42results about How to "Precise filter" patented technology

The invention discloses a road sign defect detection method and system based on Transform and a medium, and the method comprises the steps: firstly, obtaining an image, carrying out the preprocessing of the image, carrying out the road sign region segmentation of the image through employing a Mask2Former algorithm, and generating a Mask contour of a road sign; based on the minimum enclosing rectangular frame of the road identification area, segmenting the original image to obtain a plurality of sub-images containing the identification area; then, a YOLO-WORLD multi-mode model is used for conducting defect detection on the sub-graph rectangular frame of the road mark, the defects of the road mark are detected on the sub-graph rectangular frame, and a defect positioning frame is generated; and finally, in combination with the Mask contour, filtering the pixel positions of the defects, and filtering out the defects outside the road marking area. According to the method, two excellent algorithm models of Mask2Former and YOLO-WORLD are creatively combined, accurate road marking area segmentation in a complex environment is realized, the detection precision and robustness are improved, the false alarm rate is remarkably reduced, and the detection reliability is improved.

The invention discloses a method for extracting micro-plastics in soil of a refuse landfill. The method comprises the following steps: pretreating a refuse landfill soil sample to obtain sample soil; the sample soil and an extracting agent are fully mixed and oscillated and then stand for layering, first supernate is obtained, the extracting agent is added at least twice, the types of the extracting agent added each time are different, and the extracting agent added each time is one of saturated sodium chloride, saturated sodium bromide and saturated sodium iodide; the first supernate is extracted and fully mixed with a digesting agent, the mixture is shaken and then stands for layering again, second supernate is obtained, and the digesting agent is at least one of 30% hydrogen peroxide, a Fenton reagent, pepsin and trypsin; and filtering the second supernatant to obtain the micro-plastic. Therefore, the extraction method provided by the embodiment of the invention has multiple technical effects of high efficiency, deep digestion, accurate filtration, wide applicability, environmental friendliness and the like.

The invention relates to a THI millimeter wave cloud radar clear sky echo filtering method, device and equipment, and the method comprises the steps: carrying out the marking of clear sky echoes or meteorological echoes through employing the dual-mode characteristics of space-time physical information combination for sample data of each channel in a sample data set; marking the sample data of each channel, constructing a corresponding fuzzy logic membership function according to the marked sample data of each channel, obtaining membership degrees of clear sky echoes and meteorological echoes from the base data of different channels in the original THI data obtained through actual measurement by using the corresponding fuzzy logic membership functions, and inputting the membership degrees of the channels into a trained classification identification model, and identifying clear sky echoes and meteorological echoes in the original THI data to realize filtering of the clear sky echoes. By adopting the method, clear sky echoes can be accurately filtered.

The invention relates to the technical field of Internet of Vehicles, and discloses a vehicle end data processing method, a monitoring system, a domain controller, a medium and a product. The method comprises the following steps: acquiring a data packet monitoring program corresponding to a vehicle end network and an original socket; loading the data packet monitoring program to a kernel mode space, and binding the data packet monitoring program with the original socket to obtain a target socket; and monitoring each network data packet passing through the target socket, filtering out the target data packet from each network data packet by using the data packet monitoring program, and reading data in the target data packet to the user mode space. By implementing the technical scheme of the invention, memory resources are saved, the performance overhead is saved, the flexibility of data packet filtering is improved, the running stability and running safety of the vehicle-side operating system are improved, and the running effect of the vehicle-side operating system is ensured.

The invention provides a building electrical safety detection method and system, and relates to the technical field of electrical detection, and the method comprises the steps of first collection, second collection, first processing, second processing, third processing, model construction, model training, fault recognition and the like. The compressed sensing algorithm and the morphological filter are used, noise and interference signals in the data are effectively removed, the accuracy of fault detection of the electrical equipment in a complex electromagnetic environment in a high-rise building is improved, faults of the electrical equipment can be found in time, corresponding maintenance measures can be taken, and the fault detection efficiency is improved. And safe and stable operation of an electrical system is ensured.

