Unmanned aerial vehicle-mounted infrared and ultraviolet imager integrated system

By equipped with an infrared ultraviolet imager integrated system, the problems of low efficiency and high risks of traditional manual inspection are solved, efficient and accurate detection of insulators are achieved, the risk of flashover accidents is reduced, and the operation of the power system is improved Dimensional reliability.

CN118914766APending Publication Date: 2024-11-08STATE GRID HUBEI EXTRA HIGH VOLTAGE CO
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410840006.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-26
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The traditional manual inspection method is inefficient, has high working strength and high risk when detecting insulators, and cannot meet the needs of efficient and accurate detection of insulators. Especially in high-voltage transmission systems, the potential threat of flashover accidents caused by zero-value or low-value insulators. Difficult to prevent.

Method used

Design a drone equipped with an infrared ultraviolet imager integrated system, combining infrared imager and ultraviolet imager, and realize real-time video playback, status display and insulators through detector components, signal processing circuits, image circuits and abnormality detection modules. Failure positioning improves detection accuracy and efficiency.

Benefits of technology

It realizes efficient and accurate detection of insulators and electrical equipment, reduces troubleshooting time and labor costs, and improves the operation and maintenance reliability and safety of the power system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118914766A_ABST
    Figure CN118914766A_ABST
Patent Text Reader

Abstract

The invention discloses an unmanned aerial vehicle carried infrared and ultraviolet imager integrated system, which comprises an infrared imager carried on an unmanned aerial vehicle, and is characterized in that the infrared imager comprises a detector assembly, a signal processing circuit assembly, an image circuit assembly and a servo control assembly; the ultraviolet imager is carried on the unmanned aerial vehicle and comprises an optical system, an optical coupling device, a vertical conversion driving chip, an AD (analog-to-digital) conversion chip and an FPGA (field programmable gate array); the real-time video playing module is used for displaying a real-time picture shot by the unmanned aerial vehicle load camera; the unmanned aerial vehicle state display module is used for displaying whether the unmanned aerial vehicle can take off in the current state; and the anomaly detection module is used for capturing an abnormal thermal signal and a corona discharge signal of the insulator so as to realize rapid positioning of a fault. Compared with the existing single detection technology, the method has the advantages of higher detection precision, stronger real-time performance and better automation capability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of insulator detection, and in particular to an integrated system of an infrared and ultraviolet imager carried by a drone. Background Art

[0002] As a key component of power transmission, insulators on transmission lines are exposed to the natural environment for a long time, and are subjected to severe tests of electromechanical loads and extreme climates (such as sunshine, rain, and temperature changes). The combined effect of these factors will cause the insulation performance or mechanical properties of the insulators to deteriorate, resulting in zero-value or low-value insulators. This phenomenon is called insulator degradation. In a high-voltage transmission system, if there are zero-value or low-value insulators in the insulator string, it is equivalent to that part of the insulation of the insulator string is short-circuited, and its overall creepage distance is correspondingly reduced. The flashover probability of the insulator string containing degraded insulators is greatly increased. When a flashover occurs, the arc may pass through the inside of the zero-value insulator, and its steel cap will often explode or detach, resulting in serious accidents such as string drop and conductor falling, which poses a potential threat to power supply reliability. Therefore, regular and accurate insulation status detection of insulators has become an indispensable part of maintaining the reliable operation of the power grid. However, the traditional manual inspection method has problems such as low detection efficiency, high work intensity, and high risk, and cannot meet the needs of efficient and accurate detection of insulators. Summary of the invention

[0003] In order to solve the above problems, the present invention provides an integrated system of an infrared and ultraviolet imager carried by a drone.

[0004] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] An integrated system of an infrared and ultraviolet imager carried by a drone, comprising:

[0006] An infrared imager mounted on a drone, the infrared imager comprising: a detector component, a signal processing circuit component, an image circuit component, and a servo control component;

[0007] And an ultraviolet imager carried on a drone, the ultraviolet imager comprising: an optical system, an optical coupling device, a vertical conversion driver chip, an AD digital-to-analog conversion chip and an FPGA;

[0008] Real-time video playback module: used to display the real-time images taken by the drone’s payload camera;

[0009] Drone status display module: used to display whether the drone is ready to take off in its current state;

[0010] Abnormal detection module: used to capture abnormal thermal signals and corona discharge signals of insulators to quickly locate faults.

