431results about How to "Improve scene adaptability" patented technology

The invention discloses an efficient flight control system of a low-altitude manned unmanned aerial vehicle based on tilting composite wings, and relates to the technical field of unmanned aerial vehicle control systems. The low-altitude manned unmanned aerial vehicle efficient flight control system based on the tilting composite wings comprises an initialization and pre-check module, a vertical take-off and landing control module, a transition flight control module, an obstacle avoidance and path planning module and a landing control module. Through integration of vertical take-off and landing, smooth transition to cruise, intelligent obstacle avoidance, path planning and other modules, autonomous and efficient flight in a complex environment is realized, and the application scene of the unmanned aerial vehicle is greatly widened. Meanwhile, the accurate flight control and data post-processing analysis capability not only guarantees the safety and comfort of passengers, but also promotes the continuous optimization of the system performance through the real-time performance evaluation and feedback mechanism. In addition, the aerial refueling / charging technology is supported, the cruising ability is prolonged, and the task execution continuity and flexibility are enhanced.

The invention discloses a multi-sensor fusion unmanned aerial vehicle environment monitoring and early warning system, and relates to the technical field of environment monitoring and early warning, and the system comprises the following components: an unmanned aerial vehicle platform which is used as a carrying basis of sensors and data processing equipment and has excellent flight stability and strong cruising ability; according to the invention, by introducing a task planning mechanism based on autonomous learning, the unmanned aerial vehicle can continuously learn and optimize the flight strategy when executing the monitoring task, the innovation point significantly improves the efficiency of task execution, and the unmanned aerial vehicle can perform real-time monitoring according to the actual environment feedback and historical task data. The unmanned aerial vehicle can automatically adjust the parameters of the flight path, speed and height to avoid adverse factors and select the optimal path, and meanwhile, the unmanned aerial vehicle is endowed with strong scene adaptability by the autonomous learning ability, so that the unmanned aerial vehicle can quickly respond to a complicated and changeable natural environment or a sudden emergency, and smooth completion of a task is ensured.

The invention relates to the technical field of AGVs, in particular to an AGV, a passive forking method and an active forking method.By changing the arrangement mode of a fork arm assembly in an AGV body into the arrangement mode of hanging a fork arm on the upper portion, on the premise that the size of the AGV body is not increased, through the arrangement mode of a hanging fork, the long-size fork arm assembly can be contained in the AGV body; when a tray is carried, the fork arms can make contact with the three cross beams, and the safety and stability during tray carrying are enhanced. The AGV trolley can fork the wood tray by using a passive tray forking method and an active tray forking method so as to adapt to different working environments, and the scene adaptability is improved.

The invention discloses a visual detection optimization method and system based on data set classification. The method comprises the steps of introducing a scene classification label and carrying out standardization processing on scene features in a data set labeling and preprocessing stage, and carrying out scene classification multi-task learning in a model selection and training stage, so that a model dynamically adjusts feature learning in different scenes, and the multi-scene adaptive capacity is improved. And in the model evaluation and verification stage, scene-based hierarchical evaluation is introduced and corresponding optimization is carried out. In the model deployment and reasoning stage, a dynamic scene is introduced to adapt to reasoning, and the reasoning path of the model is dynamically adjusted to adapt to the characteristics of the scene, so that the reasoning accuracy is improved. In the later optimization, scene feature clustering and transfer learning are used, and the multi-scene performance of the model is optimized. According to the method, under the condition that the model is not jointly trained through multiple data sets, the multi-scene adaptive capacity of the visual detection model is improved through data set classification.

The invention discloses a fish body length monitoring method based on computer vision, and relates to the technical field of computer vision. Comprising the steps that data content is acquired, and a YOLOv8 detection model is established; the YOLOv8 detection model is trained; obtaining binocular images of the fish, wherein the binocular images comprise a left image and a right image; calibration is carried out to obtain binocular camera parameters, and binocular correction is carried out on the binocular image of the fish; the trained YOLOv8 detection model identifies the binocular image of the fish, and the head and the tail of the fish are taken as key points; establishing an image coordinate system and a three-dimensional world coordinate system, and correlating a conversion relation between the two; key point detection is carried out on the fish to obtain coordinates of key points in the image coordinate system; re-projecting the key points to a three-dimensional world coordinate system; according to the key points in the three-dimensional world coordinate system, a space straight line is fitted, and the length of the space straight line is the fish body length of the fish. The method can monitor the length of the fish body in real time, is high in accuracy and efficiency, and helps to achieve real-time accurate control of marine ranching operation.

