128 results about "Model fitting" patented technology

The invention relates to a quadruped robot laser odometer method based on ground point optimization, hardware and application, and the method comprises the steps: obtaining an original point cloud collected by a laser radar, carrying out the preprocessing, dividing a polar coordinate region, jointly extracting ground plane feature points and ground geometric edge feature points based on normal vector and local curvature features, and carrying out the calculation of the ground plane feature points and the ground geometric edge feature points; ground geometric edge feature points are utilized to optimize ground model fitting; optimizing a point set of the ground based on a time sequence reference plane of the ground; the optimized ground points are removed, and non-ground point features are extracted from the remaining point clouds; establishing feature association between the feature points in the current frame and the target point cloud, and screening effective matching pairs; a cascade weight optimization objective function associated with the ground point plane constraint, the non-ground point line feature constraint and the surface feature constraint is constructed, and continuous pose transformation of the quadruped robot is solved; hardware is realized based on the method, and the method is applied to real-time pose estimation and environment mapping of the quadruped robot in outdoor unstructured terrains and complex dynamic environments.

The invention discloses an allergic rhinitis diagnosis and allergen traceability system based on dynamic text guidance, and belongs to the technical field of intelligent medical treatment. According to the method, the problems that in the prior art, patient description is inaccurate, allergens are various in variety and have differences, and allergens are difficult to accurately determine and trace are solved, the initial situational sub-graph is verified by introducing periodic features of the patient, the causal contribution degree of the initial situational sub-graph is evaluated through a Bayesian reasoning algorithm, and a mode association result library is formed; according to the method, the crossing of the allergic rhinitis diagnosis and tracing from the traditional static judgment to the dynamic accurate inference is realized, and a personalized and scientific allergen avoidance scheme is provided for patients; through similarity retrieval, environmental information injection verification and calculation of a mode goodness-of-fit score, under the condition that patient symptoms are in multi-factor mixing, a mixed causal graph is created through a graph fusion technology, and the mode goodness-of-fit score is calculated again, so that the accuracy of obtaining an allergen traceability result in practical application is ensured.

The invention discloses a dense topology model lightweight method, and belongs to the technical field of three-dimensional model processing. The problems that an existing scheme is low in efficiency, large in structural damage, poor in mapping quality and the like are solved. The method comprises the following steps: importing a dense grid model containing high-precision geometry and texture; selecting an automatic (adaptive lattice clustering algorithm) or manual mode to generate a target simplified topology; constructing a simplified model fitting the original model through a multi-direction projection technology; uV expansion is realized through automatic blocking, LSCM algorithm optimization and manual adjustment, and textures are baked in combination with RendertoTexture and SSAA technologies; and after performance optimization, outputting a lightweight model compatible with a mainstream format and an engine. According to the invention, high efficiency, stability and light weight are realized, key features are accurately reserved, visual consistency is guaranteed, and multi-scene application is adapted.

The invention relates to a numerical control machine tool reliability modeling method and system considering a fault trend. The method comprises the following steps: checking the homogeneity of a numerical control machine tool based on the fault trend; constructing an AMSAA model of the multiple samples subjected to truncation at unequal time; evaluating the precision of the reliability model; aiming at the problem of insufficient reliability model precision of a multi-sample condition of a numerical control machine tool, a multi-sample AMSAA model modeling method of the numerical control machine tool considering a fault trend is provided, sample data classification is carried out through trend inspection and homotype inspection, and fault data of same-type machine tools are preprocessed by adopting a fault total time method; adopting a maximum likelihood method to estimate AMSAA model parameters of the same-type machine tool, and using Cramer-Von Mises to test the goodness of fit of the model; evaluating the precision of the reliability model by taking an average absolute percentage error (MAPE) of an instantaneous MTBF point estimation value and an MTBF observation value as an index; compared with a multi-sample AMSAA model established by a direct maximum likelihood method, the method is higher in prediction precision.

A method for assessing a lens capsule stability condition in an eye of a human patient includes simultaneously directing electromagnetic energy in a predetermined spectrum via an energy source onto a pupil of the eye after movement of the eye results in eye saccades occurring therein. The method further includes acquiring images of the eye indicative of the eye saccades using an image capture device and calculating a motion profile of the lens capsule using the images via an ECU. Additionally, the method includes extracting time-normalized lens capsule oscillation trajectories based on the motion profile via the ECU and then model fitting the lens capsule oscillation trajectories via the ECU, thereby assessing the lens capsule instability condition. Also disclosed herein is an automated system for performing an embodiment of the method, the automated system including an energy source, an image capture device, and an ECU.