[0011] Further: the detector assembly comprises:

[0012] Detector, preamplifier circuit, AD digital-to-analog conversion chip and readout circuit.

[0013] Further: the image circuit includes FPGA, AD components, storage components, focal temperature chips, DC / DC and LDO conversion chips.

[0014] Further: the anomaly detection module includes:

[0015] Data receiving module: performs infrared imaging on the insulators on the overhead line and collects infrared images; performs ultraviolet imaging on the insulators and collects ultraviolet images;

[0016] Data preprocessing module: spatial registration and denoising of infrared and ultraviolet images;

[0017] Feature extraction module: extracts temperature anomaly areas from infrared images and marks possible low-value insulator locations; extracts corona discharge areas from ultraviolet images and marks areas with higher discharge intensity;

[0018] Data fusion module: fuses infrared and ultraviolet feature maps to generate a comprehensive anomaly indication map and eliminate noise points;

[0019] Fault location module: calculates the abnormal distribution of each shed area and marks the faulty shed.

[0020] Compared with the prior art, the present invention has the following technical advances:

[0021] Infrared imagers can detect the temperature distribution on the surface of equipment, and are particularly suitable for capturing abnormal thermal signals of insulators, and can detect potential faults caused by overheating. Ultraviolet imagers can detect corona discharge signals, and are suitable for identifying partial discharge problems in high-voltage equipment, which helps to detect potential risks of electrical equipment in advance.

[0022] The real-time video playback module enables operators to view the images taken by the drone in real time, providing instant feedback, which helps to detect problems in a timely manner and take measures. The drone status display module ensures that operators perform detection in a safe and appropriate state, improving the reliability and safety of detection. The abnormality detection module combines infrared and ultraviolet imaging technology to quickly capture and analyze abnormal signals of insulators and electrical equipment, improving the accuracy and efficiency of fault detection. The present invention can quickly locate the detected abnormal signals, reducing the time and labor costs of troubleshooting. By integrating the detector component, signal processing circuit component, image circuit component, servo control component, and the optical system of the ultraviolet imager, optical coupling device, vertical conversion driver chip, AD digital-to-analog conversion chip and FPGA into the drone, an efficient and automated detection system is formed.

[0023] In summary, compared with the existing single detection technology, the present invention has higher detection accuracy, stronger real-time performance and better automation capability, and provides a more reliable and efficient technical means for the operation and maintenance of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0025] In the attached picture:

[0026] Figure 1 This is a system diagram of the infrared thermal imager of the present invention;

[0027] Figure 2 is a system diagram of the preamplifier circuit of the present invention;

[0028] Figure 3 is a system diagram of the readout circuit of the present invention;

[0029] Figure 4 is a system diagram of the image circuit of the present invention;

[0030] Figure 5 The software execution flow chart of the infrared detector component of the present invention;

[0031] Figure 6 It is a composition diagram of the non-uniformity correction capability of the present invention;

[0032] Figure 7 Enhanced algorithm data flow chart for the present invention;

[0033] Figure 8 It is a schematic diagram of the structure of the ultraviolet imager of the present invention;

[0034] Fig. 9 This is a schematic diagram of a real-time video playback module of the present invention;

[0035] Fig.10 This is a schematic diagram of the drone status display module of the present invention;

[0036] Fig.11 This is a schematic diagram of an abnormality detection module of the present invention; DETAILED DESCRIPTION

[0037] The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. The embodiments of the present invention will be described below in conjunction with the accompanying drawings.