The utility model relates to the technical field of grouting pipe, concretely to a foundation pit plugging grouting pipe, including the pipe body, the surface fixed mounting of pipe body has the connecting pipe and the pipe sharp head, the surface of pipe body is established grouting hole, through setting up the sliding cover screw rod, the sliding cover, the pressing plate screw rod, the filter plate, the pressing plate, the pressure -bearing plate, utilize the synergies of these structures, the sliding cover screw rod drives the sliding cover to slide, the sliding cover is closely combined with the pipe body when closing, prevents the slurry overflow to block the external impurity infiltration, when opening, it is convenient for the filter plate to take out quickly and clean or replace, the filter plate is closely combined with the pressure -bearing plate by the pressing plate screw rod, the filter plate provides stable support with the help of pressure -bearing plate, avoids the filter plate to shift under the slurry impact, also can let the filter plate accurate filtration particle impurity in the slurry, prevents the pipeline blockage, guarantees the grouting pressure stability and grouting passage smooth, has simplified the equipment maintenance process, has promoted the efficiency and reliability of foundation pit plugging grouting construction.

The invention provides a digital geological logging system, which belongs to the technical field of geological logging, and comprises an early-stage geological management system, a data server system and a geological construction real-time logging system, and the early-stage geological management system and the geological construction real-time logging system are both connected with the data server system. The early-stage geological management system is used for inputting early-stage geological engineering information and carrying out early-stage management and planning on geological engineering, the geological construction real-time logging system is used for collecting, recognizing and logging geological construction site data, and the geological construction real-time logging system has offline recognition and logging caching functions. The data server system is used for storing and verifying geological information. Early-stage data of geological engineering are managed in real time and bound with later-stage construction engineering, data support is provided for later-stage construction, real-time datamation of the geological engineering is achieved, management is convenient, working efficiency is improved, and data errors are avoided.

The application relates to the technical field of network security, in particular to a Java Restful interface security test method based on static analysis, which comprises the following steps: generating a Jimple intermediate representation of source code of a target Java application; performing data flow analysis based on the Jimple intermediate representation to generate a complete data flow graph; identifying a pollution source and tracking a pollution propagation path on the basis of the data flow graph, judging whether a pollution object reaches a convergence point, and determining a vulnerable interface; extracting an interface routing signature in the pollution analysis process to generate routing signature information; grouping interfaces based on parent routing of the interfaces, performing interface security test in combination with the routing signature information, introducing a reinforcement learning method to dynamically optimize a test strategy, performing parameter variation and attack load injection of a test case on the vulnerable interface; and setting a security monitoring strategy in a dynamic plug-in manner, and completing interface security risk judgment according to recorded logs and the test case.

The invention discloses a real-time noise filtering method based on an event camera. The objective of the invention is to improve the real-time performance and precision of data processing. According to the method, original event data are collected through an event camera, an event stream is generated, a fixed window filter is used for carrying out noise filtering on the event stream, and a Manhattan distance algorithm is used for judging and filtering noise events. The processed event stream data is displayed through a real-time visual interface, and it is ensured that the event stream data are synchronized with the output of an event camera. The method optimizes the algorithm performance of an embedded system, has an efficient noise filtering effect, and is suitable for dynamic monitoring and real-time processing in an intelligent industrial environment.

The present application relates to the technical field of smart city, especially relates to a smart city bus stop board dynamic information accurate pushing method and system, wherein a query semantic representation structure is constructed by collecting query requests and scene state information, a joint understanding representation is obtained by modeling the scene state information multi-level semantic association, scene constraint multi-hop reasoning is executed in a bus knowledge association network, a candidate answer sequence is generated by filtering, an optimal answer is selected by calculating a scene adaptation score, pushing content is generated, interactive feedback is collected, feature fusion weights are adaptively adjusted to form a closed loop optimization, and accurate dynamic pushing of bus information and user experience improvement are realized.

The utility model discloses a stirring reaction unit is used to lithium phosphate recovery, including the reaction tank, still including L type board, the outside installation of reaction tank has the fixed ring, the outside fixed of fixed ring has a plurality of and the branch leg of circumferential distribution, the top of reaction tank installs the top cover, is provided with the stirring mechanism on the top cover, the top of top cover installs the square tube, and the top of square tube is connected with the detection pipeline, and the top of detection pipeline installs the connecting flange, and the inside of square tube is provided with the dust filter mechanism, and the right side of square tube is connected with the blast pipe, and the inside fixed of blast pipe has the support, and the left side fixed of support has the gas cylinder, and the output of gas cylinder is connected with L type board, the utility model discloses through the gas cylinder drive L type board, when needing to clean the filter screen, the square tube top end is closed fast, forms a closed space, fills in high pressure nitrogen through the blast pipe and carries out the reverse flushing, can effectively remove the dust and impurity that filter screen surface accumulated, and the gas permeability of filter screen is improved significantly.

