141 results about "Spatial registration" patented technology

The application discloses a kind of multi-modal medical image segmentation method, device, computer equipment and storage medium, comprising: obtaining the multi-modal medical image of the same examinee and pre-processing and spatial registration;The target segmentation concept is standardized and analyzed to generate prompt;Using basic segmentation model obtains multiple candidate segmentation results of each modality;Constitute credibility gating model, assess the credibility score of each candidate result and carry out screening, obtain the credible instance of each modality;Establish the corresponding relationship between cross-modal credible instance to form instance group;Each instance group is executed evidence weighted fusion and conflict resolution, and the final segmentation result of each target is obtained;Output multi-instance segmentation result and corresponding instance-level uncertainty index.The application can effectively reduce the false detection and result fluctuation of basic model in multi-modal image, improve the consistency and reliability of cross-modal segmentation, and provide robust segmentation result for clinical screening.

This invention discloses an AR-assisted puncture navigation spatial registration method and system based on environmental reference features, relating to the field of medical image navigation. The system includes an augmented reality navigation device, a permanent ceiling marker, and a temporary marker board. The augmented reality navigation device includes a binocular augmented reality camera group, a global tracking camera, a global spatial registration module, and a display module. The binocular augmented reality camera group constructs a binocular overlapping field of view, acquiring the pose information of the binocular augmented reality camera group relative to the temporary marker board during the calibration phase, and acquiring near-field images of the surgical area in real time during the surgical phase and displaying them on the display module. The global tracking camera acquires the pose information of the global tracking camera relative to the global coordinate system of the ceiling. The global spatial registration module solves the extrinsic parameter matrix between cameras and the environmental relationship matrix based on the pose information. During the surgical phase, the path planned by the imaging equipment is converted into a real-time navigation path and superimposed on the display module. This solves the complex calibration problem under non-overlapping fields of view.

This invention discloses a computer vision-based method for anomaly identification in smart water conservancy dams, belonging to the field of intelligent visual monitoring technology for water conservancy dams. The method includes acquiring a time-series image set covering the entire dam area, generating a standard observation dataset through multi-scale spatial registration and illumination normalization, constructing a digital twin of the dam containing topographic relationships, recording surface material properties, and elevation distribution information, mapping its surface grid data to the standard observation dataset, and generating a texture difference map with geographic coordinates. A rule library storing the evolution of dam deformation thresholds and seepage characteristic patterns under different hydrological conditions is loaded and stored. The texture difference map is input into the library for comparison and analysis, outputting anomaly candidate areas, aggregating the spatial distribution density and duration parameters of the anomaly areas, and generating a dam safety status assessment report. This method can mitigate the interference of illumination and viewing angle, achieving accurate anomaly location and systematic determination of the overall safety status.

The present application belongs to the technical field of PCB defect detection, and particularly relates to a PCB back-drilling defect detection method, a detection device and an evaluation method. The method first identifies and preliminarily screens the holes to be rechecked through the two-dimensional image of the whole board; then performs three-dimensional topography scanning on the holes to be rechecked, and performs spatial registration on the two-dimensional image subgraph and the three-dimensional point cloud data matrix of the same hole through a unified coordinate system, so that the two-dimensional texture and the three-dimensional topography information are accurately corresponded; based on the fused data after registration, residual stubs, copper wire residues, dirt, microcavities, depressions, scratches and hole wall roughness are identified and classified. The present application corresponds the two-dimensional preliminary screening result and the three-dimensional measurement data through spatial registration, so that the copper wire and the dirt, the microcavity and the scratch can be distinguished based on the three-dimensional topography difference, and the size parameters such as the residual stub length and the hole wall roughness are included in the quality evaluation.

This invention belongs to the field of industrial vision intelligence and CNC machine tool processing technology, specifically relating to a method for determining machining allowance of CNC machine tool workpieces based on image recognition. The method includes: retrieving the design digital model point cloud; preprocessing the original point cloud data of the workpiece to be processed to obtain the workpiece point cloud; calculating the scanning confidence, solid surface weight, and liquid removal weight of each data point in the workpiece point cloud, and combining them to obtain the final usable weight; calculating the spatial alignment error based on the final usable weight; performing iterative registration between the workpiece point cloud and the design digital model point cloud; solving the three-dimensional rotation matrix and three-dimensional translation vector; aligning the spatial pose of the workpiece point cloud; and extracting the machining allowance data. This invention improves the spatial registration stability of unstructured visual data and enhances the accuracy of adaptive machining on CNC machine tools.