[0038] The present invention discloses an integrated system of an infrared and ultraviolet imager mounted on a drone, comprising:

[0039] An infrared imager mounted on a drone, the infrared imager comprising: a detector component, a signal processing circuit component, an image circuit component, and a servo control component;

[0040] And an ultraviolet imager carried on a drone, the ultraviolet imager comprising: an optical system, an optical coupling device, a vertical conversion driver chip, an AD digital-to-analog conversion chip and an FPGA;

[0041] Real-time video playback module: used to display the real-time images taken by the drone’s payload camera;

[0042] Drone status display module: used to display whether the drone is ready to take off in its current state;

[0043] Abnormal detection module: used to capture abnormal thermal signals and corona discharge signals of insulators to quickly locate faults.

[0044] Specifically, infrared imagers such as Figure 1 As shown in the figure, infrared imaging technology is a technology that uses infrared radiation emitted by an object to obtain image information. Its principle is based on the phenomenon of thermal radiation. All objects emit infrared radiation at a certain intensity, and its intensity is related to the temperature of the object. Infrared imaging technology uses infrared sensors to capture this radiation and convert it into corresponding electrical signals. After signal processing and image reconstruction, it finally forms an infrared image visible to the human eye.

[0045] The infrared thermal imager adopts modular design, which consists of detector components, signal processing circuit components, image circuit components, and servo control components. The signal processing circuit components are composed of A / D conversion circuits, image circuits, and power conversion circuits.

[0046] The detector assembly consists of a detector preamplifier circuit, an AD digital-to-analog conversion chip, and a readout circuit. The readout circuit is one of the key components of the infrared imaging system. Its main function is to preprocess the detector signal and provide an output channel. The preamplifier has always been the key to the development of the readout circuit. Complementary metal oxide semiconductor transistors (CMOS) have the advantages of low power consumption, low temperature resistance, standard IC process, and low cost, and have gradually become the mainstream technology for readout circuits.

[0047] Photoelectric detection is essentially a process of converting and amplifying light signals. Complex amplification mechanisms are completed outside the pixel, while simple amplification mechanisms can be completed inside the pixel. Pre-amplification processing (such as Figure 2 As shown). The sensitivity of the infrared system mainly depends on the detector and preamplifier. The noise of the preamplifier is the main noise source of the entire readout circuit. The charge injection efficiency of the detector signal is equivalent to the input noise in the detector and preamplifier circuit as a specific noise source. The optimal design of the preamplifier can be obtained based on the gap design of these noise sources.

[0048] The preamplifier forms a load resistance and impedance. In order to effectively transmit the signal and energy, it must work in an impedance matching state. Even if the detector resistance is equal to the load resistance, the energy transferred is maximum and the injection efficiency is also high.

[0049] After the reference voltage is determined, the charge storage capacity depends on the size of the integral capacitor. People generally hope to increase the integral capacitor while reducing the center distance. Usually the integral capacitor is built into the pixel circuit. Due to area limitations, it is impossible to make it very large. Therefore, a structure that reads the charge by directing the photocurrent to the charge amplifier outside the pixel has emerged, which can optimize the charge processing capability of the pixel. Figure 3 ) can directly determine the dynamic range of the focal plane array. Having a large dynamic range is one of the requirements for high-performance readout circuits. One approach is to introduce a current source with a gate cathode amplifier that acts as a "charge removal" in each pixel, apply a bias voltage (GPOL) to the current source throughout the chip, and subtract a fixed amount representing the background effect from each pixel to fully utilize the dynamic range. The output swing of the readout circuit refers to the swing range of the signal voltage output under different backgrounds. This value (VR) will directly affect the dynamic range of the focal plane array. When the output swing is too small, the detector is prone to saturation and the integration time is limited.

[0050] Combination Figure 3The readout processing circuit has no ability to suppress VR noise. VR noise will be 100% transmitted to the output of the processing circuit, and the VR input signal is required to have a certain driving capability. The application board needs to reduce this noise. The test process is to first test the mean square noise value V of the (VR signal source + AFE evaluation circuit) n,0 2 , then connect to the processing circuit, connect the VR signal source + processing circuit + AFE evaluation board, and test the mean square noise V n,1 2 , by calculating the square root of the difference between the two mean square noises After that, the RMS noise of the processing circuit is obtained.