The invention relates to a deep learning technology, in particular to a two-stage noise label image classification method and system based on a dual-network cooperative training mode, and the method comprises the steps: dividing a data set into multiple batches of input samples, creating two same networks, carrying out the feature extraction of the input samples, obtaining output p1 and p2, and calculating the total loss of a stage through combining with a sample label; retaining the sample loss of which the total loss value is less than the threshold value in the first stage, and averaging to update network parameters in a back propagation manner; processing records of the input samples as historical information of the samples to be reserved; performing statistics on historical sample information to obtain a discrimination matrix Y and a weight matrix W; calculating the total loss of the two stages by using the matrix Y as an indication signal; the sample loss with the second-stage total loss value smaller than the threshold value is reserved, and the matrix W is used for reweighting processing and then averaging is carried out to update network parameters in a back propagation mode; and an image classification model is obtained until the training is finished to classify the images. The interference of noise labels is resisted by selecting the clean instance training network, and the noise can be accurately filtered.

The present invention relates to a fully automatic river and lake water purification robot, comprising a carrying box, which is divided into a ballast water tank and an equipment compartment; a snorkeling device; a water purification device, which comprises a water pump, a replaceable filter element, a drainage panel, a first five-way joint, and a water purification pipe. The present invention also relates to a fully automatic river and lake water purification method, which comprises the steps of: deploying a water purification robot, and then calibrating the current GPS coordinates and water depth data; heading to a target hovering point; starting water purification work; collecting water depth data, dissolved oxygen data, conductivity data, and turbidity data; if the current hovering point is not the last hovering point in the cruise route, executing again; otherwise heading to the GPS coordinates and water depth data at the time of calibration. The present invention fundamentally solves the problem that the prior art cannot reach the deep waters for purification operations; it is accurate and efficient; it accurately filters various pollutants and brings air to where it is needed; it has first-hand water ecological data; and it does not require a special underwater navigation device.

This utility model discloses a high-temperature resistant smoke sensor, relating to the field of smoke sensor technology. A high-temperature resistant smoke sensor includes: a housing, with a smoke detection component disposed inside the housing; a top cover fixedly connected to the upper end of the housing; and an air intake grille disposed on the surface of the top cover. This invention solves the problem that traditional smoke sensors have a temperature resistance limit of only 70-90℃. When the ambient temperature rises suddenly (e.g., a kitchen oil fire can rise from room temperature to 180℃ within 30 seconds), the housing is prone to softening and deformation due to the high temperature, leading to failure of the internal chamber's seal and even short circuits. Furthermore, in industrial settings where the temperature is consistently between 80-150℃ (e.g., in metallurgical workshops and steam boiler rooms), the aging of internal components in the smoke sensor is accelerated, causing drift in the parameters of electronic components such as capacitors and resistors, a decrease in detection sensitivity of more than 50%, and a lifespan shortened to one-third of that under normal conditions. Frequent replacements not only increase costs but also create safety hazards.

The invention discloses a centrifugal filtration type film forming device which comprises a barrel body, the barrel body comprises a first barrel wall, a second barrel wall and a third barrel wall, the second barrel wall is arranged around the first barrel wall, the third barrel wall is arranged around the second barrel wall, and gaps are formed among the first barrel wall, the second barrel wall and the third barrel wall. A spacing area between the first barrel wall and the second barrel wall is a stock solution area, a spacing area between the second barrel wall and the third barrel wall is a filtrate area, a plurality of filtering holes are formed in the second barrel wall, the stock solution area and the filtrate area are communicated through the filtering holes, a filtering membrane and a filtering plate are installed at the filtering holes, one side of the filtering membrane is the stock solution area, and the other side of the filtering membrane is the filtering plate. A barrel cover matched with the barrel body in size covers the barrel body, and the barrel body is connected with a power driving piece. Efficient filtration is achieved through centrifugal force, the industrial rapid treatment requirement is met, accurate filtration is achieved by replacing the filter membrane, and the problem of membrane forming quality control is solved.

The utility model provides a self-compacting backfill material processing device which comprises a main material storage bin, the main material storage bin is connected with a sorting system through a main material conveying channel, the sorting system comprises a screening device, an air classification device and a magnetic separation device which are sequentially connected, and the magnetic separation device is connected with the main material storage bin. According to the treatment device, through the arrangement of the sorting system, unwanted impurities, particles and the like are gradually separated from the main materials, optimization of the quality of raw materials is facilitated, the quality of backfill materials is improved, and classified storage in the main material storage bin is facilitated.