The present application belongs to the technical field of medical monitoring, and particularly relates to a pipe diameter flow monitoring method and system for ECMO, which comprises the following steps: a video image of an ECMO pipe is acquired, a dynamic contour model containing a statistical shape model and a Kalman filter is constructed to adapt to non-rigid deformation of the pipe and track the non-rigid deformation; macro tracking features such as model fitting residuals and kinematic deviations, and micro features such as textures and optical flows inside the pipe are combined to calculate a multi-dimensional tracking confidence, which is used to adaptively adjust parameters of the Kalman filter, so that optimal compensation for pipe vibration and deformation is realized; furthermore, the system can accurately extract the pipe diameter and flow rate, and calculate real-time instantaneous flow. The accuracy, robustness and intelligent level of ECMO monitoring are improved.

The invention relates to a method for determining an individualized nerve regulation target spot, and the method comprises the steps: obtaining nerve image data, which are repeatedly measured for multiple times, of a subject and corresponding clinical psychological assessment data in a preset time period; calculating a brain activity index of each voxel in a preset brain region; establishing a plurality of candidate mathematical models to describe the relationship between the brain activity indexes and the clinical scores; selecting an optimal model through statistical test of fitting residual errors; generating a parameter weight map representing'symptom-activity 'association strength based on the optimal model; and finally, determining an individualized nerve regulation target according to the numerical distribution of the map. The invention also relates to a corresponding device, a processor and a computer readable storage medium thereof, and solves the problem of inaccurate target positioning caused by neglecting symptom heterogeneity and individual brain function difference in the prior art through a core methodology of ''longitudinal tracking-multi-model fitting-residual optimal selection''. The curative effect of nerve regulation and control treatment is obviously improved.

The invention is suitable for the technical field of data identification, and provides a service life test method and device for a wireless charger, and the method comprises the steps: executing a test process at a preset test temperature, and collecting observation data corresponding to a plurality of sampling moments; wherein the preset test temperature is higher than normal temperature; calculating a health index corresponding to each sampling moment according to the observation data corresponding to each sampling moment; fitting the plurality of sampling moments and the health indexes corresponding to the plurality of sampling moments based on a plurality of degradation track models to obtain a target fitting model; and calculating the service life of the wireless charger according to the target fitting model. Through systematic data acquisition, dynamic evaluation of health indexes, scientific degradation model fitting and reliable life prediction, the accuracy and effectiveness of life evaluation of the wireless charger are significantly improved.

A steel structure digital restoration device and method for digital twin construction, the device comprises a structure recognition module, the structure recognition module is configured to: the point cloud information of the acquired steel structure is executed in a virtual three-dimensional space according to its embedded space coordinate information based on the corresponding restoration operation of position, the total model field about the distribution of point cloud in space is acquired, and the restored point cloud information is in the form of a point model in the total model field;Based on the distribution of point cloud, the position with the best point model density is automatically selected, slicing is carried out to obtain the basic cross-section data of the structure;Based on the image in the basic cross-section data, the identification and judgment of the type of steel structure are executed, after correctly identifying the type of steel structure, the three-dimensional model fitting restoration of slicing is executed based on the identified type of steel structure.The present application can directly restore the three-dimensional pipeline model with high precision, and the engineering personnel can obtain the information of all steel structures in the whole factory by directly observing the three-dimensional steel structure model with sufficient rich details.

The invention discloses a radio astronomical interference data processing and analyzing method and system based on AIPS, and relates to the technical field of astronomical data processing. The method comprises the steps of data loading, imaging processing, image information extraction, source detection and parameter extraction, model fitting and result output. By integrating an AIPS data processing function and automatic workflow management, tasks such as fit, imagr, sad, jmfit and omfit of the AIPS are called, and end-to-end automatic processing from original observation data to scientific results in the whole process from astronomical data loading to source parameter extraction, model optimization and result visualization is achieved. The system comprises functional modules corresponding to the steps, supports multi-band data processing, can automatically identify a radio power source, performs Gaussian model fitting, analyzes parameters such as flow, position, form and polarization of the source, and generates a high-quality visual result. According to the method, the efficiency and precision of astronomical data processing are improved, the risk of manual operation errors is reduced, the efficiency and consistency of radio astronomical data processing are remarkably improved, and the method is suitable for batch processing and analysis of radio astronomical observation data.