The invention discloses an all-season hot rod device and a permafrost stability maintenance method, and belongs to the technical field of permafrost protection.The all-season hot rod device comprises a hot rod unit and a refrigeration unit, the hot rod unit is sequentially divided into a condensation section, a heat insulation section and an evaporation section from top to bottom, and the condensation section is arranged above the ground surface; the heat insulation section is arranged in a permafrost season activity layer, the evaporation section is arranged in a permafrost layer, and a gaseous heat-carrying agent on the lower portion of the evaporation section is injected into the upper condensation section through an ejector in the heat insulation section; the refrigeration unit is arranged at the top of the condensation section and can absorb heat of the condensation section and release the heat into the atmosphere; the operation of the hot rod unit and the refrigeration unit is controlled according to the permafrost temperature. The heat rod unit and the refrigeration unit are combined, cold can be continuously conveyed to permafrost all year round, and the heat stability of permafrost engineering is maintained; and meanwhile, the device has the advantages of zero energy consumption, efficient heat transfer, convenience in transportation and installation, is reliable in operation, and can achieve the dual purposes of permafrost protection, energy conservation and efficiency improvement.

The invention provides an unmanned aerial vehicle conflict resolution method based on reinforcement learning and a self-attention mechanism, and the method comprises the steps: firstly, modeling a dynamic relation between an unmanned aerial vehicle and an airspace as an interaction process between an intelligent agent and an environment in reinforcement learning; then, designing an action space, an observation space and a reward function of the intelligent agent, and building a conflict resolution training framework based on reinforcement learning; and finally, designing an unmanned aerial vehicle attention model based on a self-attention mechanism, which comprises an encoder for extracting unmanned aerial vehicle positions and conflict characteristics in a dynamic environment and a decoder for generating unmanned aerial vehicle conflict resolution actions by using an autoregression mode. According to the novel unmanned aerial vehicle free flight conflict resolution method, reasonable combined action distribution can be carried out on multiple unmanned aerial vehicles flying in the airspace, conflict resolution is completed, and flight safety is guaranteed.

The present invention discloses a weather-resistant refrigeration system and a permafrost protection method applied to the Qinghai-Tibet Plateau, belonging to the field of permafrost protection technology. The refrigeration system includes a plurality of refrigeration units on the ground and a plurality of permafrost coolers in the permafrost. The refrigeration unit is provided with a plateau environmental performance improvement component, and the permafrost cooler is connected to the refrigeration unit; the protection method is to design a refrigeration system according to permafrost, control the refrigeration unit through a control component to cool the refrigerant and then transport it to the permafrost cooler for heat exchange with the permafrost, so that the temperature of the permafrost is reduced. The plateau environmental performance improvement component can make the refrigeration component adapt to the environment of drastic fluctuations in the plateau climate, and improve the weather resistance of the refrigeration system operation; the refrigeration system using an assembled structure is convenient for transportation and installation, and has an adaptive adjustment function for the plateau climate and the permafrost geological environment, ensuring the balance of refrigeration supply and demand and the permafrost protection effect.

The invention discloses an engineering vehicle scheduling method and device, a processor and a machine readable storage medium, and relates to the technical field of intelligent logistics. The method comprises the steps of extracting scheduling features based on acquired historical operation data, wherein the historical operation data comprises historical transportation data, historical task data and historical working condition data; and the scheduling features are respectively input into the first random forest model and the second random forest model to obtain the prediction task demand and the scheduling strategy, so that the scene adaptability of the prediction task demand and the scheduling strategy is improved. Determining an objective function and a constraint condition of a mixed integer programming model based on the prediction task demand, the scheduling strategy and the resource limitation condition of the engineering vehicle; optimizing a target function of the mixed integer programming model based on constraint conditions; and the accuracy of the mixed integer programming model is improved. And outputting an engineering vehicle scheduling scheme corresponding to the optimized target function based on the mixed integer programming model, thereby improving the effectiveness and robustness of engineering vehicle scheduling.

