A laser obstacle removal remote control system for remote obstacle removal of overhead lines

Abstract

The present invention discloses a laser obstacle removal remote control system for remote obstacle removal of overhead lines, which relates to the field of laser obstacle removal technology, in order to solve the problem of inaccurate obstacle acquisition by laser obstacle removal equipment. The integration and optimization of steps such as image processing, dynamic adjustment and batch normalization of the present invention can significantly improve the processing efficiency of image data, so that the laser obstacle removal equipment can respond and identify obstacles more quickly, and the obstacle features in the optimized image data are trained and calculated with the obstacle model using a computer learning algorithm, which can greatly improve the system's adaptability and robustness. By continuously learning new obstacle features, the system can gradually improve its recognition performance, calculate the characteristics and operation mode of the obstacle through a particle filtering method, and adjust the direction of the laser beam in real time according to the calculation results, thereby ensuring accurate tracking of the obstacle position.

Description

Technical Field

[0001] The invention relates to the technical field of laser obstacle removal, and in particular to a laser obstacle removal remote control system for remote obstacle removal of overhead lines. Background Art

[0002] Laser obstacle removal refers to the use of laser technology to remove obstacles remotely, efficiently and accurately. It mainly achieves the purpose of removing obstacles by emitting high-energy laser beams to instantly heat and evaporate the surface of the target object.

[0003] The Chinese patent with announcement number CN208890233U discloses a long-distance laser obstacle removal system, which mainly transmits the target image to the smart device through the camera to form a field of view containing the target, wherein the automatic tracking module collects the data information of the foreign object, realizes the automatic tracking function of the foreign object through the control center module, and realizes the automatic obstacle removal function with the cooperation of the control pan / tilt; at the same time, the area safety detector in the laser obstacle removal instrument detects that people, animals or other obstacles that affect the laser emission enter the emission safety area, and the control center module automatically cuts off the laser emission and sounds an alarm to ensure the safety of the surrounding personnel. Although the above patent solves the problem of equipment obstacle removal, there are still the following problems in actual operation:

[0004] 1. When the laser obstacle removal equipment collects images of the obstacle removal area, it does not further optimize and process the collected images, resulting in inaccurate obstacle acquisition.

[0005] 2. The obstacles collected by the laser obstacle removal equipment are not accurately tracked, resulting in the inability to remove the obstacles after they are offset.

[0006] 3. No further confirmation is made based on the surrounding environment and the actual distance of the obstacle, which reduces the safety of the laser obstacle removal equipment. Summary of the invention

[0007] The purpose of the present invention is to provide a laser obstacle removal remote control system for remote obstacle removal of overhead lines. The integration and optimization of steps such as image processing, dynamic adjustment and batch normalization can significantly improve the processing efficiency of image data, so that the laser obstacle removal equipment can respond and identify obstacles more quickly. The obstacle features in the optimized image data are trained and calculated with the obstacle model using a computer learning algorithm, which can greatly improve the adaptive ability and robustness of the system. By continuously learning new obstacle features, the system can gradually improve its recognition performance, calculate the characteristics and operation mode of the obstacle through the particle filtering method, and adjust the direction of the laser beam in real time according to the calculation results, thereby ensuring the accurate tracking of the obstacle position, which can solve the problems in the prior art.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A laser obstacle removal remote control system for remote obstacle removal of overhead lines, comprising:

[0010] Laser obstacle removal equipment confirmation detection unit, used for:

[0011] Perform functional tests on each component in the laser obstacle removal equipment, and determine whether the laser obstacle removal equipment is in normal operation based on the test results;

[0012] Obstacle recognition and analysis unit, used for:

[0013] The laser obstacle removal equipment collects image data of the environment in the obstacle removal area, optimizes and analyzes the collected image data, determines the position of the obstacle in the obstacle removal area according to the analysis results, and marks the position of the obstacle in the obstacle removal area as obstacle position data;

[0014] Obstacle targeting and tracking unit for:

[0015] The laser obstacle removal device aims at the obstacle location data, and then tracks the obstacle location data. At the same time, the obstacle location data is monitored in real time during tracking, and the monitoring data is marked as obstacle tracking data;

[0016] Laser rangefinder controlled clearing unit for:

[0017] The laser obstacle removal equipment transmits obstacle tracking data to the control center for distance measurement. After the distance measurement is completed, the laser clearance parameters are adjusted according to the characteristics and distance of the obstacle. After the laser clearance parameters are adjusted, laser clearance is performed.

[0018] Preferably, the components in the laser obstacle removal equipment include an auxiliary red dot sight, a human body sensing sensor, a quick-release mechanism, an electric pan / tilt, a tripod, a camera, an optical fiber, a battery assembly and a photoelectric sensor;

[0019] The laser obstacle removal equipment confirmation detection unit is also used for:

[0020] After confirming the components in the laser obstacle removal equipment, functional testing is carried out, and each component is tested separately according to its characteristic requirements.

[0021] Preferably, the obstacle identification and analysis unit comprises:

[0022] Acquisition image optimization module for:

[0023] The camera in the laser obstacle removal equipment is used to shoot the environment of the obstacle removal area, and multiple environmental image data are obtained after the environment shooting is completed;

[0024] Optimizing environmental image data, including image processing, dynamic adjustment and batch normalization of environmental image data;

[0025] Among them, the image processing is to reduce noise, enhance, segment and compress the environmental image data in sequence;

[0026] After the image processing is completed, dynamic adjustment is performed, and the dynamic adjustment is to dynamically train the environmental image data after the image processing by using exponential decay, and the learning rate of the environmental image data is gradually reduced during the training process;

[0027] After the dynamic adjustment is completed, batch normalization is performed. Batch normalization is to use the neural network model to normalize the input of each layer of the dynamically adjusted environmental image data;

[0028] Finally, the optimized environment image data is obtained and marked as optimized image data.

[0029] Preferably, the obstacle identification and analysis unit further includes:

[0030] Optimized image analysis modules for:

[0031] The optimized image data is first subjected to edge detection;

[0032] The optimized image data is meshed, where the meshing is to divide the optimized image data into a plurality of mesh regions of uniform size;

[0033] The grayscale value of the pixel in each grid area is extracted, and the edge information of each grid area is obtained according to the relationship between the grayscale value of each pixel and the grayscale change within the grid area;

[0034] Each edge information is refined, denoised and connected to obtain the edge contour in the grid area;

[0035] Extract features of edge contours in the grid area, and the extracted features include shape features, texture characteristics and color features of the image;

[0036] Using a computer learning algorithm to train and calculate obstacle features and obstacle models in the optimized image data, wherein the obstacle model is retrieved from a database;

[0037] Performing a similarity comparison between the obstacle features in the optimized image data and the obstacle model, and confirming the obstacle position in the optimized image data according to the similarity comparison result;

[0038] Among them, the higher the similarity, the more obvious the obstacle features in the optimized image data;

[0039] After confirming the obstacle position in the optimized image data, the obstacle position is marked as obstacle position data.