[0051] The function of the image circuit is to perform A / D conversion on the output of the analog signal board to form an infrared digital image. The FPGA in the movement mainboard uses algorithms to perform non-uniformity correction, bad pixel removal, image enhancement, dimming and other algorithms on the infrared digital image, and then output the image to the interface board.

[0052] like Figure 4 As shown, the image circuit is mainly composed of: FPGA, AD components, storage components, focal temperature chip, DC / DC and LDO conversion chip, while retaining the signal board / refrigeration machine / interface board interface. The interface board external interface function is expanded to the maximum extent. The image circuit power conversion circuit generally adopts high-efficiency DC / DC circuit, and through comprehensive optimization, reasonable selection of related devices, and careful configuration of circuit parameters, the DC / DC conversion circuit basically works at its highest conversion efficiency state. The average conversion efficiency of the entire system is as high as more than 90%. At the same time, the main heat-generating components are placed on one side to facilitate the heat dissipation of the structure.

[0053] The image processing algorithm of the image circuit is as follows Figure 5 Shown:

[0054] Ideally, when the infrared focal plane detector is irradiated with uniform light, the video output signal amplitude of each pixel should be completely consistent. In fact, due to the non-uniformity of the semiconductor material used to make the device, the non-uniformity of doping (proportions of various components), mask errors, and the instability of production process control, the video output amplitude will be uneven. That is, different pixels of the infrared focal plane array have different video output signal amplitudes under the same uniform incident radiation. This is the so-called non-uniformity (NU) of the infrared focal plane array response.

[0055] The non-uniformity of the detector is affected by many factors. In different environments, the non-uniformity characteristics of the detector are different, which is difficult to express with a fixed mathematical formula. Therefore, we use the segmented calibration method to perform non-uniformity correction. This method infers the non-uniformity characteristics through experimental data and has been proven to be a very effective non-uniformity correction method in practice.

[0056] In order to prevent the influence of the non-uniformity of the infrared detector itself, the infrared imaging component software adopts a two-point correction non-uniformity correction method. This method calculates the correction coefficient K and collects the background data B in real time, and calculates the corrected data according to Y=K×(X_B), where X represents the data before correction, K is the correction coefficient, B is the background data collected in real time, and Y is the result after correction. The infrared imaging component has automatic correction and manual correction functions. The infrared imaging component software automatically performs a non-uniformity correction when it is turned on; the manual correction function can be completed by receiving serial port commands. The non-uniformity correction capability consists of the following components: Figure 6 shown.

[0057] 1) Two-point calibration

[0058] FPGA realizes two-point correction through Y16=K(X16_B)+Baver calculation. K value and B value are read from DDR respectively, and Baver is read from NIOS by dual-port RAM.

[0059] 2) Correction coefficient collection and storage

[0060] Each pixel corresponds to a correction coefficient K. The current double reference radiation source method is used to calculate the correction coefficient. The same detector is facing radiation sources of different temperatures, and two sets of response data are collected respectively. The temperature response characteristics of each pixel of the detector, that is, the correction coefficient K value, are calculated according to the following formula.

[0061]

[0062] in, is the high temperature background mean; is the low temperature background average; AD2 (x,y) AD1 is the high temperature background value corresponding to the specific pixel point; (x,y) is the low temperature background value corresponding to the specific pixel.

[0063] In order to observe the weak targets in the infrared image more clearly, the infrared imaging component provides an electronic zoom function, which can observe the details of the image by magnifying the central 640×512 image by 2 or 4 times, thereby improving the image recognition ability. In order to enhance the image details, the image enhancement algorithm is used for processing, and the high-frequency and low-frequency separation are performed to perform high-frequency compression and low-frequency dimming respectively, which can better enhance the image details while meeting the visual observation.

[0064] The core part of the enhancement algorithm is low-frequency filtering. Its basic idea is to decompose the image into a low-frequency image and a detail image in two scales, then compress the low-frequency image and enhance the detail image, and finally fuse the compressed low-frequency image and the enhanced detail image to obtain a low dynamic range image that retains image details. It can retain image details as much as possible while completing the compression of large dynamic range digital images. The specific data flow of the algorithm is as follows: Figure 7 shown.