The invention discloses an anomaly monitoring method for panoramic smart energy cloud edge collaboration, and the method comprises the following steps: S10, edge side credibility perception calculation: carrying out the real-time analysis of locally collected energy time sequence data through an edge side device, generating a preliminary anomaly probability Pe, and carrying out the calculation of the edge side credibility based on the probability Pe and the corresponding data; calculating an edge credibility score Re corresponding to the preliminary anomaly probability Pe by analyzing the difference with a historical normal mode and the uncertainty of local model fitting; and step S20, cloud credibility enhancement analysis: the cloud performs deep analysis according to the data uploaded by the edge side, generates a cloud anomaly probability Pc, and calculates a cloud credibility score Rc corresponding to the cloud anomaly probability Pc by analyzing the novelty of the data in global distribution and the stability of model prediction based on the probability Pc and the corresponding data. According to the invention, abnormity monitoring of panoramic smart energy cloud edge collaboration can be carried out more comprehensively.

The invention discloses a control method and equipment of a knowledge distillation system and a storage medium, and belongs to the technical field of data processing. The method comprises the steps of obtaining a condition vector of a middle layer of a to-be-learned model based on a classification processing action of the to-be-learned model on sampling sub-graphs; carrying out noise addition on the sampling sub-graph to obtain a noise-added sub-graph; through the first teacher model, performing a denoising action on the noise-added sub-graph according to the condition vector to obtain a reconstructed sub-graph, and calculating reconstruction loss of the reconstructed sub-graph and the sampling sub-graph; calculating the model loss of the to-be-learned model according to the reconstruction loss and the fitting loss of the to-be-learned model and a second teacher model; and iterating the to-be-learned model according to the model loss to obtain a student model. Through the double-teacher model, in the student model fitting learning process, the student model is guided to construct an understanding mechanism for the graph result, so that the essential performance is improved.

The application provides a kind of park lane line perception method and system based on multi-source data fusion and prior map constraint, it is related to map and positioning navigation technical field, this method includes: collecting multi-source data and pre-processing image;Extract color, edge and semantic feature for weighted fusion, generate candidate lane line heat map;Obtain high-precision map prior information and project to image coordinate system, spatial screening is carried out to candidate point;Adaptive model fitting is carried out to effective point set, and lane line observation value is obtained;Kalman filtering tracking and prediction are carried out in combination with vehicle kinematic model;Comprehensive multi-dimensional index calculates confidence and carries out hierarchical processing.The application improves the detection robustness in complex environment by multi-feature fusion, uses map prior to carry out flexible constraint to eliminate noise, introduces motion tracking to ensure continuity when shielding, and realizes abnormal response through confidence evaluation, effectively solves the perception problem caused by unclear park lane line feature and variable environment.

A ball machine system comprising a ball launching system, a memory configured to store a court model that includes dimensions of a court type in relation to a three-dimensional (3D) coordinate system, an imaging system connected to the ball launching system and configured to capture a plurality of images of a court, and a processor. The processor configured to analyze the plurality of images to detect feature points in each of the plurality of images, superimpose a plurality of different transformations of the court model onto the plurality of images to generate a plurality of model fits, select as a final model fit, from the plurality of model fits, a model fit based on a degree of overlap among features of the transformed court model and the respective image, and calculate the position of the ball launching system on a playing surface of the court based on the selected final model fit and intrinsic parameters of the imaging system.

The application discloses a parameter identification method for improving an industrial robot GMS friction model, and comprises the following steps: step 1, introducing a Sigmoid function to design a GMS friction model satisfying a switch point differential; step 2, constructing an improved Stribeck effect model of the GMS friction model; step 3, designing an excitation trajectory for Stribeck effect identification and identifying parameters based on a SOS algorithm; and step 4, designing a GMS dynamic parameter identification method based on SOS state point determination. The application satisfies the switch point differential, avoids model oscillation, reduces resonance of a servo system for robot joint control, more accurately describes the Stribeck effect, improves the precision of a classical GMS model, avoids empirical determination of state points, improves model fitting precision, and improves the identification precision of the GMS friction model as a whole.