[0040] Preferably, the optimized image data is gridded, including:

[0041] Extracting the number of pixel blocks of the optimized image data;

[0042] Obtaining a first division coefficient using the grayscale value of each pixel block of the optimized image data;

[0043] Extracting the number of pixel blocks corresponding to the maximum grayscale value of the optimized image data and the number of pixel blocks corresponding to the minimum grayscale value among the non-zero grayscale values ​​of the optimized image data;

[0044] Obtain a second division coefficient using the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the number of pixel blocks corresponding to the grayscale minimum value in the non-zero grayscale value of the optimized image data and their corresponding grayscale maximum values ​​and grayscale minimum values;

[0045] The upper limit value and the lower limit value of the grid size range are obtained by using the first division coefficient and the second division coefficient; wherein the upper limit value and the lower limit value of the grid size range are obtained by the following formula:

[0046] ;

[0047] Among them, M d represents the lower limit of the grid size range, and the lower limit of the grid size range is rounded down; M u represents the upper limit of the grid size range, and the upper limit of the grid size range is rounded up; M d0 and M u0 Respectively represent the lower limit and upper limit of the preset initial grid size range; R represents the gray value coefficient; E represents the second division coefficient;

[0048] The optimized image data is grid-divided according to the upper limit value and the lower limit value of the grid size range.

[0049] Preferably, obtaining the first division coefficient using the grayscale value of each pixel block of the optimized image data comprises:

[0050] Extracting the grayscale value of each pixel block of the optimized image data;

[0051] Extracting the median grayscale value corresponding to the pixel blocks included in the optimized image data;

[0052] The reference grayscale value corresponding to the optimized image data is obtained by using the median grayscale value corresponding to the pixel blocks included in the optimized image data, wherein the reference grayscale value corresponding to the optimized image data is obtained by the following formula:

[0053] ;

[0054] Where r represents the reference grayscale value; n represents the total number of pixel blocks contained in the optimized image data; H i Represents the grayscale value corresponding to the i-th pixel block; H c Indicates the median grayscale value corresponding to the pixel block contained in the optimized image data;

[0055] Extracting the reference grayscale value;

[0056] The gray value coefficient is obtained by combining the gray value of each pixel block of the optimized image data with the reference gray value; wherein the gray value coefficient is obtained by the following formula:

[0057] ;

[0058] Among them, R represents the gray value coefficient; H min and H max They respectively represent the minimum grayscale value and the maximum grayscale value of the pixel block contained in the optimized image data; r represents the reference grayscale value, and the reference grayscale value is obtained by the following formula.

[0059] Preferably, the second division coefficient is obtained by using the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the number of pixel blocks corresponding to the grayscale minimum value in the non-zero grayscale value of the optimized image data and their corresponding grayscale maximum value and grayscale minimum value, including:

[0060] Extracting the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the value corresponding to the grayscale maximum value;

[0061] The first gray value parameter is obtained by using the number of pixel blocks corresponding to the gray value maximum value of the optimized image data and the value corresponding to the gray value maximum value, wherein the first gray value parameter is obtained by the following formula:

[0062] ;

[0063] Among them, E 01 represents the first gray value parameter; m represents the number of pixel blocks corresponding to the gray maximum value; k represents the number of pixel blocks corresponding to the gray minimum value among the non-zero gray values ​​of the optimized image data; H max Indicates the maximum grayscale value of the pixel block contained in the optimized image data; H p represents the grayscale average value of the pixel blocks included in the optimized image data; n represents the total number of pixel blocks included in the optimized image data;

[0064] Extract the number of pixel blocks corresponding to the minimum grayscale value among the non-zero grayscale values ​​of the optimized image data and the value corresponding to the minimum grayscale value among the non-zero grayscale values;

[0065] The second gray value parameter is obtained by using the number of pixel blocks corresponding to the gray minimum value in the non-zero gray value of the optimized image data and the value corresponding to the gray minimum value in the non-zero gray value, wherein the second gray value parameter is obtained by the following formula:

[0066] ;

[0067] Among them, E 02 represents the second gray value parameter; k represents the number of pixel blocks corresponding to the minimum gray value among the non-zero gray values ​​of the optimized image data; m represents the number of pixel blocks corresponding to the maximum gray value; H fmin Indicates the value corresponding to the minimum grayscale value among non-zero grayscale values; H p represents the grayscale average value of the pixel blocks included in the optimized image data; n represents the total number of pixel blocks included in the optimized image data;

[0068] The second division coefficient is obtained by using the first gray value parameter and the second gray value parameter; wherein the second division coefficient is obtained by the following formula:

[0069] ;

[0070] Wherein, E represents the second partition coefficient; E 01 Represents the first gray value parameter; E 02 represents the second gray value parameter; k represents the number of pixel blocks corresponding to the minimum gray value among the non-zero gray values ​​of the optimized image data; m represents the number of pixel blocks corresponding to the maximum gray value.

[0071] Preferably, the obstacle aiming and tracking unit comprises:

[0072] Obstacle aiming module for:

[0073] Obtaining the coordinate coefficients in the obstacle position data;

[0074] The auxiliary red dot sight in the laser obstacle removal equipment emits a laser beam to aim and illuminate the obstacle according to the coordinate coefficients;

[0075] At the same time, confirm the changing distance and time of the laser beam each time it is aimed at the obstacle.

[0076] Preferably, the obstacle aiming and tracking unit further includes:

[0077] Obstacle tracking module for:

[0078] When the laser beam is aimed at the obstacle in the obstacle position data, the obstacle is monitored and tracked in real time;

[0079] The particle filter method is used to calculate the characteristics and operation mode of obstacles, and the direction of the laser beam is adjusted in real time according to the calculation results;

[0080] Before the laser beam tracks the obstacle, the particle filter method initializes a group of particles according to the initial position of the obstacle in the obstacle position data, and the initialized particles are the first position of the obstacle in the image data obtained by the laser obstacle removal device;

[0081] According to the confirmation of the changing distance and time of the laser beam each time it is aimed at and irradiated on the obstacle, the particle data of each change is acquired;

[0082] Compare the states of the particles that change each time with those of the initialization particles, and derive several weight values ​​based on the comparison results;

[0083] Each weight value is judged by particles, and particles with low weight values ​​are deleted;

[0084] The obstacle tracking data of the laser beam tracking obstacle is obtained by deleting the particles with low weight values ​​in the weight values.