The application relates to the technical field of power transformation installation, and relates to a power transformation station equipment installation and wiring guiding system based on augmented reality, which comprises a three-dimensional scene reconstruction module, a visual inertial positioning module, a space registration module, an installation guiding module and a wiring path guiding module. Three-dimensional point cloud data of a target power transformation station scene is subjected to point cloud fusion reconstruction to obtain a three-dimensional space model. An environment image and motion data of the target power transformation station scene are collected through a wearable augmented reality device, feature points are matched and optimized to obtain a real-time space pose of the equipment, coordinate registration is completed, and a virtual-real fusion space is constructed. A three-dimensional model of equipment to be installed is superimposed on a corresponding physical position, an installation-in-place signal is generated when the spatial coincidence degree meets preset conditions, a three-dimensional wiring path is generated according to terminal definition and is projected to guide wiring operation. The application can improve the efficiency of power transformation station equipment installation and wiring.

The application provides a multi-modal caries depth intelligent interpretation method fusing intraoral photos and X-ray films, and through standardized oral multi-modal medical image and pathological data synchronous collection, spatial registration and pixel-level normalization, the spatiotemporal consistency of clinical images and histopathological data is realized, the application utilizes a double-channel convolutional neural network and a lightweight LSTM, applies a physical constraint gate unit to perform channel re-labeling on a double-modal shallow feature map, generates a re-labeled feature map, and after splicing, outputs a caries depth grading interpretation result through a decoding network, so that the accuracy and consistency of caries depth grading interpretation are effectively improved.

The application relates to a three-dimensional spectrum combined PCB detection method and system based on multi-modal fusion. The detection method comprises the following steps: acquiring three-dimensional topographic data and multi-band spectral reflection data, and collecting the three-dimensional topographic data after structure light projection; performing cross-modal space registration processing according to the three-dimensional topographic data and the spectral reflection data to generate corresponding registration feature data; determining the defect state of the PCB surface; when the defect state is an appearance similar defect state and a material attribute difference state, corresponding fusion feature encoding signals are generated, spectral set consistency verification is performed, and a direct output defect recognition action is performed or dynamic compensation processing is performed until the conditions for the direct output defect recognition action are met. The same detection area can be described from two physical dimensions of geometric height change and material optical reflection difference, and the difference can be distinguished, so that more comprehensive and more reliable defect determination results can be obtained.

The application relates to the technical field of industrial nondestructive testing, and particularly discloses a multi-modal industrial data three-dimensional fusion and visual detection method and system based on brain-like perception. The method is aimed at the problem of multi-modal data heterogeneity and fusion difficulty in synchronous detection of internal and external defects of a weld, and surface optical and internal ultrasonic data are synchronously collected through a binocular camera and an ultrasonic array sensor. After denoising, time sequence alignment and spatial registration, a human brain multi-modal information processing mechanism is simulated, time and space features are respectively extracted based on EEG and fMRI inverse problem solving ideas, and a composite space-time metric is constructed for weighted fusion. Then, three-dimensional reconstruction of internal and external defects of the weld is realized through tensor decomposition and joint inverse solution, and a visual image is generated. Finally, positioning, identification and quantitative analysis of the defects are completed through a lightweight detection network. The application realizes integrated, high-precision and visual synchronous detection of internal and external defect states, and significantly improves the comprehensiveness and reliability of industrial weld nondestructive testing.

The present application relates to the technical field of vacuum cup production detection and intelligent manufacturing, in particular to a vacuum cup surface flaw rejection method based on visual detection, comprising: obtaining two-dimensional image data and three-dimensional depth data of the vacuum cup surface and performing spatial registration combination; extracting the surface flaw contour based on an image segmentation network, calculating the geometric sharpness, depth gradient and physical distance of the relative key stress area; inputting a stress concentration prediction proxy model and outputting the predicted local maximum stress multiple under simulated thermal cycling and vacuum negative pressure environment; comparing with the fatigue limit multiple threshold to generate a functional failure probability score, and combining with the appearance evaluation benchmark to generate an appearance degradation score; executing classification rejection decision according to the two types of scores and outputting the judgment result; the present application can distinguish between appearance flaws with less functional impact and surface flaws with weak visual representation but may lead to structural failure, and both detection accuracy and long-term reliability of the vacuum cup are taken into account.