The invention discloses an underground water flow field physical information neural network modeling method considering spatial heterogeneity, and belongs to the technical field of underground water flow fields, and the method comprises the steps: 1, selecting a two-dimensional underground water unstable flow equation as a physical law, collecting hydrogeological parameters, rainfall, underground water exploitation amount and geographic coordinate data, and carrying out the normalization preprocessing; 2, constructing a PINNs model based on the preprocessed data, dividing a data set according to 7: 3, dynamically adjusting the weight of a loss function by means of an NTK technology, and updating parameters through an optimization algorithm; 3, extracting a simulation residual error of the PINNs model, inputting the simulation residual error into a GTWR model for fitting, performing Kriging interpolation to obtain residual error space distribution, and superposing PINNs simulation results to obtain a final underground water flow field simulation result; according to the method, the weight of the loss function of the PINNs is dynamically adjusted by fusing the neural tangent kernel technology, the spatial heterogeneity characteristics of the hydrogeological parameters can be adapted, the weight imbalance of physical constraints and data information is avoided, and the model is guaranteed to meet the physical law and fit the actual data distribution.

The application discloses a kind of precision instrument with thermosetting adhesive curing kinetics model construction method;First, differential scanning calorimetry (DSC) experiment is carried out, and corresponding heat flow density curve is obtained;Second, each heat flow density curve is carried out peak fitting, and multiple exothermic peak curve is obtained;Third, the exothermic quantity curve, degree of cure curve and curing rate curve of each exothermic peak are obtained;Fourth, the kinetic model parameters of each exothermic peak are obtained by using autocatalytic model fitting;Fifth, the percentage of the exothermic quantity corresponding to each exothermic peak in the total exothermic quantity of adhesive curing is calculated;Finally, the real-time curing rate of each exothermic peak is calculated according to the corresponding kinetic model, and the adhesive real-time curing rate is obtained by linear superposition after being multiplied by the corresponding percentage;The application can significantly improve the accuracy of adhesive curing process analysis results, provide reliable theoretical support for precision instrument with thermosetting adhesive curing forming process optimization, and improve the performance of precision instrument.

The present application relates to the technical field of ship motion identification modeling, and discloses an intelligent ship maneuvering motion machine learning identification modeling method based on PINN. The present application first constructs an intelligent ship maneuvering motion fusion model: a mechanism model considering the influence of hydrodynamic force, rudder angle and propeller is constructed by referring to Abkowitz's overall model, a data-driven model fitting the response of uncertain disturbance is constructed, and then the two are weighted and fused; subsequently, based on physical information neural network modeling, the input and output of PINN are determined, a total loss function coupled with the physical loss function and the data loss function is designed, and the network parameters are trained and optimized through normalization processing and cyclic iteration. The Z-shaped maneuvering test data of the Mariner ship is verified, and the model is significantly improved in accuracy, robustness and interpretability, which can meet the needs of intelligent ship autonomous navigation.

A method for identifying a material boundary within volumetric image data is based on use of a model boundary transition function, which models the expected progression of voxel values across the material boundary, as a function of distance. Each voxel is taken in turn, and voxel values within a subregion surrounding the voxel are fitted to the model function, and the corresponding fitting parameters are derived, in addition to a parameter relating to quality of the model fit. Based on these parameters for each voxel, for each of at least a subset of the voxels, a candidate spatial point is identified, estimated to lie on the material boundary within the 3-D image dataset. The result is a cloud of candidate spatial points which spatially correspond to the outline of the boundary wall. Based on these, a representation of the boundary wall can be generated, for example a surface mesh.

The application discloses a traffic flow basic graph construction and parameter estimation method for fusing data of different time intervals, and comprises the following steps: establishing the correlation between the traffic flow of a short time interval and the traffic flow of a long time interval; determining the probability distribution function of the traffic density of the short time interval by using the traffic flow density of different time intervals, the traffic flow of the long time interval and the flow variance; constructing the traffic flow basic graph by using the traffic flow speed of the short time interval, the traffic flow of the long time interval, the traffic density of different time intervals and the variance of the traffic flow of the short time interval; and estimating the parameters of the traffic flow basic graph model by using the Bayesian hierarchical model and the updating method, so that the model fitting goodness is best after multiple iterations and updating. The application can effectively reduce the error of the parameter estimation value of the traffic flow basic graph, is helpful to estimate and predict the traffic state of a road section under the condition that complete traffic data is lacked, and has application value for relieving the urban road traffic congestion.