[0085] Preferably, the laser ranging control clearing unit comprises:

[0086] Obstacle control ranging module, used for:

[0087] When the laser beam in the laser obstacle removal equipment contacts an obstacle, the photoelectric sensor in the laser obstacle removal equipment acquires the laser signal reflected from the obstacle, converts the laser signal into an electrical signal, and transmits the electrical signal to the control center;

[0088] The control center performs signal characteristic analysis on the transmitted electrical signal, wherein the signal characteristic analysis is to analyze the amplitude, frequency and phase of the transmitted electrical signal;

[0089] Calculate the distance between the laser beam and the obstacle based on the propagation speed and propagation time of the laser beam in the air;

[0090] Obstacle control and clearance module for:

[0091] Determine the laser beam clearing parameters according to the distance between the laser beam and the obstacle, the material, shape and size of the obstacle;

[0092] Among them, the cleaning laser parameters include the power, frequency and scanning angle of the laser beam;

[0093] The human sensing sensor confirms whether there is anyone in the obstacle clearance area based on the clearing laser parameters;

[0094] If the human body sensing sensor senses that there is someone in the obstacle clearance area, the laser obstacle removal equipment stops emitting the laser beam;

[0095] If the human body sensing sensor does not sense that there is anyone in the obstacle clearance area, the obstacle clearance laser is emitted to clear the obstacle, and the power, frequency and scanning angle of the obstacle clearance laser are consistent with the laser beam.

[0096] Compared with the prior art, the present invention has the following beneficial effects:

[0097] 1. The present invention provides a laser obstacle removal remote control system for remote obstacle removal of overhead lines. The integration and optimization of steps such as image processing, dynamic adjustment and batch normalization can significantly improve the processing efficiency of image data, so that the laser obstacle removal equipment can respond and identify obstacles more quickly. The computer learning algorithm is used to optimize the obstacle features in the image data and the obstacle model for training and calculation, which can greatly improve the system's adaptability and robustness. By continuously learning new obstacle features, the system can gradually improve its recognition performance.

[0098] 2. The present invention provides a laser obstacle removal remote control system for remote obstacle removal of overhead lines. The use of an auxiliary red dot sight simplifies the aiming process, allowing the operator to complete the aiming operation more quickly and accurately. The characteristics and operation mode of the obstacle are calculated by the particle filtering method, and the direction of the laser beam is adjusted in real time according to the calculation results, thereby ensuring accurate tracking of the obstacle position. By accurately tracking the position and movement mode of the obstacle, the laser obstacle removal equipment can aim and clear the obstacle more accurately.

[0099] 3. The present invention provides a laser obstacle removal remote control system for remote obstacle removal of overhead lines. By measuring the distance between the laser beam and the obstacle in real time, the operator can ensure that the laser obstacle removal equipment is operated within a safe distance to avoid potential harm to the surrounding environment and personnel. By adjusting the laser parameters such as power, frequency and scanning angle according to the characteristics of the obstacle, it can be ensured that the laser beam can accurately clear the obstacle, avoiding unnecessary energy waste and potential impact on the surrounding environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0100] Figure 1 It is a schematic diagram of a module for remote obstacle removal of overhead lines of the present invention;

[0101] Figure 2 The present invention is a flow chart of a method for remotely clearing obstacles of overhead lines. DETAILED DESCRIPTION

[0102] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0103] In order to solve the problem that in the prior art, when laser obstacle removal equipment collects images of the obstacle removal area, the collected images are not further optimized and processed, resulting in inaccurate obstacle acquisition, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions:

[0104] A laser obstacle removal remote control system for remote obstacle removal of overhead lines, comprising:

[0105] Laser obstacle removal equipment confirmation detection unit, used for:

[0106] Perform functional tests on each component in the laser obstacle removal equipment, and determine whether the laser obstacle removal equipment is in normal operation based on the test results;

[0107] Obstacle recognition and analysis unit, used for:

[0108] The laser obstacle removal equipment collects image data of the environment in the obstacle removal area, optimizes and analyzes the collected image data, determines the position of the obstacle in the obstacle removal area according to the analysis results, and marks the position of the obstacle in the obstacle removal area as obstacle position data;

[0109] Obstacle targeting and tracking unit for:

[0110] The laser obstacle removal device aims at the obstacle location data, and then tracks the obstacle location data. At the same time, the obstacle location data is monitored in real time during tracking, and the monitoring data is marked as obstacle tracking data;

[0111] Laser rangefinder controlled clearing unit for:

[0112] The laser obstacle removal equipment transmits obstacle tracking data to the control center for distance measurement. After the distance measurement is completed, the laser clearance parameters are adjusted according to the characteristics and distance of the obstacle. After the laser clearance parameters are adjusted, laser clearance is performed.

[0113] Specifically, by using the laser obstacle removal equipment confirmation detection unit to conduct a comprehensive functional test of the laser obstacle removal equipment regularly or before use, potential problems or faults can be discovered in time, so that repairs or replacements can be made in advance to avoid problems in the actual obstacle removal process. The obstacle recognition and analysis unit uses a computer learning algorithm to optimize the obstacle features in the image data and the obstacle model for training and calculation, which can greatly improve the system's adaptability and robustness. The obstacle aiming and tracking unit accurately tracks the position and movement mode of the obstacle, and the laser obstacle removal equipment can more accurately aim at and clear the obstacle, thereby improving the safety and efficiency of the operation. The laser ranging control and clearing unit adjusts the laser parameters such as power, frequency and scanning angle according to the characteristics of the obstacle, which can ensure that the laser beam can accurately clear the obstacle and avoid unnecessary energy waste and potential impact on the surrounding environment.

[0114] The laser obstacle removal equipment confirmation detection unit is also used for:

[0115] The various components in the laser obstacle removal equipment include auxiliary red dot sights, human body sensing sensors, quick-release mechanisms, electric pan / tilts, tripods, cameras, optical fibers, battery components, and photoelectric sensors;

[0116] After each component in the laser obstacle removal equipment is confirmed, a functional test is performed;

[0117] Among them, each component in the laser obstacle removal equipment is tested separately according to the characteristic requirements of each component.

[0118] Specifically, by conducting a comprehensive functional test on the laser obstacle removal equipment regularly or before use, potential problems or failures can be discovered in time, so that they can be repaired or replaced in advance, avoiding problems in the actual obstacle removal process, and improving the reliability and stability of the equipment. Laser obstacle removal equipment has certain dangers, especially if a failure occurs during use, it may cause damage to the human body or the environment. Through the functional test of the detection unit, it can be ensured that the equipment is running in a safe state, reducing the possibility of accidents.

