, ∑x represents the statistics of the relative depth corresponding to all target power lines, ∑y represents the statistics of the absolute depth corresponding to all target power lines, ∑xy represents the statistics of the product of the absolute depth and relative depth corresponding to all target power lines, ∑x 2 A statistic representing the square of the relative depths corresponding to all target power lines.

6. The UAV ranging method according to claim 5, characterized in that: The fitting function is: y * =ax * +b Among them, y * Indicates absolute depth, x * Indicates relative depth.

7. The UAV ranging method according to claim 1, characterized in that: The method of measuring the distance between the power line to be inspected and the drone by using the power line depth estimation network and the fitting function to obtain a final distance between the power line to be inspected and the drone includes: Acquire a final drone-photographed image of the power line to be inspected; Segmenting the final image captured without a drone to obtain a final power line image of the power line to be inspected; Using a power line depth estimation network to perform distance estimation on the final power line image to obtain a final relative depth between the power line to be inspected and the drone; The final relative depth is substituted into the fitting function to calculate the absolute depth corresponding to the power line to be inspected, and the absolute depth is used as the final distance between the power line to be inspected and the UAV.

8. A UAV distance measuring device for power line inspection, characterized in that: include: an acquisition module, used to acquire drone-photographed images of multiple target power lines; The drone-photographed image is an image obtained by a drone used for power line inspection photographing a target power line; a segmentation module, configured to segment each of the drone-photographed images to obtain a power line image of the target power line corresponding to the drone-photographed image; a distance estimation module, configured to perform distance estimation on each of the power line images using a power line depth estimation network to obtain a relative depth between each target power line and the drone; the relative depth being the estimated distance between the target power line and the drone; a construction module for obtaining the absolute depth between each target power line and the UAV, and constructing a fitting function based on all relative depths and all absolute depths; the absolute depth is the actual distance between the target power line and the UAV, and the fitting function is used to describe the mapping relationship between the relative depth and the absolute depth; The measuring module is used to measure the distance between the power line to be inspected and the UAV by using the power line depth estimation network and the fitting function to obtain the final distance between the power line to be inspected and the UAV.

9. A terminal device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the drone ranging method for power line inspection according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the drone ranging method for power line inspection according to any one of claims 1 to 7 is implemented.

Images