The utility model provides a stirring device, and relates to the technical field of automobile corrosion prevention. The device comprises an emulsification tank body and a stirring mechanism, a filtering mechanism is arranged at one end of an extraction mechanism, the filtering mechanism comprises a filtering box and supporting blocks, the surfaces of one sides of the supporting blocks are installed on the inner wall of the filtering box, and detachable filtering plates are installed on the surfaces of the supporting blocks. And a second valve penetrating through the inner wall is fixedly mounted on the outer side of the filter box. In the working process of the device, the filtering mechanism is additionally arranged, the filtering box has a space for efficiently filtering a large amount of undissolved materials, and the materials are accurately filtered through two layers of filtering plates installed inside, so that the product quality damage and the influence on the equipment practicability are reduced; the problems that if a large number of undissolved materials exist in the emulsifying tank main body, excessive materials are difficult to effectively filter only by means of a filtering device installed at the bottom, residues are possibly accumulated in a filtering pipe, and then serious blockage is caused are solved.

The present application belongs to the technical field of valve, especially a pressure speed measuring device and full bore bidirectional sealing C type valve, aiming at the intelligent sensor of speed measuring valve being easily interfered by turbulent flow and vortex, flow field being disorderly changed, lacking of filtering, reducing speed measuring precision and affecting the judgment of flow data, the following scheme is proposed, including water inlet pipe and straightening pipeline, the water inlet end of the water inlet pipe is connected with upstream pipeline through bolts, the same precision speed measuring module is installed on the water inlet pipe and the straightening pipeline, and the precision speed measuring module includes triangular support one. The pressure speed measuring device and full bore bidirectional sealing C type valve disclosed by the present application has the effects that the integrated impurity filter screen and the hydraulic drive annular cleaning brush in the water inlet pipe realize online self-cleaning of the filtering surface, the converging flow guide cone and the multi-stage multi-hole straightening plate are arranged at the water outlet end, the eccentric jet flow after the valve is converted into uniform and stable flow field, and the annular multi-point intelligent speed measuring sensor is combined to realize the effect of precise speed measurement.

The invention provides a focusing and leveling sensor and an ambient light noise elimination method, which are applied to the technical field of semiconductor processing and manufacturing, and are characterized in that a light source emits parallel light, and the parallel light penetrates through a slit of a modulation slit plate to form a slit image; the projection lens assembly projects a slit image to the surface of the wafer; the detection lens assembly detects a slit image on the surface of the transmission wafer; detecting a part of slits of the slit plate to transmit the transmitted slit image as signal light; ambient light penetrates through a slit of the detection slit plate; a slit of the detection slit plate is arranged corresponding to the detector; the detectors corresponding to one part of slits of the detection slit plate receive ambient light and signal light, and the detectors corresponding to the other part of slits of the detection slit plate receive the ambient light. The detector without the signal light is used for ambient light detection, the detector with the signal light is used for defocusing amount detection of the corresponding position, the influence of the ambient light can be filtered out, and the system defocusing amount detection precision is effectively improved.

This invention provides a machine learning-based method and device for anomaly detection in DDR wafer testing. It acquires timing eye diagrams and timing waveform data related to total jitter, deterministic jitter, and random jitter synchronously during DDR wafer testing. The method performs clock edge phase segmentation alignment on the differential timing sampling points in the timing waveform data to construct dense sampling data in the phase domain. Combined with DDR jitter physical priors, adaptive sparse coding is performed to remove noise and components related to normal process fluctuations, obtaining defect-sensitive sparse components. Variational mode decomposition is then performed to obtain components related to process drift, normal process fluctuations, and abrupt anomalies. The components related to abrupt anomalies are extracted as target detection data, input into a preset machine learning model, and anomaly pattern association results are generated. Based on these, DDR wafer testing anomaly detection results are output, and corresponding test setting adjustment instructions are generated. This invention can improve the overall efficiency and reliability of DDR wafer testing results.

This utility model relates to solid-liquid separation equipment technology, and discloses a multi-stage pressurized filter cylinder structure for a vinegar residue extrusion dewatering machine. It includes a base plate, a support fixedly connected to the rear top of the base plate, a hydraulic assembly fixedly connected to the inner top wall of the support, a connecting base slidably connected to the center of the top of the base plate, a liquid collection tank fixedly connected to the top of the connecting base, a motor fixedly connected to a groove in the top of the connecting base, and a connecting shaft fixedly installed at the output end of the motor. In this utility model, the motor drives the connecting shaft to rotate, causing the connecting plate and adjusting components to rotate around the fixed filter cylinder. When the adjusting components rotate, they block the filter holes, allowing for rapid adjustment of the filter hole diameters of multiple filter cylinders according to different materials and processing levels. This ensures that different sized particles remain in the corresponding filter cylinders during the extrusion filtration of the vinegar residue, achieving precise filtration and solving the problem of not being able to quickly and accurately adjust the filter hole diameter for the dewatered material.