[0119] Obstacle recognition and analysis unit, including:

[0120] Acquisition image optimization module for:

[0121] The camera in the laser obstacle removal equipment is used to shoot the environment of the obstacle removal area, and multiple environmental image data are obtained after the environment shooting is completed;

[0122] Optimizing environmental image data, including image processing, dynamic adjustment and batch normalization of environmental image data;

[0123] Among them, the image processing is to reduce noise, enhance, segment and compress the environmental image data in sequence;

[0124] After the image processing is completed, dynamic adjustment is performed, and the dynamic adjustment is to dynamically train the environmental image data after the image processing by using exponential decay, and the learning rate of the environmental image data is gradually reduced during the training process;

[0125] After the dynamic adjustment is completed, batch normalization is performed. Batch normalization is to use the neural network model to normalize the input of each layer of the dynamically adjusted environmental image data;

[0126] Finally, the optimized environment image data is obtained and marked as optimized image data.

[0127] Specifically, through noise reduction processing, the noise in the image can be effectively removed and the image clarity can be improved. Image enhancement technology can enhance the contrast, brightness and other attributes of the image, making the image easier to observe and identify. Segmentation processing can separate different areas or objects in the image for subsequent analysis and processing. Compression processing can reduce the storage space and transmission bandwidth of image data and improve processing efficiency. Using exponential decay for dynamic training, the model parameters can be continuously adjusted according to the feedback during the training process to improve the adaptability and accuracy of the model. Gradually reducing the learning rate can prevent the model from overfitting in the later stages of training and further improve the performance of the model. By normalizing the input of each layer, the model can be more easily converged to the optimal solution while reducing the dependence on initialization parameters. The integration and optimization of steps such as image processing, dynamic adjustment and batch normalization can significantly improve the processing efficiency of image data, allowing laser obstacle removal equipment to respond and identify obstacles more quickly.

[0128] Optimized image analysis modules for:

[0129] The optimized image data is first subjected to edge detection;

[0130] The optimized image data is meshed, where the meshing is to divide the optimized image data into a plurality of mesh regions of uniform size;

[0131] The grayscale value of the pixel in each grid area is extracted, and the edge information of each grid area is obtained according to the relationship between the grayscale value of each pixel and the grayscale change within the grid area;

[0132] Each edge information is refined, denoised and connected to obtain the edge contour in the grid area;

[0133] Extract features of edge contours in the grid area, and the extracted features include shape features, texture characteristics and color features of the image;

[0134] Using a computer learning algorithm to train and calculate obstacle features and obstacle models in the optimized image data, wherein the obstacle model is retrieved from a database;

[0135] Performing a similarity comparison between the obstacle features in the optimized image data and the obstacle model, and confirming the obstacle position in the optimized image data according to the similarity comparison result;

[0136] Among them, the higher the similarity, the more obvious the obstacle features in the optimized image data;

[0137] After confirming the obstacle position in the optimized image data, the obstacle position is marked as obstacle position data.

[0138] Specifically, by dividing the optimized image data into grids and extracting edge information based on the grayscale value and its change relationship within each grid, the scheme can capture the edge contours in the image more accurately. This fine edge detection is crucial for subsequent obstacle recognition. The process of refining, denoising and connecting each edge information helps to remove irrelevant noise data and retain important image features. This processing can significantly improve the accuracy of image analysis. The feature extraction process not only focuses on shape features, but also includes texture characteristics and color features, which enables the system to more comprehensively describe and optimize obstacles in image data. This comprehensive feature extraction helps to improve the reliability of obstacle recognition. Using computer learning algorithms to train and calculate obstacle features in optimized image data with obstacle models can greatly improve the system's adaptability and robustness. By continuously learning new obstacle features, the system can gradually improve its recognition performance. By comparing the similarity between obstacle features in optimized image data and obstacle models, the system can accurately confirm the location of obstacles. This similarity-based location confirmation method is simple and effective, and can quickly respond to and mark obstacle location data.

[0139] Specifically, the optimized image data is meshed, including:

[0140] Extracting the number of pixel blocks of the optimized image data;

[0141] Obtaining a first division coefficient using the grayscale value of each pixel block of the optimized image data;

[0142] Extracting the number of pixel blocks corresponding to the maximum grayscale value of the optimized image data and the number of pixel blocks corresponding to the minimum grayscale value among the non-zero grayscale values ​​of the optimized image data;

[0143] Obtain a second division coefficient using the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the number of pixel blocks corresponding to the grayscale minimum value in the non-zero grayscale value of the optimized image data and their corresponding grayscale maximum values ​​and grayscale minimum values;

[0144] The upper limit value and the lower limit value of the grid size range are obtained by using the first division coefficient and the second division coefficient; wherein the upper limit value and the lower limit value of the grid size range are obtained by the following formula:

[0145] ;

[0146] Among them, M d represents the lower limit of the grid size range, and the lower limit of the grid size range is rounded down; M u represents the upper limit of the grid size range, and the upper limit of the grid size range is rounded up; M d0 and M u0 Respectively represent the lower limit and upper limit of the preset initial grid size range; R represents the gray value coefficient; E represents the second division coefficient;

[0147] The optimized image data is grid-divided according to the upper limit value and the lower limit value of the grid size range.

[0148] The technical effect of the above technical solution is: by dynamically calculating the size range of the grid division (i.e., the upper and lower limits), the technology can adjust the grid size according to the specific characteristics of the optimized image data (such as the grayscale distribution of the pixel block, the extreme value of the grayscale value and the corresponding number of pixel blocks). This adaptive adjustment mechanism makes the grid division more in line with the actual characteristics of the image data and improves the flexibility and adaptability of image processing.

[0149] Grid partitioning is a commonly used technique in image processing and analysis. A reasonable grid size can more effectively extract image features and reduce information loss or redundancy. By calculating the first partition coefficient (based on the grayscale value of the pixel block) and the second partition coefficient (based on the extreme value of the grayscale value and the corresponding number of pixel blocks), the scheme can more accurately determine the grid size, thereby optimizing the effect of image analysis, such as more accurately identifying targets, textures or edges in the image.

[0150] In image processing, a grid that is too small may lead to a sharp increase in the amount of calculation, while a grid that is too large may lose important information. Through the preset initial grid size range (Md0 and Mu0) and the grid size range (Md and Mu) dynamically adjusted according to the image data characteristics, this technology can reasonably control the amount of calculation and improve the calculation efficiency while ensuring the analysis effect.