[0065] When the image dynamic range is large, the Y16 image contains more scene information. The dimming algorithm cannot map all the information in the Y16 image to the Y8 image for display, which will cause the loss of detail information in the image. Therefore, the IDE algorithm is used to extract the existing detail information in the Y16 image, enhance the image, and improve the image clarity.

[0066] For images with a very small dynamic range, the IDE algorithm can only extract limited high-frequency information, and the enhancement effect is not obvious.

[0067] The IDE enhancement algorithm itself cannot achieve the purpose of highlighting the target energy, but it can only extract detail information that exists in the image but cannot be displayed by simple dimming. By enhancing the image details, the image can achieve a good visual effect when displayed.

[0068] After receiving the image freeze command, the image data is no longer written to DDR, and then the DDR is controlled to read the image of the current frame repeatedly. The image is first cached in DDR, and the vertical or horizontal flip of the image display is achieved by controlling the address of reading DDR. Polarity switching can reverse the display of black and white images with a grayscale of 256 levels, that is, "white hot" and "black hot" display. This function is convenient for observing targets and highlighting key points. Pseudo-color distinguishes the 14-bit original grayscale data between normal temperature background and high temperature target, sets the demarcation threshold according to the integral file, optimizes the high temperature display, and maps pseudo-color images.

[0069] For the image obtained by infrared temperature measurement, the image is processed. First, the grayscale of the image is grayed and segmented. In order to extract the image quantization parameters, each discharge area image needs to be segmented from the ultraviolet image. At present, the main algorithms for image segmentation are: edge operator method, region growing method and threshold method. Then mathematical morphological filtering is performed. The basic morphological operations include corrosion and expansion. The opening operation can eliminate isolated points higher than its neighboring points, eliminate the thin points between the boundaries of objects without significantly changing their areas, and play a smoothing role; the closing operation can eliminate isolated points lower than its neighboring points, fill and smooth the boundaries of neighboring objects without significantly changing their areas. Considering that the discharge area is usually circular, the selected structural element is a circular structure. The next step is to eliminate the small area area. The small area area elimination algorithm of the binary image is used to remove the interference points in the image. Finally, the regional contour is extracted. To obtain the quantitative parameters of the discharge area image, the contour boundary points of each discharge area need to be extracted. Since there are often more than one discharge point in the ultraviolet image, the coordinate extraction of the multi-region boundary contour points is a core link in this algorithm, which can be implemented by the multi-region boundary tracking algorithm.

[0070] Ultraviolet imaging detection is a non-contact leakage detection technology. When there is an abnormality in the electrical performance of the insulator, it will cause surface discharge, ionize atmospheric molecules, and release ultraviolet radiation. The ultraviolet imager separates and detects ultraviolet radiation in the solar radiation blind area through a solar-blind spectrometer, and uses dual-channel fusion technology to obtain a visual image after the fusion of visible light and ultraviolet radiation, and accurately locates the corona position. The ultraviolet imaging detection method uses an ultraviolet imager to take ultraviolet images or videos at a long distance, and extracts characteristic parameters such as the amount of photons or the spot area to characterize the discharge amount of the insulating equipment and determine the operating status of the insulator. Compared with traditional detection methods, ultraviolet image detection has the advantages of easy operation, safety, high sensitivity, strong anti-interference ability, and the ability to accurately locate the discharge position.

[0071] UV detectors are the basis for the realization of UV detection technology. Commonly used UV detection devices include photomultiplier tubes (PMTs), second-generation enhanced charge-coupled devices (ICCDs) and electron multiplier charge-coupled devices (EMCCDs). Among them, UV detectors can accurately receive UV radiation and have the advantages of strong recognition ability, small size, light weight and high detection sensitivity. They have been widely used in UV detection systems at home and abroad.