The invention discloses a weather-resistant refrigerating system applied to the Qinghai-Tibet Plateau and a permafrost protection method, and belongs to the technical field of permafrost protection.The refrigerating system comprises a plurality of refrigerating units on the ground and a plurality of permafrost cold conveying devices in permafrost, and each refrigerating unit is provided with a plateau environment performance improving assembly; the permafrost cold conveying device is connected with the refrigerating unit; according to the protection method, the refrigerating system is designed according to the permafrost, the refrigerating unit is controlled by the control assembly to cool the refrigerant, and then the refrigerant is conveyed to the permafrost cold conveying device for heat exchange with the permafrost, so that the temperature of the permafrost is reduced. Through the plateau environment performance improving assembly, the refrigerating assembly can adapt to the environment with severe fluctuation of plateau climate, and the weather resistance of operation of the refrigerating system is improved; the refrigerating system of an assembly type structure is convenient to transport and install, has a self-adaptive adjusting function aiming at the plateau climate and the permafrost geological environment, and guarantees the refrigerating capacity supply and demand balance and the permafrost protection effect.

The utility model discloses an impedance matcher, a radio frequency assembly and plasma processing equipment. Each impedance matcher comprises an impedance matching network and a control unit, at least one impedance matcher is electrically connected between the radio frequency source and the load end through a radio frequency link, and the impedance matching network is electrically connected with the control unit through a high-speed communication bus, the radio frequency source is electrically connected with the control unit through a high-speed communication bus, and the load end is electrically connected with the control unit through a high-speed communication bus. The impedance matching network, the control unit, the radio frequency source and the load end are provided with industrial Ethernet control automation technology ports and / or gigabit transmit-receive interfaces, so that the high-speed communication bus carries out high-speed signal transmission, and the delay is obviously reduced.

The invention relates to an image-text matching analysis method, system and device based on a large model and multiple modes. The method comprises the following steps: 1) image-text analysis: analyzing an image into an image and a text; 2) carrying out fine-tuning multi-mode large model processing on the image, and carrying out word segmentation processing on the characters; 3) constructing a bisection weight; 4) performing Few-shot cue word construction on the processed characters and the constructed equally-divided weight; 5) performing training / fine tuning on the characters and the pictures in the step 4 and the step 2 to serve as a large oracle model; according to the multi-mode image-text matching method, cooperative work of the multi-mode large model and the large language model is introduced, the advantages of the multi-mode large model and the large language model in semantic comprehension, context analysis and dynamic adjustment capacity are comprehensively utilized, a comprehensive image-text matching framework is established, and the accuracy, flexibility and interpretability of matching can be improved.

The invention discloses an intelligent monitoring image defogging method and system based on coring attention. Constructing a pairing data set of the haze image and the corresponding clear image by improving an atmospheric light scattering model; a coring attention defogging network structure is designed, the coring attention defogging network structure comprises a shallow feature extraction layer and a local-global coring attention module, and defogging image restoration is carried out by using convolution operation; training the coring attention defogging network by using three loss functions, namely a mean square error loss function, an edge loss function and a Gaussian kernel perception comparison loss function; and a user inputs a to-be-processed haze image into the trained coring attention defogging network to obtain a restored clear image. According to the invention, haze in the intelligent monitoring scene can be removed, the scene adaptability is high, and the defogging effect is obvious. Random feature mapping in a Gaussian kernel Fourier domain is adopted in the constructed nucleation attention module, and the nucleation attention module is helped to focus on a haze area with higher discrimination in a feature map.

The invention relates to the field of computer servers, and discloses a modularized computing power and storage integrated server which comprises a computing module, a storage module, a bus module, a control module and an optimization module. The calculation module and the storage module support modular hot plug through a standardized interface, and can dynamically adjust configuration according to task requirements; the bus module is used for realizing high-speed data transmission between the calculation module and the storage module and supporting dynamic allocation of bandwidth priority; the control module generates a resource extension suggestion and a bandwidth allocation strategy based on fuzzy reasoning; and the optimization module performs offline optimization on the fuzzy rule base and membership function parameters of the control module through a genetic algorithm so as to improve the overall performance of system regulation and control. According to the method, elastic expansion and dynamic management of computing and storage resources are achieved, the resource utilization rate, the bandwidth allocation efficiency and the system operation stability are improved, and the method is suitable for high-performance computing and large-scale storage tasks with complex loads and multi-scene requirements.

The present invention provides a collaborative electromagnetic signal intelligent identification method and system, which belongs to the technical field of electromagnetic signal analysis and processing. The collaborative sensing node and the collaborative identification controller cooperate with each other. The collaborative sensing node is responsible for air interface electromagnetic signal reception, preprocessing, time-frequency analysis and potential electromagnetic target identification and uses machine learning classification methods to identify the current scene; the collaborative identification controller is responsible for collecting the electromagnetic target list reported by the subordinate collaborative sensing nodes and comprehensively judging and weighting the status, time-frequency position consistency, and graphic morphology consistency of each collaborative sensing node, thereby improving the accuracy of the electromagnetic identification results and the scene adaptation ability.