[0151] Since the grid size is dynamically calculated based on the actual characteristics of the image data, this technology has good robustness for different types of image data (such as different resolutions, different grayscale distributions, etc.). Even in the face of complex image scenes, this technology can ensure the effectiveness and accuracy of image analysis through adaptive grid division.

[0152] By providing the upper and lower limits of the grid size range, the technology also supports multi-scale grid division and analysis of images. This helps capture image features at different scales and provides a richer information basis for subsequent image recognition, classification or compression tasks.

[0153] In summary, this technical solution improves the flexibility, analysis effect, computational efficiency and robustness of image processing by dynamically calculating the grid division size range, and supports multi-scale analysis, which is of great significance to the field of image processing and analysis.

[0154] Specifically, obtaining a first division coefficient by using the grayscale value of each pixel block of the optimized image data includes:

[0155] Extracting the grayscale value of each pixel block of the optimized image data;

[0156] Extracting the median grayscale value corresponding to the pixel blocks included in the optimized image data;

[0157] The reference grayscale value corresponding to the optimized image data is obtained by using the median grayscale value corresponding to the pixel blocks included in the optimized image data, wherein the reference grayscale value corresponding to the optimized image data is obtained by the following formula:

[0158] ;

[0159] Where r represents the reference grayscale value; n represents the total number of pixel blocks contained in the optimized image data; H i Represents the grayscale value corresponding to the i-th pixel block; H c Indicates the median grayscale value corresponding to the pixel block contained in the optimized image data;

[0160] Extracting the reference grayscale value;

[0161] The gray value coefficient is obtained by combining the gray value of each pixel block of the optimized image data with the reference gray value; wherein the gray value coefficient is obtained by the following formula:

[0162] ;

[0163] Among them, R represents the gray value coefficient; H min and H max They respectively represent the minimum grayscale value and the maximum grayscale value of the pixel block contained in the optimized image data; r represents the reference grayscale value, and the reference grayscale value is obtained by the following formula.

[0164] The technical effect of the above technical solution is: by extracting the gray value of each pixel block and calculating the gray value median (Hc) of the entire optimized image data, the reference gray value (r) is further obtained. This technology can more comprehensively consider the overall gray distribution of the image. The reference gray value is used as a benchmark to help more accurately measure the relative position of the gray values ​​of each pixel block, so as to more effectively extract image features in subsequent processing.

[0165] The calculation of the gray value coefficient (R) combines the gray value of each pixel block, the minimum gray value (Hmin), the maximum gray value (Hmax) and the reference gray value (r). This comprehensive consideration makes the calculation of the gray value coefficient more accurate and can more accurately reflect the relationship between the gray value of the pixel block and the overall gray distribution. This helps to obtain more accurate results in subsequent image processing tasks, such as image segmentation and feature extraction.

[0166] Although the calculation of the reference grayscale value and grayscale value coefficient increases the processing steps, these calculations are based on the overall characteristics of the image data and only need to be performed once. Therefore, in subsequent meshing or other image processing tasks, these pre-calculated results can be directly used, thereby improving the efficiency of the overall processing.

[0167] By introducing reference grayscale values ​​and grayscale coefficients, the technology can better cope with noise or outliers in image data. The reference grayscale value is used as a global benchmark to reduce the impact of local noise on grayscale coefficient calculation; while the grayscale coefficient calculation method takes into account the extreme values ​​and overall distribution of grayscale values, making the algorithm more robust to small changes in image data.

[0168] Since the gray value coefficient reflects the relationship between the gray value of the pixel block and the overall gray distribution, this technology can not only be used for grid division, but also support deeper image analysis tasks. For example, the pixel blocks can be sorted, grouped or clustered according to the gray value coefficient to further explore the potential information or patterns in the image.

[0169] In summary, this technical solution improves the accuracy of image feature extraction, the calculation accuracy of gray value coefficients, and the efficiency and robustness of image processing by calculating reference gray values ​​and gray value coefficients, and supports multi-level image analysis tasks. This is of great significance in the field of image processing and analysis.

[0170] Specifically, the second division coefficient is obtained by using the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the number of pixel blocks corresponding to the grayscale minimum value in the non-zero grayscale value of the optimized image data and their corresponding grayscale maximum value and grayscale minimum value, including:

[0171] Extracting the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the value corresponding to the grayscale maximum value;

[0172] The first gray value parameter is obtained by using the number of pixel blocks corresponding to the gray value maximum value of the optimized image data and the value corresponding to the gray value maximum value, wherein the first gray value parameter is obtained by the following formula:

[0173] ;

[0174] Among them, E 01 represents the first gray value parameter; m represents the number of pixel blocks corresponding to the gray maximum value; k represents the number of pixel blocks corresponding to the gray minimum value among the non-zero gray values ​​of the optimized image data; H max Indicates the maximum grayscale value of the pixel block contained in the optimized image data; H p represents the grayscale average value of the pixel blocks included in the optimized image data; n represents the total number of pixel blocks included in the optimized image data;

[0175] Extract the number of pixel blocks corresponding to the minimum grayscale value among the non-zero grayscale values ​​of the optimized image data and the value corresponding to the minimum grayscale value among the non-zero grayscale values;

[0176] The second gray value parameter is obtained by using the number of pixel blocks corresponding to the gray minimum value in the non-zero gray value of the optimized image data and the value corresponding to the gray minimum value in the non-zero gray value, wherein the second gray value parameter is obtained by the following formula:

[0177] ;

[0178] Among them, E 02 represents the second gray value parameter; k represents the number of pixel blocks corresponding to the minimum gray value among the non-zero gray values ​​of the optimized image data; m represents the number of pixel blocks corresponding to the maximum gray value; H fmin Indicates the value corresponding to the minimum grayscale value among non-zero grayscale values; H p represents the grayscale average value of the pixel blocks included in the optimized image data; n represents the total number of pixel blocks included in the optimized image data;

[0179] The second division coefficient is obtained by using the first gray value parameter and the second gray value parameter; wherein the second division coefficient is obtained by the following formula:

[0180] ;

[0181] Wherein, E represents the second partition coefficient; E 01 Represents the first gray value parameter; E 02represents the second gray value parameter; k represents the number of pixel blocks corresponding to the minimum gray value among the non-zero gray values ​​of the optimized image data; m represents the number of pixel blocks corresponding to the maximum gray value.