[0072] like Figure 8 As shown in the figure, the composition of the ultraviolet imager includes an optical system, an optical coupling device, a vertical conversion driver chip, an AD digital-to-analog conversion chip, and an FPGA. From the analysis of the CCD principle, it can be concluded that in the focal plane circuit including the CCD and ADC chips, the design points of the timing signal and the driver circuit are as follows:

[0073] 1) Generate charge shift drive signals: including vertical shift signals for shifting charges from pixels to shift registers, horizontal shift signals for shifting charges out of horizontal shift registers in sequence, and reset pixel control signals (electronic shutters) for releasing pixel collected charges, etc.;

[0074] 2) Generate multiple bias voltages for CCD: multiple bias voltages including positive and negative voltages, including some high voltage bias requirements (>20V);

[0075] 3) Generate ADC timing control signal: It belongs to the same clock domain as the CCD clock signal, including ADC control signal, CDS double sampling point signal, black area clamping signal, blank pixel marking signal and output data clock, etc. Among the KLI2113 chip pins, LOGn is an electronic shutter, which is used to control the charge integration time of the pixel; TG1 and TG2 are vertical shift transmission gates, which are used to transfer the charge in the pixel as a whole to the horizontal shift register: D1 and D2 (hereinafter referred to as H1, H2) are horizontal shift register controllers, which are used to output the charge in the pixel to the output end in sequence; D2s is the pixel merging (Bimning) control pin Reset is the output end capacitor reset signal control pin: VIDn is the output signal pin, and the analog voltage signal generated by charge conversion is output from here: ID.

[0076] Real-time video playback module such as Fig. 9 Shown:

[0077] 1) If Fig. 9 Display the real-time images taken by the drone's payload camera. Push the video stream through the payload camera, and then decode and view it through the remote control.

[0078] 2) Click EO to switch the current camera video stream, which supports the playback of visible light, ultraviolet light, infrared light, ultraviolet + visible light fusion, and ultraviolet + infrared fusion video streams.

[0079] 3) Click the third button in the upper right corner (from right to left) to control the payload, and control its rotation up, down, left, and right to control the rotation of the payload camera screen.

[0080] 4) Click the second button in the upper right corner (from right to left) to quickly return the load to its original position.

[0081] 5) Click the photo button on the right to take a photo, click Switch to switch to video recording mode, and click Start and End to complete the recording and save it to the drone's memory card.

[0082] 6) Slide and select the focus on the left to control the focus ratio of the loaded camera.

[0083] The implementation steps are as follows:

[0084] Push to the bitstream channel (Viedeo Channels) through PSDK, and then obtain the corresponding video stream through the bitstream channel on the MSDK side. You can use IVideo Decoder to decode and display it on the surface View. Send command data to the payload through the management class IPayload Manager provided by the drone SDK. It can also be called an instruction. This instruction is the corresponding instruction protocol negotiated by both PSDK and MSDK. Then send instructions to the payload through this wireless communication method, and the payload will perform corresponding operations after receiving it.

[0085] The drone status display module is as follows: Fig.10 Shown:

[0086] 1) The status bar at the top of the remote control interface can show whether the drone is currently ready to take off.

[0087] 2) The top status bar shows the number of satellites connected to the drone, the power of the remote controller, the mobile network signal, the WIFI signal, and the battery power of the drone.

[0088] 3) The bottom of the interface displays: D: horizontal distance from the return point H: flight altitude H(s): horizontal speed V(s): vertical speed VPS: absolute height of the aircraft from the ground.

[0089] The implementation steps are as follows:

[0090] 1) The small controls in the corresponding status bar of the remote control are implemented according to the small controls provided by DJI

[0091] 2) Regarding the custom control of the drone flight status, the Flight Controller Key related flight control API in MSDK is called through polling at intervals of 1s to refresh and display the current flight information.

[0092] Anomaly detection module such as Fig.11 Shown:

[0093] 1) Video playback data includes gain value, laser ranging, temperature, humidity, and the number of photons in the red selected box

[0094] 2) Single-point temperature measurement: Click a point in the video on the screen to display the highest temperature of the corresponding point in the video stream.