The embodiment of the invention relates to a vehicle thermal management control method and device, electronic equipment and a storage medium. The method comprises the steps that at least one current working mode of a target vehicle is determined; determining a target working mode meeting a preset thermal management high-power working starting condition from the at least one working mode; determining a thermal management power limit value corresponding to the target working mode; a signal is generated that permits operation of a high-power drive device included in the thermal management system, and the operating power of the thermal management system is limited not to exceed a thermal management power limit. According to the embodiment of the invention, the control strategy of the thermal management system is combined with the working mode of the vehicle, so that the thermal management system can carry out adaptive control according to different working modes, the problems of insufficient thermal management power, insufficient vehicle power and the like can be avoided, and the energy utilization efficiency and scene adaptability of thermal management operation are improved.

The invention discloses a solar refrigeration device and a permafrost protection method, and belongs to the technical field of permafrost protection.The refrigeration device comprises a solar driving unit, an absorption type refrigeration unit and a cooling capacity transmission unit, and the solar driving unit provides a heat source for a bubble generator of the absorption type refrigeration unit and provides power for circulating flowing of a refrigerant; and a condensation section of the cold energy transmission unit is arranged in the cold storage assembly to absorb cold energy, and the cold energy is transmitted to a permafrost degeneration disease work site where the traffic infrastructure is located through a refrigeration section to protect permafrost. The green refrigeration technology is introduced into the field of traffic infrastructure from the field of building environment, the refrigeration device and the condensation section and the refrigeration section of the cooling capacity transmission unit are separately arranged according to application scenes, the function of protecting permafrost is achieved, and the scene adaptability is high; solar energy is used as a driving source, external power supply is not needed, self-driven operation is achieved, the intelligent degree is high, and green and energy-saving effects are achieved; no moving part exists, and the service life is long; installation is convenient, operation is automatic, and reliability is high.

The invention discloses a method for implementing intelligent management and control of a special vehicle operation process based on a virtual space. Only a laser radar and an optical camera which are deployed in a special vehicle operation site and have overlapped detection view fields are used as external sensing equipment to synchronously collect operation environment data; based on the working environment data, constructing a three-dimensional virtual space containing the semantic information of the electrified equipment, and setting a virtual electronic fence to form a security defense area; based on laser radar real-time point cloud data, point cloud clusters of different rigid parts of the special vehicle are identified through segmentation and clustering, and a theoretical three-dimensional contour is generated; extracting a two-dimensional edge of the vehicle part through semantic segmentation based on the optical camera image data, and generating a three-dimensional semantic contour; fusing the theoretical three-dimensional contour with the three-dimensional semantic contour to obtain a high-precision dynamic three-dimensional contour of the moving part of the special vehicle; and calculating the shortest distance between the high-precision dynamic three-dimensional contour and the boundary of the virtual electronic fence, and taking the shortest distance as a reference for executing graded early warning or operating intervention on a special vehicle.

The invention relates to the technical field of distributed optical fiber sensing, in particular to an SVM-based optical cable knocking positioning method and system, and the method comprises the following steps: knocking a manhole cover or the ground above a target optical cable through a hammer; collecting and caching vibration signal data of the optical cable; extracting the vibration signal data to obtain knocking waveform feature information; and inputting the extracted knocking waveform feature information into the SVM model to obtain a judgment result. According to the method, six types of features including an average value, a standard deviation, a variance, a maximum value, a median and a variable coefficient in a to-be-identified signal are extracted and then sent to an SVM model for identification. Non-manual knocking interference is eliminated, optical cable positioning is efficiently and rapidly achieved, and the false alarm rate of the system is reduced. Moreover, the vibration threshold value does not need to be set manually, the scene adaptability is enhanced, and the intelligence of the optical cable routing detector system is further improved. Therefore, the problem that an existing cable knocking positioning method is low in recognition precision is solved.