[0182] The technical effect of the above technical solution is: by introducing the number of pixel blocks corresponding to the maximum grayscale value (m), the number of pixel blocks corresponding to the minimum grayscale value in the non-zero grayscale value (k), and their corresponding grayscale values ​​(Hmax, Hfmin), the technology can more comprehensively consider the grayscale distribution characteristics of the image data. The first grayscale value parameter (E01) and the second grayscale value parameter (E02) calculated using these parameters further integrate the extreme conditions and overall average conditions of the grayscale values ​​in the image, making the grid division more consistent with the actual characteristics of the image and improving the rationality of the division.

[0183] The maximum and minimum grayscale values ​​usually represent the bright and dark areas in the image, which often contain important information or details of the image. By focusing on the number of pixel blocks and their grayscale values ​​corresponding to these extreme grayscale values, this technology can enhance the ability to capture image details, which helps to better retain and extract these important information in subsequent image processing tasks.

[0184] The adaptability of meshing refers to the ability of the mesh size to be dynamically adjusted according to the different characteristics of the image data. This technical solution further optimizes the meshing size by calculating the second division coefficient (E) and combining the first gray value parameter with the second gray value parameter. This optimization enables the meshing to better adapt to the characteristics of different image data and improves the flexibility and adaptability of the meshing.

[0185] Although this technical solution introduces multiple parameters and calculation steps, these calculations are based on the overall characteristics of the image data and only need to be performed once. Therefore, in subsequent meshing or other image processing tasks, these pre-calculated results can be directly used to improve the efficiency of the overall processing.

[0186] By comprehensively considering multiple factors such as the maximum grayscale value, the minimum grayscale value, the number of corresponding pixel blocks, and the average grayscale value, this technology can reduce the impact of noise or outliers in image data on grid division. This makes the algorithm more robust when facing complex or changing image data.

[0187] In summary, the technical solution further optimizes the grid division size by introducing the concepts of grayscale maximum value, grayscale minimum value and the number of corresponding pixel blocks, and calculating the first grayscale value parameter and the second grayscale value parameter, and then obtaining the second division coefficient. This improves the rationality of grid division, the ability to capture image details, adaptability and calculation efficiency, and enhances the robustness of the algorithm.

[0188] In order to solve the problem in the prior art that the obstacles collected by the laser obstacle removal equipment are not accurately tracked, resulting in the obstacle being unable to be removed after being offset, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions:

[0189] Obstacle aiming and tracking unit, including:

[0190] Obstacle aiming module for:

[0191] Obtaining the coordinate coefficients in the obstacle position data;

[0192] The auxiliary red dot sight in the laser obstacle removal equipment emits a laser beam to aim and illuminate the obstacle according to the coordinate coefficients;

[0193] At the same time, confirm the changing distance and time of the laser beam each time it is aimed at the obstacle.

[0194] Specifically, by obtaining the coordinate coefficients in the obstacle position data, it can be ensured that the laser beam can be accurately aimed and irradiated onto the obstacle. This high-precision aiming is essential to ensure the accuracy of the obstacle removal operation. The changing distance and time of each aiming and irradiation of the laser beam to the obstacle can be confirmed and recorded. This real-time feedback allows the operator to understand the aiming effect in a timely manner and make adjustments as needed. The use of auxiliary red dot sights simplifies the aiming process, allowing operators to complete the aiming operation more quickly and accurately. This helps to improve the efficiency of the entire obstacle removal work. Accurate aiming reduces the risk of accidental injury to surrounding objects and improves the safety of operations. At the same time, real-time feedback also helps operators to detect potential dangers in a timely manner and take measures.

[0195] Obstacle tracking module for:

[0196] When the laser beam is aimed at the obstacle in the obstacle position data, the obstacle is monitored and tracked in real time;

[0197] The particle filter method is used to calculate the characteristics and operation mode of obstacles, and the direction of the laser beam is adjusted in real time according to the calculation results;

[0198] Before the laser beam tracks the obstacle, the particle filter method initializes a group of particles according to the initial position of the obstacle in the obstacle position data, and the initialized particles are the first position of the obstacle in the image data obtained by the laser obstacle removal device;

[0199] According to the confirmation of the changing distance and time of the laser beam each time it is aimed at and irradiated on the obstacle, the particle data of each change is acquired;

[0200] Compare the states of the particles that change each time with those of the initialization particles, and derive several weight values ​​based on the comparison results;

[0201] Each weight value is judged by particles, and particles with low weight values ​​are deleted;

[0202] The obstacle tracking data of the laser beam tracking obstacle is obtained by deleting the particles with low weight values ​​in the weight values.

[0203] Specifically, the obstacle tracking module can monitor and track the position of obstacles in real time, calculate the characteristics and operation mode of obstacles through the particle filtering method, and adjust the direction of the laser beam in real time according to the calculation results, thereby ensuring the accurate tracking of the obstacle position. The particle filtering algorithm initializes a set of particles to represent the possible position of the obstacle, and then updates it according to the changing distance and time of each aiming and irradiation. This probability-based tracking method makes the system more robust to noise and interference. The particle filtering algorithm can dynamically adjust the weight of particles according to the characteristics and operation mode of the obstacle, and optimize the tracking performance by deleting particles with low weights. This dynamic adaptability enables the system to cope with obstacles of different types and motion modes. Although the particle filtering algorithm involves a large number of particle updates and weight calculations, through reasonable algorithm design and optimization, higher computing efficiency can be achieved to ensure that the system can respond to and track obstacles in real time. By accurately tracking the position and motion mode of obstacles, laser obstacle removal equipment can more accurately aim and remove obstacles, thereby improving the safety and efficiency of operation.

[0204] In order to solve the problem that in the prior art, when confirming the laser removal of obstacles, no further confirmation is made based on the surrounding environment and the actual distance of the obstacles, which reduces the safety of the laser obstacle removal equipment, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions:

[0205] Laser rangefinder controlled clearing unit, including:

[0206] Obstacle control ranging module, used for:

[0207] When the laser beam in the laser obstacle removal equipment contacts an obstacle, the photoelectric sensor in the laser obstacle removal equipment acquires the laser signal reflected from the obstacle, converts the laser signal into an electrical signal, and transmits the electrical signal to the control center;

[0208] The control center performs signal characteristic analysis on the transmitted electrical signal, wherein the signal characteristic analysis is to analyze the amplitude, frequency and phase of the transmitted electrical signal;

[0209] The distance between the laser beam and the obstacle is calculated based on the propagation speed and propagation time of the laser beam in the air.