This invention discloses a method and system for correcting errors in galvanometer scanning, relating to the field of galvanometer scanning technology. The method includes acquiring galvanometer scanning system information, obtaining a basic error dataset based on calibration tests using the galvanometer scanning system information, and constructing a structured parameter sequence based on the basic error dataset. This invention achieves multi-dimensional and accurate characterization of error features by constructing a structured parameter sequence. It establishes a benchmark model using a nine-point correction method, divides sector regions, and calculates density correction coefficients based on feature distortion rates and benchmark distortion gradients, achieving differentiated distribution of sampling points. By selecting highly correlated swing angle feature datasets, the purity and reliability of the error model fitting are improved. By splitting the geometric error model and constructing a multi-dimensional comprehensive error model, comprehensive coverage and accurate separation of geometric distortion and inter-axis coupling errors are achieved, improving the accuracy and reliability of error correction.

This invention discloses a method for predicting the transport distance of landslides around tunnels based on multi-stage AI regression, belonging to the field of tunnel landslide prediction. The method includes the following steps: S1, collecting topographic remote sensing data, land use data, rainfall data, historical landslide data, sensor monitoring data, tunnel engineering data, and ecological data within the boundary of the study area; S2, dividing the study slope into tunnel disturbance zone, collapse zone, transport zone, and sedimentation zone to obtain a key feature dataset; S3, based on the multi-stage regional division results, targeting the key feature datasets of landslide transport distance for each stage, and using a multi-stage AI regression model to obtain the predicted total transport distance. This method for predicting the transport distance of landslides around tunnels based on multi-stage AI regression overcomes the limitations of traditional CTD (Cross-Tunnel Landslide Division) that ignore engineering disturbances and have insufficient single-model fitting capabilities, achieving high-precision prediction and engineering application of landslide transport distance in complex scenarios around tunnels.

The invention belongs to the field of geological modeling, and relates to a geological semantic constraint-fused implicit potential field three-dimensional modeling method, which comprises the following steps of: constructing an original mixed data set; encoding the unstructured text data, and constructing a knowledge graph; the fuzzy occurrence attributes are converted into space occurrence control points, and a soft constraint data set is formed; mapping the drilling data and the spatial occurrence control points in the soft constraint data set to a normalized three-dimensional coordinate system, and constructing a structured input matrix; calculating an error covariance matrix of the drilling data and the spatial occurrence control point in the soft constraint data set; in combination with the structured input matrix and the error covariance matrix, an implicit potential field interpolation is obtained through calculation, and an initial three-dimensional geologic model is constructed according to the implicit potential field interpolation; constructing a verification function calculation model fitting residual error to judge whether the local model and the prior geological knowledge have significant conflicts or not; and outputting the three-dimensional geologic model. The method has the beneficial effect that the compatibility and stability of the modeling process to multi-source heterogeneous data are effectively improved.

The application discloses a kind of mud pipeline extension automatic assembly control method and system based on mechanical arm, solve the existing technology in tunnel construction slurry shield mud pipeline extension pipeline assembly docking difficult, low degree of automation Problem.Through being arranged machine vision / radar sensor on pipeline extension walking device trailer, obtain mud pipeline grabbing, transfer and the plane pixel / space point cloud information in the process of assembly, adopt a series of methods such as edge detection, pipeline feature extraction, model fitting to realize the identification and spatial pose positioning of pipeline, realize real-time obstacle avoidance, alarm and optimal trajectory planning in the process of mechanical arm grabbing, transferring pipeline by carrying mechanical arm obstacle avoidance path and trajectory planning module running in industrial computer;Finally realize slurry shield mud pipeline extension automatic assembly whole-process operation, improve the degree of automation of tunnel construction, improve pipeline docking precision, reduce operation time, directly reduce construction period and construction cost.

The invention discloses a meteorological feature prediction method and system based on a deep residual network model, and relates to the technical field of meteorological feature prediction management. The meteorological feature prediction method based on the deep residual network model comprises the following steps: S1, trying to solve an air pressure tendency; s2, correcting the air pressure tendency; s3, carrying out model generalization evaluation; and S4, performing model fitting evaluation. According to the method, deep residual network model calculation is carried out through neural network parameters and meteorological element data to obtain an air pressure tendency trial solution, then a corrected air pressure tendency value is obtained according to the air pressure tendency trial solution and the meteorological element data, and then model generalization evaluation is carried out according to model training parameters to judge whether overfitting correction is carried out or not. And finally, performing model fitting evaluation through the obtained prediction error parameters to judge whether to adjust the neural network parameters, thereby achieving the effect of improving the accuracy of meteorological feature prediction, and solving the problem of low accuracy of meteorological feature prediction in the prior art.