This application provides a method for image fingerprint recognition of small unmanned aerial vehicles (UAVs) based on heterogeneous information fusion, belonging to the field of UAV detection and recognition technology. The method includes: acquiring visible light image sequences and infrared image sequences of the target UAV, performing time synchronization, spatial registration, and scale normalization; extracting structural texture features, dynamic differential features, and infrared energy criterion feature vectors; inputting these into a cross-scene decoupled hypergraph fingerprint network for fusion encoding to obtain a low-dimensional unique image fingerprint vector; training the network using cross-scene consistency constraints and scene identity orthogonality constraints to establish a UAV image fingerprint database; matching the fingerprint vector of the target to be identified with the fingerprint database, using an open set unknown individual recognition mechanism to determine the matching result, and outputting the recognition result or the result for an unknown target. This application relies solely on image information to achieve fine-grained recognition at the model and individual levels, and has the advantages of low deployment cost, strong cross-scene stability, and high reliability of open set recognition.

The application relates to the technical field of electronic data digital processing, and discloses a two-end full-frame support beam stress state in-situ finite element simulation method based on mixed reality. The method obtains a three-dimensional geometric model of a support beam and a support base through a mixed reality device, and utilizes a high-frame-rate camera to capture full-field displacement data under an excited state; a finite element model taking the rotation stiffness of the support base as an unknown parameter is constructed, and a real stiffness value is reversely identified through an iterative optimization algorithm; the corrected model is solved under a design load, and stress, strain and deformation nephograms are spatially registered and superimposed on the real structure through the mixed reality device. The system comprises five modules, namely, three-dimensional perception, micro-variation capture, boundary identification, model updating and mixed reality visualization. The application realizes accurate in-situ boundary condition identification, non-contact nondestructive testing and immersive visual evaluation, and significantly improves the accuracy and efficiency of structural safety analysis.

The present application relates to the technical field of data processing, in particular to a data processing method and system based on unmanned aerial vehicle video and vector data fusion. By calculating the deviation of the predicted position and the actual detection position of each reference point in the image, the reference points with large errors are screened and the camera pose parameters are adjusted using the random consistency algorithm and the iterative optimization method to improve the projection accuracy and stability. Through grid division, corner detection and feature descriptor extraction, a reference point set is established, the parallax change standard deviation is introduced, and the unstable reference points with large fluctuations are removed to ensure the reliability of the pose solving process and the consistency of spatial registration. According to the registration and fusion of unmanned aerial vehicle video and vector geographic data, the problems of current vector element drift, occlusion determination difficulty and attitude error dynamic correction are solved, the spatial reality and dynamic stability of the augmented reality scene are enhanced, and the industrial needs of various application scenarios such as underground pipe network inspection are met.

This invention discloses a single-snapshot spatial frequency domain imaging and morphology correction method, system, and device, comprising: acquiring a set of monochromatic fringe images with different center wavelengths based on the same projected image; simultaneously inputting each monochromatic fringe image in the set into two parallel and independent deep learning networks, wherein the first deep learning network is used to extract the DC and AC components of the image; the second deep learning network is used to extract the wrapping phase component and fringe hierarchy information of the image, and perform absolute phase calculation to obtain the three-dimensional morphology before correction; performing height correction on the three-dimensional morphology and calculating the corrected diffuse reflectance image; performing background segmentation on the corrected diffuse reflectance image and spatially aligning it to the same coordinate system; and retrieving the optical characteristic parameters and physiological parameters of the tissue based on the spatially registered multi-wavelength diffuse reflectance data. This invention ultimately achieves real-time, wide-field, high-precision, and highly robust quantitative imaging of tissue physiological parameters.

The application relates to a virtual reality-based intangible cultural heritage immersive experience method, device, equipment and medium. The method comprises the following steps: acquiring multidimensional data of a target intangible cultural heritage scene; performing grid processing on original three-dimensional point cloud data to obtain an initial three-dimensional geometric model; performing spatial registration on a spectral reflection curve and the initial three-dimensional geometric model to obtain a three-dimensional geometric model containing material spectral properties; analyzing a facet spectral emission curve in combination with a scene space-time cultural database to generate a parameterized physical material library; generating a basic light image in combination with the parameterized physical material library and the scene space-time cultural database according to experience space-time information selected by a user; fusing environmental special effect physical description parameters and the basic light image to obtain a synthesized visual picture result; and generating an immersive experience instruction based on the synthesized visual picture result and environmental special effect data. The method can dynamically adapt to material characteristics and space-time changes of different intangible cultural heritage scenes, and improves the restoration degree of the intangible cultural heritage scene and the user immersion.