[0210] Specifically, the laser signal reflected by the obstacle is obtained through the photoelectric sensor and converted into an electrical signal for analysis. Combined with the propagation speed and propagation time of the laser beam in the air, the distance between the laser beam and the obstacle can be accurately calculated. This high-precision measurement capability is crucial to ensure the accuracy and safety of laser obstacle removal equipment. The laser obstacle removal equipment can immediately obtain and transmit the laser signal when it contacts the obstacle, and the control center can analyze the signal characteristics and calculate the distance in real time. This real-time feedback mechanism enables operators to understand the working status of the equipment and the distance of the obstacle in a timely manner, so as to make a quick and accurate response. By measuring the distance between the laser beam and the obstacle in real time, the operator can ensure that the laser obstacle removal equipment is operated within a safe distance to avoid potential harm to the surrounding environment and personnel. Accurate distance measurement helps operators locate obstacles more accurately and remove them, thereby improving work efficiency. At the same time, real-time feedback also helps operators to adjust equipment parameters and operating methods in a timely manner to meet different work requirements.

[0211] Obstacle control and clearance module for:

[0212] Determine the laser beam clearing parameters according to the distance between the laser beam and the obstacle, the material, shape and size of the obstacle;

[0213] Among them, the cleaning laser parameters include the power, frequency and scanning angle of the laser beam;

[0214] The human sensing sensor confirms whether there is anyone in the obstacle clearance area based on the clearing laser parameters;

[0215] If the human body sensing sensor senses that there is someone in the obstacle removal area, the laser obstacle removal equipment stops emitting the laser beam;

[0216] If the human body sensing sensor does not sense that there is anyone in the obstacle clearance area, the obstacle clearance laser is emitted to clear the obstacle, and the power, frequency and scanning angle of the obstacle clearance laser are consistent with the laser beam.

[0217] Specifically, the laser clearing parameters, such as power, frequency and scanning angle, are automatically calculated and determined based on the distance between the laser beam and the obstacle, and the material, shape and size of the obstacle. This intelligent setting ensures the pertinence and effectiveness of the clearing process. The human body sensing sensor detects whether there is anyone in the obstacle clearing area, ensuring that the device can automatically stop emitting the laser beam when someone enters, thereby avoiding possible injury accidents and significantly improving the safety of operation. Adjusting laser parameters, such as power, frequency and scanning angle, according to the characteristics of the obstacle can ensure that the laser beam can accurately clear the obstacle, avoiding unnecessary energy waste and potential impact on the surrounding environment. The laser clearing parameters can be adjusted according to the characteristics of different obstacles. This flexibility enables the device to adapt to a variety of different clearing needs.

[0218] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0219] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A laser obstacle removal remote control system for remote obstacle removal of overhead lines, characterized in that: include: Laser obstacle removal equipment confirmation detection unit, used for: Perform functional tests on each component in the laser obstacle removal equipment, and determine whether the laser obstacle removal equipment is in normal operation based on the test results; Obstacle recognition and analysis unit, used for: The laser obstacle removal equipment collects image data of the environment in the obstacle removal area, optimizes and analyzes the collected image data, determines the position of the obstacle in the obstacle removal area according to the analysis results, and marks the position of the obstacle in the obstacle removal area as obstacle position data; Obstacle targeting and tracking unit for: The laser obstacle removal device aims at the obstacle location data, and then tracks the obstacle location data. At the same time, the obstacle location data is monitored in real time during tracking, and the monitoring data is marked as obstacle tracking data; Laser rangefinder controlled clearing unit for: The laser obstacle removal equipment transmits obstacle tracking data to the control center for distance measurement. After the distance measurement is completed, the laser clearance parameters are adjusted according to the characteristics and distance of the obstacle. After the laser clearance parameters are adjusted, laser clearance is performed; The obstacle recognition and analysis unit further includes: Optimized image analysis modules for: The optimized image data is first subjected to edge detection; The optimized image data is meshed, where the meshing is to divide the optimized image data into a plurality of mesh regions of uniform size; The grayscale value of the pixel in each grid area is extracted, and the edge information of each grid area is obtained according to the relationship between the grayscale value of each pixel and the grayscale change within the grid area; Each edge information is refined, denoised and connected to obtain the edge contour in the grid area; Extract features of edge contours in the grid area, and the extracted features include shape features, texture characteristics and color features of the image; The obstacle features and obstacle models in the optimized image data are trained and calculated using a computer learning algorithm, and the obstacle model is retrieved from the database; Performing a similarity comparison between the obstacle features in the optimized image data and the obstacle model, and confirming the obstacle position in the optimized image data according to the similarity comparison result; The higher the similarity, the more obvious the obstacle features in the optimized image data; After confirming the obstacle position in the optimized image data, the obstacle position is marked as obstacle position data; The optimized image data is meshed, including: Extracting the number of pixel blocks of the optimized image data; Obtaining a first division coefficient using the grayscale value of each pixel block of the optimized image data; Extracting the number of pixel blocks corresponding to the maximum grayscale value of the optimized image data and the number of pixel blocks corresponding to the minimum grayscale value among the non-zero grayscale values ​​of the optimized image data; Obtain a second division coefficient using the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the number of pixel blocks corresponding to the grayscale minimum value in the non-zero grayscale value of the optimized image data and their corresponding grayscale maximum values ​​and grayscale minimum values; The upper limit and lower limit of the grid size range are obtained by using the first division coefficient and the second division coefficient; the upper limit and lower limit of the grid size range are obtained by the following formula: ； M d represents the lower limit of the grid size range, and the lower limit of the grid size range is rounded down; M u represents the upper limit of the grid size range, and the upper limit of the grid size range is rounded up; M d0 and M u0 Respectively represent the lower limit and upper limit of the preset initial grid size range; R represents the gray value coefficient; E represents the second division coefficient; Gridding the optimized image data according to the upper limit value and the lower limit value of the grid size range; Obtaining a first division coefficient using the grayscale value of each pixel block of the optimized image data comprises: Extracting the grayscale value of each pixel block of the optimized image data; Extracting the median grayscale value corresponding to the pixel blocks included in the optimized image data; The reference grayscale value corresponding to the optimized image data is obtained by using the median grayscale value corresponding to the pixel blocks included in the optimized image data. The reference grayscale value corresponding to the optimized image data is obtained by the following formula: ； r represents the reference grayscale value; n represents the total number of pixel blocks contained in the optimized image data; H i Represents the grayscale value corresponding to the i-th pixel block; H c Indicates the median grayscale value corresponding to the pixel block contained in the optimized image data; Extracting the reference grayscale value; The gray value coefficient is obtained by combining the gray value of each pixel block of the optimized image data with the reference gray value; the gray value coefficient is obtained by the following formula: ； R represents the gray value coefficient; H min and H max They represent the minimum grayscale value and the maximum grayscale value of the pixel block contained in the optimized image data respectively; r represents the reference grayscale value; The second division coefficient is obtained by using the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the number of pixel blocks corresponding to the grayscale minimum value in the non-zero grayscale value of the optimized image data and their corresponding grayscale maximum value and grayscale minimum value, including: Extracting the number of pixel blocks corresponding to the grayscale maximum value of the optimized image data and the value corresponding to the grayscale maximum value; The first gray value parameter is obtained by using the number of pixel blocks corresponding to the gray value maximum value of the optimized image data and the value corresponding to the gray value maximum value. The first gray value parameter is obtained by the following formula: ； E 01 represents the first gray value parameter; m represents the number of pixel blocks corresponding to the gray maximum value; k represents the number of pixel blocks corresponding to the gray minimum value among the non-zero gray values ​​of the optimized image data; H max Indicates the maximum grayscale value of the pixel block contained in the optimized image data; H p represents the grayscale average value of the pixel blocks included in the optimized image data; n represents the total number of pixel blocks included in the optimized image data; Extract the number of pixel blocks corresponding to the minimum grayscale value among the non-zero grayscale values ​​of the optimized image data and the value corresponding to the minimum grayscale value among the non-zero grayscale values; The second gray value parameter is obtained by using the number of pixel blocks corresponding to the minimum gray value in the non-zero gray value of the optimized image data and the value corresponding to the minimum gray value in the non-zero gray value. The second gray value parameter is obtained by the following formula: ； E 02 represents the second gray value parameter; k represents the number of pixel blocks corresponding to the minimum gray value among the non-zero gray values ​​of the optimized image data; m represents the number of pixel blocks corresponding to the maximum gray value; H fmin Indicates the value corresponding to the minimum grayscale value among non-zero grayscale values; H p represents the grayscale average value of the pixel blocks included in the optimized image data; n represents the total number of pixel blocks included in the optimized image data; The second division coefficient is obtained by using the first gray value parameter and the second gray value parameter; the second division coefficient is obtained by the following formula: ； E represents the second division coefficient; E 01 Represents the first gray value parameter; E 02 represents the second gray value parameter; k represents the number of pixel blocks corresponding to the minimum gray value among the non-zero gray values ​​of the optimized image data; m represents the number of pixel blocks corresponding to the maximum gray value.