The invention relates to the technical field of composite nanofiber aerogel, in particular to polyester-PAN / PCL composite nanofiber aerogel and a preparation method thereof. Preparing a PAN (Polyacrylonitrile) nanofiber membrane and a PCL (Polycaprolactone) nanofiber membrane; preparing a suspension by taking the nanofiber membrane and polyester fibers as raw materials; and finally, uniformly stirring the suspension, pouring into a culture dish, sealing, freezing, drying and taking out to obtain the PAN / PCL composite nanofiber aerogel. According to the prepared polyester-PAN / PCL composite nanofiber aerogel, graded pore structures exist in the aerogel, and when the proportion of micron-sized polyester staple fibers to nanofibers is 1: 1, due to the fact that the content of PAN and PCL nanofibers is large, the aerogel is high in filtering efficiency, small in resistance pressure drop, high in quality factor and better in filtering performance; when the proportion of the polyester fibers to the nanofibers is fixed, the air permeability and the filtering efficiency of the aerogel have the trend of increasing along with the increase of the mass fraction of the PCL nanofibers.

The application provides a kind of early fault warning method of bridge machine, for the technical problems that weak fault characteristics are easy to be covered by strong noise under complex operation condition, traditional model generalization ability is weak and early warning threshold is rigid, the application first collects multiple source operation data and carries out denoising reconstruction using variational mode decomposition algorithm, realizes weak fault signal enhancement;Secondly, extract multi-dimensional statistical features and use local preserving projection algorithm to reduce dimension, obtain low-dimensional sensitive feature vector;Then construct adaptive kernel spectrum clustering model, use data-driven scale factor and introduce mahalanobis structure penalty similarity matrix to identify high-precision early weak fault state;Finally, combined with the mechanism of non-linear fatigue damage of equipment and real-time working condition variable updates dynamic warning threshold, and then trigger multi-level response to prevent false alarm.The application greatly improves the purity of feature extraction and cross-condition generalization ability in strong noise environment, realizes low false alarm, zero false alarm and accurate active early warning.

The invention provides a wafer graph defect semantic reasoning method driven by a retrieval enhancement large model. The method comprises the following steps: constructing a multi-source data set fusing a wafer graph sample, a domain document and a process document; designing a wafer graph defect knowledge entity and relation extraction mechanism driven by prompt learning, designing a wafer graph defect visual prototype attribute embedding mechanism, and constructing a bimodal defect knowledge graph with text semantics and visual prototype attributes; introducing a defect prototype matching mechanism to realize visual-semantic alignment of the input wafer graph and the knowledge graph, and constructing a fine screen graph through graph path screening and a semantic connection scoring function; and designing a structured cue word construction and role adaptive generation strategy, and guiding the large model to output a query-oriented wafer graph defect semantic reasoning result. The method has higher visual semantic fusion capability, knowledge screening accuracy and multi-role semantic adaptability, and is suitable for defect root cause tracing and intelligent question and answer system construction tasks in the semiconductor manufacturing process.

The invention relates to the technical field of intelligent decision making, can be applied to business system platforms such as financial science and technology, medical health and the like, and discloses a question answer generation method, device, equipment and medium, and the method comprises the steps: collecting a plurality of question answers generated by questions, collecting the question answers generated by each question into a question and answer sample set, calculating a reward model score and a reference model benchmark score of each question and answer sample in the question and answer sample set one by one, fitting into a confidence coefficient curve, and dividing an answer retention interval; screening answers to the questions according to the answer retention interval, and calculating and updating an answer retention proportion in the question and answer sample set; optimizing the question and answer model according to the answer retention proportion, and generating a plurality of candidate answers of the user question by using the optimized question and answer model; and calculating candidate reward scores, retaining candidate answers corresponding to the candidate reward scores in the answer retaining interval, and performing semantic fusion on the answers to generate a final question answer of the user question. The accuracy of question answer generation can be improved.