This invention relates to the field of process control technology, and discloses an automatic control method and system for actual drilling trajectory comparison deviation. The method includes: performing coordinate analysis on the inclination angle and azimuth angle at the well depth and the spatial coordinate sequence of the designed trajectory at the well depth to obtain a spatial coordinate sequence; performing sliding window comparison analysis on the spatial coordinate sequence and the spatial coordinate sequence of the designed trajectory at the same well depth to obtain a multi-scale deviation feature set; performing nonlinear mapping inference on the multi-scale deviation feature set to obtain a dynamic control target vector; performing correlation analysis on the tool face angle in the dynamic control target vector and the measured tool face angle to obtain a bottom-level drive command sequence; performing spatial registration evaluation on the spatial coordinate sequence and the preset design trajectory to obtain a spatial deviation vector sequence; and adaptively adjusting the bottom-level drive command sequence to obtain a target drilling scheme. This invention can improve the efficiency of automatic control of actual drilling trajectory comparison deviation.

The application relates to the field of edge detection, and specifically discloses an edge detection and correction method in an additive manufacturing process, which synchronously collects a visible light image and a temperature distribution image of a printing layer edge region by using a visible light camera and an infrared thermal imager, extracts an actual edge coordinate sequence from the visible light image through contour extraction and fitting, and performs spatial registration and geometric deviation calculation on the actual edge coordinate sequence and target layer edge contour data to generate an edge deviation map. Furthermore, a deep learning algorithm is used to perform multi-level correlation analysis on the actual temperature distribution image and the edge deviation map to deeply mine the dynamic response characteristics of edge deformation under thermal coupling, so that specific reasons for causing edge deviation can be intelligently judged, and process parameter correction instructions can be generated based on the deviation reasons and the quantized edge deviation map. The method can realize intelligent identification and targeted adjustment of edge defect causes in the additive manufacturing process, and thus improves the yield and stability of the additive manufacturing.

This invention discloses an adaptive intelligent welding robot system and method based on multi-source information fusion, relating to the fields of intelligent welding and robot control technology. The system integrates a multi-source heterogeneous sensing module, a spatiotemporal registration and synchronization module, a three-level information fusion module, a digital twin and incremental learning module, a trajectory-parameter collaborative control module, and an execution drive module. It synchronously collects multi-dimensional welding data through multiple sensors, including laser 3D vision. After time synchronization and spatial registration, it completes three-level information fusion using Kalman filtering, hybrid feature selection, and D-S evidence theory. Combined with multiphysics twin simulation and federated incremental learning, a fractional-order PID controller synchronously generates trajectory correction and optimized process parameters. Adaptive welding is then performed by a six-axis robotic arm and other execution components. This invention significantly improves welding accuracy, stability, and adaptability, meeting the high-precision and high-efficiency welding requirements of fields such as bridges and new energy equipment.

This application relates to the field of interdisciplinary technology, and in particular to a stroke rehabilitation training system based on a mixed reality digital human. The system includes: a data acquisition module configured to: acquire multimodal data of the subject performing standard movement tasks under the guidance of the rehabilitation therapist's digital human; the multimodal data includes: visual data, electromyographic data, and inertial data; perform perception and heterogeneous spatial registration on the multimodal data; and drive the subject's mirror digital human for synchronous display in real time; and a processing module configured to: identify the subject's movement intentions, assess upper limb motor function, detect compensatory movements, and assess lower limb motor function based on the multimodal data; and generate real-time prompts, visual feedback, and training reports, thus achieving an integrated closed loop of stroke rehabilitation training, functional assessment, and feedback adjustment. It is applicable to stroke rehabilitation training, functional grading assessment, and home-based continuous rehabilitation intervention.

This invention relates to the field of artificial intelligence and industrial nondestructive testing, and discloses a multi-processor nondestructive defect detection method based on multimodal information fusion. The method includes: spatial registration of CT data and infrared thermal imaging data; dividing the registered data into multiple image blocks and extracting primary features; screening based on D-S evidence theory to identify suspicious image blocks; inputting the CT image blocks and infrared image blocks corresponding to the suspicious image blocks into a dual-stream convolutional neural network to output a set of suspicious image blocks; marking the connected regions of the output set, merging adjacent image blocks of the same category, calculating defect parameters, filtering out noise regions, and generating the final detection result. This invention can achieve accurate detection of various internal defects such as pores and cracks by fully utilizing the internal structural information provided by CT and the thermal diffusion anomaly information provided by infrared without damaging the multi-processor workpiece, and effectively handles sensor information conflicts and uncertainties.