[0095] 3) Area test: A rectangular frame is generated by sliding your finger on the screen, and the highest and lowest temperatures in the rectangular frame can be displayed in the video stream.

[0096] 4) You can also slide your finger to generate a rectangular frame, which can display the number of photons in the red frame in the video stream.

[0097] The implementation steps are as follows:

[0098] The finger slides to generate a rectangle, which is implemented by a custom View. After the screen touches, according to the touch event, click, slide and then automatically draw a rectangular area with a red border according to the coordinates. Then the coordinates of the four fixed points in the red area are used as command parameters and transmitted to PSDK through wireless communication in DJI MSDK. Then PSDK calculates and counts the temperature or number of photons in the rectangular area according to the received command, and adds the information to the video stream. After the remote control receives the video stream, it is decoded by the decoder and displayed to the user.

[0099] The anomaly detection module includes:

[0100] Data receiving module: performs infrared imaging on the insulators on the overhead line and collects infrared images; performs ultraviolet imaging on the insulators and collects ultraviolet images;

[0101] Data preprocessing module: spatial registration and denoising of infrared and ultraviolet images;

[0102] Feature extraction module: extracts temperature anomaly areas from infrared images and marks possible low-value insulator locations; extracts corona discharge areas from ultraviolet images and marks areas with higher discharge intensity;

[0103] Data fusion module: fuses infrared and ultraviolet feature maps to generate a comprehensive anomaly indication map and eliminate noise points;

[0104] Fault location module: calculates the abnormal distribution of each shed area and marks the faulty shed.

[0105] The data receiving module includes:

[0106] An infrared imager is used to image the insulators on overhead lines. The infrared image is represented by I IR (x, y), where (x, y) are the pixel coordinates in the image;

[0107] The same insulator is imaged using an ultraviolet imager, and the ultraviolet image is represented as I UV (x, y), where (x, y) are the pixel coordinates in the image.

[0108] The data preprocessing module includes:

[0109] From the infrared image I IR (x, y) and UV image I UV Extract feature points from (x, y), let {p i} and {q i} respectively represent the feature point sets in the infrared image and ultraviolet image;

[0110] For each feature point {p i} and {q i}, calculate its local feature descriptor and

[0111] Descriptors using feature points and Match and find the corresponding feature point pairs in the infrared image and ultraviolet image (p i ,q i );

[0112] According to the matched feature point pairs (p i ,q i ), the following transformation model is used to describe the transformation relationship from infrared image to ultraviolet image:

[0113]

[0114] Among them, a, b, c, d are the parameters of the transformation matrix H, t x and t y is the translation parameter; the transformation matrix H makes the feature point p in the infrared image i After transformation H, it can correspond to the feature point q in the ultraviolet image i :

[0115] p′ i =H·p i

[0116] Among them, p′ i is the position of the infrared image feature point after transformation; where the transformation matrix H is:

[0117] H=arg min H ∑ i ||q i -H·p i ||

[0118] For infrared image I IR (x, y) and UV image I UV (x, y) are filtered respectively;

[0119]

[0120]

[0121] Among them, K σ (x′, y′) is the kernel function used to calculate the weighted average, and Ω is the neighborhood around the current pixel (x, y);

[0122] Kσ The calculation formula for (x′, y′) is:

[0123]

[0124] Here, σ is the standard deviation of the Gaussian filter, which controls the smoothness of the filter.

[0125] The feature extraction module includes:

[0126] Extract the abnormal temperature area from the denoised infrared image, set the threshold T, and transform the infrared image I IR_filtered (x, y) is divided into abnormal temperature area and normal area:

[0127]

[0128] A binary image is obtained, in which a pixel value of 1 indicates an area with abnormal temperature, and a pixel value of 0 indicates an area with normal temperature.