The invention relates to automatic hoisting equipment for disassembling and assembling power equipment. The automatic hoisting equipment comprises carrying trolleys, and the carrying trolleys are located on the left side and the right side of the hoisting equipment; the control platform is located on the carrying trolley. The vertical lifting mechanical arm assembly is installed on the carrying trolley and used for vertical lifting when equipment is used. The electric hoist is used for hoisting equipment and is controlled by a remote control device; the first horizontal rails are installed at the upper end of the vertical lifting mechanical arm assembly and connected with the second horizontal rails on the same side, and the second horizontal rails on the left side and the right side are connected with each other to form a walking rail used as an electric hoist. The connecting assembly is used for connecting the second horizontal rails on the left and right sides; the position coordinate part is used for position calibration when the second horizontal rails on the left side and the right side are in butt joint; the remote control device is used for moving control of the wireless carrying trolley and lifting control of the vertical lifting mechanical arm. The mechanical hoisting problem of places with different heights and different widths is solved, and good universality and scene adaptability are achieved.

The invention discloses a lithium battery system with a remote upgrading function, belongs to the technical field of lithium battery management, and aims to solve the problems that an existing system is high in field upgrading cost, easy to fail in upgrading, split in operation and maintenance functions, lack of outdoor theft prevention and the like. The system comprises a lithium battery body, a BMS (Battery Management System), an embedded Web UI (User Interface) and a multi-protocol communication module, the BMS is integrated with an Ethernet, a CAN (Controller Area Network) and an RS485 (Recommended Standard 485) communication unit, an upgrade failure three-level protection mechanism is set, and host-slave parallel operation collaborative upgrade and configuration persistence are supported; the embedded Web UI integrates refined log management, three-level password authority control and anti-theft closed-loop control functions; the multi-protocol communication module is compatible with ModbusTCP and SNMP protocols, and has interface backup and fault redundancy design. According to the system, remote upgrading, integrated operation and maintenance and safety protection of the lithium battery are achieved, the upgrading success rate reaches 99.8% or above, the operation and maintenance cost is reduced by 90%, the fault processing efficiency is improved by 85%, and the system can flexibly adapt to scenes such as a communication base station and a data center and has high practicability and market competitiveness.

The invention relates to the technical field of sensor detection, and discloses a deep learning-based nitrogen-oxygen sensor aging test method, which comprises the following steps: extracting universal aging trend data from preprocessed historical large sample data; training the self-attention mechanism model by using the universal aging trend data to obtain a universal basic model; taking the trained general basic model as a transfer learning model, and performing model fine adjustment on the transfer learning model; adding a new full-connection layer, connecting the output of the fine-tuned transfer learning model to the new full-connection layer, activating the output of the new full-connection layer, and training the fine-tuned transfer learning model by using special aging trend data by taking minimization of a predefined loss function as a target to obtain a special detection model; performing an aging test on the to-be-detected nitrogen-oxygen sensor by using the special detection model; according to the invention, the accuracy and robustness of aging detection of the nitrogen-oxygen sensor can be improved.

The invention relates to a roadside pedestrian protection method, device and equipment, a medium and a program product. The roadside pedestrian protection method comprises the following steps: acquiring a pedestrian walking speed when a pedestrian walks along a pedestrian crossing and a vehicle driving speed corresponding to a target vehicle on a target lane associated with the pedestrian crossing; according to the vehicle running speed, determining vehicle running time corresponding to the target vehicle running to a first position of the pedestrian crossing; the first position is the position of a pedestrian crossing stop line of the pedestrian crossing in the passing direction of the target lane; according to the pedestrian walking speed, determining pedestrian walking time when the pedestrian walks to a second position of the pedestrian crosswalk; the second position is a preset position corresponding to the passing direction of the target lane in the pedestrian passing direction; according to the size relation between the vehicle driving time and the pedestrian walking time, generating indication information; the indication information is used for indicating whether the pedestrian safely passes through the pedestrian crosswalk. By adopting the method, the safety of pedestrian passing is improved.

The invention discloses a parking method, device and equipment based on path planning and a storage medium, and relates to the technical field of automatic driving, and the parking method based on path planning comprises the steps: obtaining parking space information and vehicle information; calculating a tangent intersection point coordinate, a parking point coordinate, a first circle, a parking path middle point coordinate, a second circle and a parking space anti-collision suspension point coordinate according to the parking space information, the vehicle information and preset data; according to the parking point coordinate, the second circle, the parking path middle point coordinate, the tangent intersection point coordinate, the first circle and the parking space anti-collision suspension point coordinate, a parking path is obtained; and parking according to the parking path. According to the method, the reasonable parking path can be quickly planned, and the accuracy and safety of automatic parking are guaranteed, so that the driving experience is improved, and the parking pressure of a driver is relieved.