This application relates to the field of circuit breaker status sensing technology, specifically to a method for collaborative telemetry sensing of the operating status of a feeder terminal unit (FTU) and a pole-mounted circuit breaker. The method includes: collecting internal temperature data of the FTU and current data of the circuit breaker's opening and closing coils; defining the start time of opening and closing based on the current data and a preset threshold, extracting current feature vectors and analyzing their anomaly degree, and constructing operational wear feature values; fitting the data with temperature as the independent variable and operational wear feature values ​​as the dependent variable to obtain residual temperature drift error and calculate residual temperature drift interference intensity; optimizing the observation noise covariance of the Kalman filter based on the residual temperature drift interference intensity to optimally estimate the circuit breaker's operating status. This application solves the problems of distortion in circuit breaker status measurement caused by hardware temperature drift and the difficulty of balancing response speed and smoothing accuracy in traditional fixed-parameter filtering, thus improving the reliability of early minor fault detection in pole-mounted circuit breakers.

The invention relates to the technical field of air conditioning equipment, in particular to a fresh air purification combined type energy-saving air conditioning all-in-one machine which comprises an outer shell, an air inlet module comprises a mixed air inlet pipeline, a fresh air inlet pipeline and a circulating air inlet pipeline, and the opposite faces of the fresh air inlet pipeline and the circulating air inlet pipeline are each provided with an insertion opening; a partition plate is slidably mounted between the fresh air inlet pipeline and the circulating air inlet pipeline, a rotating shaft is rotatably mounted at one end of the mixed air inlet pipeline, an electric control air cylinder is mounted at the upper end of the partition plate, two resistance rods are mounted on the side face of the partition plate through an L-shaped sliding rail and a sliding block, and the rotating shaft is fixedly sleeved with a resistance wheel. According to the air conditioning equipment, the partition plate is driven to move between the fresh air inlet pipeline and the circulating air inlet pipeline, flexible switching of three air inlet modes of fresh air inlet, circulating air inlet and mixed air inlet is achieved, the proper air inlet mode can be selected according to different use scenes and requirements, and the applicability of the air conditioning equipment is improved.

The invention belongs to the technical field of adsorption material preparation, and particularly relates to a one-step preparation method of a morphology-controllable graded ordered porous ZrO2 adsorption material. According to the preparation method, two kinds of PS microsphere primary pulp emulsion with uniform particle size from micron-scale to nano-scale particle size distribution can be prepared, subsequent treatment is not needed, and a binary PS colloidal crystal template with controllable morphology is directly self-assembled; in the process, the particle sizes of the two kinds of PS microspheres are regulated and controlled by controlling the adding amount, the dripping rate, the adding time and the reaction time of KPS; regulating and controlling the proportion of the PS microspheres with two different particle sizes by regulating the mass ratio of the secondarily added styrene monomer to the PS microspheres in the primary pulp emulsion; and mixing a ZrO2 precursor with the binary PS colloidal crystal, dipping, carrying out suction filtration, and calcining to obtain the morphology-controllable graded ordered porous ZrO2 adsorption material with different aperture matching from micron to nano. The material can more accurately filter pollutants such as suspended particles with different sizes in water, and has longer service life and more stable filtering performance.

The present invention discloses a network large flow elasticity measurement method supporting traffic jitter, the method comprising: firstly parsing the network packet and extracting a quintuple as a flow identifier; then designing a small flow filter based on a scalable Sketch cycle chain, and then proposing a filter expansion and contraction strategy that can adaptively adjust the number of Sketches in the cycle chain according to the dynamic change of the network packet rate, so as to ensure that all packets within a given time period are recorded, so as to continuously and accurately filter the small flow; then combining a multi-segment scalable hash algorithm, a large flow recorder based on a scalable hash table is designed, which can expand or contract segments according to the dynamic change of the number of recorded network large flows, so as to record all large flows as much as possible, while improving the utilization rate of storage space; according to the solution of the present invention, the problem of insufficient query accuracy and the problem of being unable to realize adaptive measurement in large flow measurement when the traffic surges on a high-speed network link is effectively solved, and the accuracy and range of large flow measurement are further improved.

The invention belongs to the technical field of computer vision and deep learning, and provides a structured light point cloud denoising method combining image processing and local statistical filtering, which comprises the following steps: firstly, converting a point cloud into a three-coordinate image, carrying out binaryzation, carrying out connected region detection on the binary image to obtain a logic mask value of a maximum connected region, and carrying out local statistical filtering on the logic mask value; and filtering the three-coordinate image according to the logic mask value, performing local statistical filtering on a new three-coordinate image, and finally obtaining a clean point cloud. The method can effectively process the features of the structured light point cloud noise, is not sensitive to parameters, and can accurately eliminate the noise without special debugging.