2. A laser obstacle removal remote control system for remote obstacle removal of overhead lines according to claim 1, characterized in that: The various components in the laser obstacle removal equipment include an auxiliary red dot sight, a human body sensing sensor, a quick-release mechanism, an electric pan / tilt, a tripod, a camera, an optical fiber, a battery assembly, and a photoelectric sensor; The laser obstacle removal equipment confirmation detection unit is also used for: After confirming the components in the laser obstacle removal equipment, functional testing is carried out, and each component is tested separately according to its characteristic requirements.

3. A laser obstacle removal remote control system for remote obstacle removal of overhead lines according to claim 2, characterized in that: The obstacle recognition and analysis unit comprises: Acquisition image optimization module for: The camera in the laser obstacle removal equipment is used to shoot the environment of the obstacle removal area, and multiple environmental image data are obtained after the environmental shooting is completed; Optimizing environmental image data, including image processing, dynamic adjustment and batch normalization of environmental image data; Among them, the image processing is to reduce noise, enhance, segment and compress the environmental image data in sequence; After the image processing is completed, dynamic adjustment is performed, and the dynamic adjustment is to dynamically train the environmental image data after image processing using exponential decay, and gradually reduce the learning rate of the environmental image data during the training process; After the dynamic adjustment is completed, batch normalization is performed. Batch normalization is to use the neural network model to normalize the input of each layer of the dynamically adjusted environmental image data; Finally, the optimized environment image data is obtained and marked as optimized image data.

4. A laser obstacle removal remote control system for remote obstacle removal of overhead lines according to claim 3, characterized in that: The obstacle aiming and tracking unit comprises: Obstacle aiming module for: Obtaining coordinate coefficients in obstacle position data; According to the coordinate coefficients in the obstacle position data, the auxiliary red dot sight in the laser obstacle removal device is used to aim and illuminate the obstacle through the laser beam; the auxiliary red dot sight in the laser obstacle removal device emits a laser beam to aim and illuminate the obstacle according to the coordinate coefficients; Confirm the changing distance and time of the laser beam each time it is aimed at the obstacle.

5. A laser obstacle removal remote control system for remote obstacle removal of overhead lines according to claim 4, characterized in that: The obstacle aiming and tracking unit further comprises: Obstacle tracking module for: When the laser beam is aimed at the obstacle in the obstacle position data, the obstacle is monitored and tracked in real time; The particle filter method is used to calculate the characteristics and operation mode of obstacles, and the direction of the laser beam is adjusted in real time according to the calculation results; Before the laser beam tracks the obstacle, the particle filter method initializes a group of particles according to the initial position of the obstacle in the obstacle position data, and the initialized particles are the first position of the obstacle in the image data obtained by the laser obstacle removal device; According to the changing distance and time of the laser beam aiming at the obstacle each time, the particle data of each change is acquired; Compare the states of the particles that change each time with those of the initialization particles, and derive several weight values ​​based on the comparison results; Each weight value is judged by particles, and particles with low weight values ​​are deleted; The obstacle tracking data of the laser beam tracking obstacle is obtained by deleting the particles with low weight values ​​in the weight values.

6. A laser obstacle removal remote control system for remote obstacle removal of overhead lines according to claim 5, characterized in that: The laser ranging control clearing unit comprises: Obstacle control ranging module, used for: When the laser beam in the laser obstacle removal equipment contacts an obstacle, the photoelectric sensor in the laser obstacle removal equipment acquires the laser signal reflected from the obstacle, converts the laser signal into an electrical signal, and transmits the electrical signal to the control center; The control center performs signal characteristic analysis on the transmitted electrical signal, and the signal characteristic analysis is to analyze the amplitude, frequency and phase of the transmitted electrical signal; Calculate the distance between the laser beam and the obstacle based on the propagation speed and propagation time of the laser beam in the air; Obstacle control and clearance module for: Determine the laser beam clearing parameters according to the distance between the laser beam and the obstacle, the material, shape and size of the obstacle; The cleaning laser parameters include the power, frequency and scanning angle of the laser beam; The human sensing sensor confirms whether there is anyone in the obstacle clearance area based on the clearing laser parameters; If the human body sensing sensor senses that there is someone in the obstacle removal area, the laser obstacle removal equipment stops emitting the laser beam; If the human body sensing sensor does not sense anyone in the obstacle clearance area, it will emit an obstacle clearance laser to clear the obstacle. The power, frequency and scanning angle of the obstacle clearance laser are consistent with the laser beam.

Images