[0129] After obtaining the binary image, the following equation is used to detect the edge of the temperature abnormality area in the image:

[0130]

[0131] Among them, G x (x, y) and G y (x, y) is the gradient of the image in the horizontal and vertical directions;

[0132] According to the detected edge image E(x, y), find all edge pixels representing the boundaries of the temperature anomaly area to extract the infrared feature map F IR (x, y);

[0133] Extract the corona discharge area from the denoised UV image and set the threshold T UV , the ultraviolet image I UV_filtered (x, y) is divided into areas with higher discharge intensity and other areas:

[0134]

[0135] A binary image is obtained, where the pixel value of 1 represents the area with higher discharge intensity, and the pixel value of 0 represents other areas;

[0136] After obtaining the binary image, the following equation is used to detect the edge of the temperature abnormality area in the image:

[0137]

[0138] Among them, G x (x, y) and Gy (x, y) is the gradient of the image in the horizontal and vertical directions;

[0139] According to the edge image E(x, y) obtained by edge detection, all edge pixels representing the boundary of the corona discharge area are found to extract the ultraviolet feature map F UV (x, y).

[0140] The data fusion module includes:

[0141] Generate a comprehensive anomaly indication map:

[0142] F combined (x, y) = F IR (x,y)OR F UV (x, y)

[0143] Expand each foreground pixel to its surrounding pixels:

[0144]

[0145] Among them, A is the binary comprehensive abnormal indication map F combined , B is the expansion kernel.

[0146] The fault location module includes:

[0147] The insulator is divided into several shed areas. Assume that there are n sheds in the insulator image, and the area of ​​each shed in the image is R. i (i=1, 2, ..., n);

[0148] For each shed area R i , calculate the number of abnormal pixels inside it, assuming A i Represents the number of abnormal pixels in the i-th umbrella skirt area:

[0149]

[0150] Among them, F combined (x, y) is 1, indicating that the pixel is an abnormal pixel, and 0, indicating that the pixel is a normal pixel;

[0151] Set the threshold T fault , if a certain umbrella skirt area R i The number of abnormal pixels in A i Greater than the threshold T fault , then mark the shed as a faulty shed:

[0152]

[0153] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the claims of the present invention.

Claims

1. An integrated system of infrared and ultraviolet imagers carried by unmanned aerial vehicles, characterized in that: include: An infrared imager mounted on a drone, the infrared imager comprising: a detector component, a signal processing circuit component, an image circuit component, and a servo control component; And an ultraviolet imager carried on a drone, the ultraviolet imager comprising: an optical system, an optical coupling device, a vertical conversion driver chip, an AD digital-to-analog conversion chip and an FPGA; Real-time video playback module: used to display the real-time images taken by the drone’s payload camera; Drone status display module: used to display whether the drone is ready to take off in its current state; Abnormal detection module: used to capture abnormal thermal signals and corona discharge signals of insulators to quickly locate faults.

2. The integrated system of infrared and ultraviolet imagers mounted on unmanned aerial vehicles according to claim 1, characterized in that: The detector assembly comprises: Detector, preamplifier circuit, AD digital-to-analog conversion chip and readout circuit.

3. The infrared and ultraviolet imager integrated system mounted on a drone according to claim 2, characterized in that: The image circuit includes FPGA, AD components, storage components, focal temperature chips, DC / DC and LDO conversion chips.

4. The integrated system of infrared and ultraviolet imagers mounted on unmanned aerial vehicles according to claim 3, characterized in that: The anomaly detection module comprises: Data receiving module: performs infrared imaging on the insulators on the overhead line and collects infrared images; performs ultraviolet imaging on the insulators and collects ultraviolet images; Data preprocessing module: spatial registration and denoising of infrared and ultraviolet images; Feature extraction module: extracts temperature anomaly areas from infrared images and marks possible low-value insulator locations; extracts corona discharge areas from ultraviolet images and marks areas with higher discharge intensity; Data fusion module: fuses infrared and ultraviolet feature maps to generate a comprehensive anomaly indication map and eliminate noise points; Fault location module: calculates the abnormal distribution of each shed area and marks the faulty shed.

Citation Information

Cited By

  • Complex environment target identification approaching method based on dual-band imaging under satellite platform

    CN120722376A

  • Insulating part interface defect detection method and device and electronic equipment

    CN121231959A