The invention relates to a drug sensitivity analysis and judgment system and method based on deep learning and a microbial growth curve. The core of the method is as follows: fitting is carried out by adopting parameterized models such as Gompertz or Logistic to extract kinetic parameters such as the maximum growth rate and the hysteresis period; then calculating a residual sequence between a growth curve observation value and a model fitting value; jointly inputting the kinetic parameters, the residual sequence and the metadata into a mixed deep learning model (1D-CNN-BiLSTM-Attention) so as to estimate a drug sensitivity category posterior probability at each time point; and finally, dynamically calculating the optimal stop time based on the posterior probability through a set of sequential decision-making module, and outputting a drug sensitivity judgment result in advance before the microorganisms grow completely when preset confidence and stability conditions are met. According to the method, the defects that a traditional endpoint method is long in period and an existing rapid method depends on single-point imaging are overcome, rapid, accurate and interpretable drug sensitivity analysis can be achieved within 6-8 hours, and timely guidance is provided for clinical anti-infection treatment.

The application discloses a multi-modal depth perception and grasping system based on transparent objects, and relates to the field of robot operation, comprising a multi-spectral perception module, a depth correction module, a grasping posture generation module and a control module. The application comprehensively acquires visual information and thermal radiation information of the transparent object by combining two perception methods of RGB-D image and thermal imaging (TIR) image, and analyzes systematic errors of the depth map to detect error sources of the RGB-D camera and the TIR camera. The system adopts an encoder-decoder model as a depth correction core, the model extracts complementary features of the RGB-D image and the TIR image through a modal exclusive encoder, and utilizes a feature fusion module to complete feature alignment and integration, thereby effectively improving the depth estimation accuracy on the transparent surface. Meanwhile, a Bayesian optimization method is adopted to optimize hyperparameters of the depth correction model, through hyperparameter optimization and model fitting processing, the system effectively avoids overfitting and underfitting problems, and further improves the robustness of the depth correction model and the grasping accuracy of the transparent object.

The invention belongs to the field of ecological environment protection and treatment, and discloses a water ecological environment deep learning prediction method based on space-time heterogeneity nesting. The method comprises the following steps: S1, preprocessing water ecology simulation prediction data; s2, fitting a water ecology basic model; s3, watershed spatial element information embedding and serialized interpolation are carried out; and S4, constructing a sampling point water ecological index prediction model. On the basis of a deep neural network algorithm coupling gradient lifting method, a nonlinear response relationship between environmental elements and water ecology evaluation indexes is fitted, and the problem of simulation difficulty caused by a complex physical, chemical and biological multi-process interaction mechanism is solved; the spatial distribution data and the time sequence data are nested and coupled, the spatial correlation data are serialized and converted, and the time sequence data are embedded, so that the time resolution of the sampling data is improved, and the accuracy and applicability of model simulation are improved.

The application discloses a kind of navigation navigation test method and system, method includes: using camera and rudder angle signal collector to collect the navigation image and rudder angle data of model boat in the test navigation section of hydraulic model, constructs camera calibration model, fitting time mapping relationship model, using linear interpolation to obtain time synchronization navigation image, rudder angle data;To the time synchronization navigation image is preprocessed and identified, obtain the coordinates of feature point in pixel coordinate system in the continuously collected navigation image;Calculate the heading angle, roll angle and pitch angle of model ship at different intervals;Draw the time-drift angle / rudder angle relationship curve of model ship in the test navigation section during test process, the navigation safety of test navigation section is evaluated.The system includes evenly arranged on the test navigation section of hydraulic model N Camera, model boat, rudder angle signal collector and remote controller of model boat installed on model boat.

The invention relates to the technical field of image processing, and discloses a multi-camera correction and boundary multi-scale seamless splicing method and system. The method comprises the following steps: acquiring a punctuation pair set between a video frame of each path of deployed camera and a reference overlook base map; fitting through a primary function model to obtain a nonlinear geometric mapping function from a camera pixel coordinate system to a reference overlook base map coordinate system; resampling and correcting the video frames of the cameras to a reference overlook base map coordinate system in real time by using a geometric mapping function, and performing multi-scale feature alignment on the obtained overlook correction frames of the cameras in an overlapping region to estimate local refinement transformation, and carrying out local refinement transformation on one of every two adjacent overlook correction frames, and then carrying out multi-scale frequency band fusion to generate a panoramic spliced video frame. The invention provides a multi-camera correction and boundary multi-scale seamless splicing method based on sparse punctuation constraint, which can realize real-time panorama of one image.