The invention discloses a dangerous area intrusion rapid detection method based on multi-scale feature fusion, and belongs to the technical field of computer vision, and the method comprises the following steps: firstly, collecting a video and an image in which a normal behavior and a violation behavior coexist; preprocessing, screening and marking the collected video images, and dividing the collected video images into a training set, a verification set and a test set; inputting the training set into the multi-scale feature fusion convolutional network for training, verifying the validity of the algorithm in the iteration process by the verification set until the network converges, and obtaining a weight file; and then detecting a dangerous area entering behavior on a test set and in a real-time scene by using the weight file, selecting a detection result, and transmitting and displaying the result in a front-end interface in real time. According to the method, the problems of low scene adaptability and low efficiency of identifying the dangerous area invaded by the worker in the real scene are solved, and the safety and reliability of the worker in the production process are improved.

The invention relates to an electric grounding wire clamp device with an intelligent monitoring function and an operation monitoring method, compared with a traditional manual rotating and twisting grounding wire clamp, an electric device in the electric grounding wire clamp device can quickly complete a clamping task of a grounding wire clamp head through motor driving and force arm design, traditional manual operation needs several minutes or even longer time, and the operation efficiency is greatly improved. However, the device can be easily completed within several seconds, so that the working efficiency is improved; in the actual operation process, an operator can remotely open or close the grounding wire clamp, and the scene adaptability of the electric wire clamp device is further improved. Through the operation monitoring method, whether the operation of the grounding wire clamp accords with the operation specification can be automatically judged in real time, and if not, an alarm signal is sent to prompt an operator to process as soon as possible.

The invention relates to a lamp bead position identification method and device, equipment and a product, and the method comprises the steps: constructing a lamp chain data structure according to the total number of lamp beads to represent the serial position of each lamp bead, and distributing each lamp bead to at least two lamp groups according to the serial position; a special color matching set of each lamp group is determined, and a unique color matching sequence of each lamp bead in the lamp group where the lamp bead is located is coded according to each color in the special color matching set; according to the color matching sequence corresponding to each lamp bead, a synchronous color developing instruction is constructed and sent to a lamp chain, and the lamp beads are driven to synchronously switch light colors; acquiring image frames corresponding to synchronous switching of light colors each time by using a camera unit to form an image sequence, and identifying a color developing sequence of the lamp beads according to the image sequence; and according to the correspondence between the color developing sequence and the color matching sequence, determining the spatial position of the lamp bead in the graphic space constructed by the image plane of the corresponding image sequence. According to the invention, the accuracy and efficiency of identifying the lamp beads can be improved.

The invention discloses a full-hydraulic crawler-type orchard plant protection robot and a control method thereof. The robot comprises a crawler chassis system, a plant protection device, a power system, a hydraulic transmission system and a control system, and the crawler chassis system is used for walking; the plant protection device is used for protecting plants; the power system is used for providing power for the crawler chassis system and the plant protection device; the hydraulic transmission system is used for transmitting power of the power system to the crawler chassis system and the plant protection device; and the control system is used for monitoring the environment, controlling the hydraulic transmission system, driving the crawler chassis system to walk according to a planned path and driving the plant protection device to perform plant protection according to a preset target. By using the robot for operation, the labor intensity of agricultural operation personnel is greatly reduced, effective separation of people and pesticide is realized, and the body health of the operation personnel is guaranteed.

The invention relates to the technical field of three-dimensional point cloud data processing, and provides a subsidence-water-area-oriented point cloud stripe noise LTCF elimination method and system, and the method comprises the steps: obtaining point cloud data, preprocessing, filtering, and optimizing point cloud output. The filtering processing comprises the following steps: local covariance feature extraction: aiming at each target point in the preliminarily optimized point cloud obtained by preprocessing, constructing a k-nearest neighbor domain, and calculating local covariance contribution values of each target point in X, Y and Z spatial directions; compared with the prior art, the method has the following beneficial effects that the industrial pain point of a subsidence water area scene is solved, a three-direction covariance joint judgment mechanism is innovatively introduced, exclusive recognition logic is designed for specific low-elevation strip noise only containing XY-direction covariance in the subsidence water area, the lamellar strip noise which cannot be recognized completely by a traditional algorithm can be accurately captured, and the recognition accuracy of the subsidence water area scene is improved. And the problem of eliminating the point cloud core noise of the subsidence water area is solved fundamentally.