The application relates to a brain region segmentation method based on multi-sequence magnetic resonance imaging collaborative feature fusion, which comprises the following steps: collecting brain amplitude graph data, phase graph data and T1 structure image data; carrying out pretreatment and reconstruction processing on the amplitude graph data and the phase graph data to obtain a quantitative magnetization rate image; carrying out segmentation labeling after spatial registration of the T1 structure image data and the quantitative magnetization rate image to construct a brain region segmentation label data set; training a CIA-Net deep learning model through a training set in the brain region segmentation label data set to obtain a trained CIA-Net deep learning model; inputting the quantitative magnetization rate image to be segmented and the T1 structure image data into the trained CIA-Net deep learning model to output an automatic segmentation result of the brain region. The application can realize automatic fine segmentation of a cranial magnetic resonance image under the conditions of fine granularity and a large number of brain regions, and improve the accuracy, structural consistency and stability of the segmentation result.

This invention discloses a neutron-gamma dual-mode synchronous imaging method and system based on single-pixel detection, belonging to the field of nuclear imaging technology. The method includes: irradiating the sample under test with modulated neutron beams of different encoding modes, identifying the neutron event count and gamma-ray energy spectrum for each irradiation; dividing the sample under test into several pixel units, reconstructing a neutron transmission structure image based on the neutron event count for each irradiation; obtaining the detection efficiency of the characteristic gamma rays received by each pixel unit of the sample under test, and constructing a gamma response modulation matrix in combination with the neutron transmission structure image; extracting the net count of the characteristic peaks of the target element from the gamma-ray energy spectrum, reconstructing an element distribution image in combination with the gamma response modulation matrix, and fusing it with the neutron transmission structure image for output. This invention achieves synchronous acquisition of structural and elemental information, automatic spatial registration, and joint high-precision reconstruction, and is particularly suitable for the engineering application of miniaturized neutron sources.

This invention discloses a method, device, equipment, and medium for monitoring and early warning of agricultural food-water-energy-carbon emissions, belonging to the field of smart agriculture technology. The method includes: acquiring production big data of a target area; performing spatial registration based on the production big data and corresponding units in a preset static basic spatial database to construct a performance-environmental condition paired dataset; inputting the performance-environmental condition paired dataset into a causal inference engine to identify direct environmental factors that have a direct causal relationship with crop yield, and eliminating interference factors that only have a correlation relationship, generating a set of sensitive factors for triggering early warning; based on the set of sensitive factors, using a threshold search algorithm to automatically identify the optimal early warning threshold for each key factor affecting crop yield; and monitoring energy crops in the target area based on the set of sensitive factors and the corresponding optimal early warning threshold to obtain monitoring results. This application realizes the monitoring and early warning of land, food, energy, and carbon emissions.

The invention provides a satellite cable assembly system and method based on AR visual guidance, and the method comprises the steps: a model preprocessing step: converting a satellite cable model format, and exporting an FBX format model; the FBX format model is compatible with a Unity engine; an AR tracking and positioning step: establishing an image target in a Unity engine, and establishing a virtual-real space registration reference; the FBX format model is imported into a target folder of a Unity engine; a process visualization guiding step: converting a port report containing cable assembly information into a specified format, importing the port report into Unity, and configuring a script, so that an assembly path and port number information can be fused and displayed with the model in an AR scene, and a configured Unity project is obtained; and a deployment step: packaging the configured Unity project into an installation package compatible with the AR equipment, and deploying the installation package to the AR equipment. According to the method, the operation cognition burden is reduced through visual guidance of virtual-real fusion, the assembly precision and efficiency are improved, AR equipment such as PICO and HoloLens is adapted, and good compatibility and practicability are achieved.

The application provides a soft tissue real-time tracking and navigation method based on inter-frame deformation field guidance and a related device. The method comprises the following steps: S1. acquiring an intraoperative endoscope image sequence, and extracting a current frame and a historical frame; S2. performing dense optical flow estimation on the current frame and the historical frame to obtain an inter-frame deformation field; S3. performing spatial transformation on a historical mask based on the inter-frame deformation field to obtain a deformation compensation segmentation prior; S4. performing prior guided fusion on the segmentation prior and a current frame feature to obtain a refined anatomical segmentation mask; S5. performing key point matching based on the refined mask and a pre-stored atlas to obtain a spatial registration parameter; and S6. rendering a safety boundary based on the registration parameter and the deformation field, and outputting a safety navigation view. The application also provides a related device corresponding to the method, and the related device comprises a device, an electronic device, a computer readable storage medium and a computer program product.

The application belongs to the technical field of intelligent medical treatment assistance, and provides a multi-organ motion coupling compensation system based on a space-time prediction model and an execution method thereof, which comprises a multi-source heterogeneous data acquisition and preprocessing module, a double-dimension gold standard construction module, a space-time fusion prediction model construction module, a training and verification module, a time sequence trajectory prediction module and a motion compensation instruction generation module. Through time sequence synchronization and space registration of multi-modal data, the application uses a space-time Transformer architecture to construct a prediction model, accurately quantifies multi-organ motion coupling coefficients, and predicts an output motion trajectory that is physically reasonable and meets accuracy standards. Meanwhile, relying on double gold standards and an optimized training strategy, combined with a parameter efficient fine-tuning strategy, the application realizes personalized adaptation for specific patients, solves defects such as low multi-source data synchronization accuracy and insufficient coupling relationship modeling in existing radiotherapy motion compensation technologies, and promotes the clinical landing and upgrading of precise radiotherapy technologies.

This invention relates to the field of commercial vehicle control, specifically to a method and system for the safety and energy conservation of hybrid fleets using multi-data fusion and reinforcement learning. The method involves collecting multi-source data from each vehicle in the fleet, performing time synchronization and spatial registration of the multi-source data to obtain a fused dataset, inputting the fused dataset into an adaptive multi-source extended Kalman filter for road slope estimation, and generating slope estimation results. Based on the slope estimation results and a pre-set desired speed trajectory, an improved vehicle longitudinal dynamics model is constructed. Based on the improved vehicle longitudinal dynamics model, desired acceleration and target distance control signals are obtained. The desired acceleration is converted into vehicle power demand, and a power control command is output. Based on the power control command and the target distance control signal, braking energy recovery is performed through multiple mode switching. This invention achieves high-precision perception of slope and environment, adaptive energy allocation, and coordinated safety control of the fleet.

This application provides a multimodal ultrasound intelligent identification method and precision diagnosis system for thyroid lesions. The method determines image data blocks with multimodal spatial registration and grayscale-contrast temporal alignment. Based on the grayscale ultrasound image sequence in the image data blocks, a first depth feature map of the target user's thyroid region is determined. A second depth feature map of the target user's thyroid region is determined based on the elastography map in the image data blocks. The hemodynamic spatial distribution characteristics of the parametric map in the image data blocks are captured to determine a third depth feature map of the target user's thyroid region. Cross-modal fusion is performed on the first, second, and third depth feature maps, and then the lesion segmentation mask, benign / malignant probability value, and risk classification of the target user's thyroid region are inferred and output in parallel. Using the scheme of this application, high-precision spatiotemporal alignment and cross-modal depth feature fusion of multimodal ultrasound images can be achieved to complete end-to-end intelligent identification of thyroid lesions.

This invention discloses a method and apparatus for quality inspection of polarizers on liquid crystal glass substrates based on image processing, belonging to the field of image inspection technology. The method includes: firstly, structural identification and region modeling of the liquid crystal glass substrate under test are performed, and a detection benchmark is established; then, multi-condition images of the same detection area are acquired under different detection states, and a reference image set and a detection image set are constructed; subsequently, spatial registration, background decoupling, defect enhancement, and layer discrimination are performed on the images; finally, quality evaluation results and process feedback are output based on defect type, layer position, and location information. This scheme can distinguish and detect surface defects, bonding interface defects, conductive layer-related anomalies, and liquid crystal functional anomalies, improving detection accuracy, evaluation consistency, and defect traceability in the manufacturing process, while also adapting to the differentiated detection needs of the central optical region, edge bonding region, conductive connection region, and partition boundary region.

This invention discloses a method, system, device, and storage medium for hail cloud feature analysis and identification based on multi-source data. The method includes: obtaining satellite feature products, derived feature products, and radar feature products strongly correlated with the hail cloud formation and dissipation process; performing spatial registration processing on the satellite feature products, derived feature products, and radar feature products to obtain spatiotemporally aligned satellite feature sets and radar feature sets; constructing a dataset based on the spatiotemporally aligned satellite feature sets and radar feature sets; constructing a GSCF deep learning recognition model; and training and testing the model. This application effectively solves the problem of heterogeneous fusion of multi-source data, fully leverages the advantages of macroscopic meteorological information from satellite data and fine structural information from radar data, improves the quality and representational ability of fused features, provides reliable data support for accurate hail cloud identification, and is of great significance for improving disaster early warning efficiency.