47 results about "Frequency domain decomposition" patented technology

This invention belongs to the field of image processing and computer vision technology, and discloses a lightweight single-image super-resolution reconstruction method based on a directional stripe hybrid network. It captures directional structural information of the image from both the spatial and frequency domains through horizontal stripes, vertical stripes, local window attention of the directional stripe self-attention module, and direction-specific convolution of the directional high-frequency enhancement module. The directional stripe self-attention module effectively captures long-distance dependencies extending along specific directions; the directional high-frequency enhancement module extracts high-frequency components through frequency domain decomposition, combines them with direction-specific convolution, and introduces a dual-guided adaptive enhancement mechanism. It identifies texture-rich regions through local variance maps and structural boundaries through edge maps, generating adaptive weights to guide feature enhancement, accurately enhancing high-frequency edges and textures with different orientations while effectively suppressing noise. The two work together to achieve spatial-frequency domain dual-level directional modeling, significantly improving the reconstruction quality of images rich in directional textures, such as architectural scenes and cartoon images, while maintaining a lightweight design.

The application discloses a rice disease image segmentation method based on frequency domain modulation and topological constraint, which comprises the following steps: 1, frequency domain decomposition is carried out on rice disease image data to obtain amplitude spectrum and phase spectrum features; 2, a frequency domain adaptive modulation module is constructed to selectively enhance the frequency components of the structural key information; 3, the topological invariants of the image are extracted through a topological constraint module, and the reference topological structure is obtained by down-sampling the label to construct a topological consistency loss; 4, the topological consistency loss and the rice disease image segmentation loss are jointly optimized, the model training under multi-level structure constraint is realized, and a rice disease image segmentation model is constructed. The sensitivity of the model to the structure features of the rice disease lesion is enhanced through frequency domain modulation, and the morphological consistency of the segmentation result is maintained through topological constraint, so that the robustness of the model under natural conditions such as light change and background complexity difference is effectively improved, and the performance of the disease image segmentation in an open farmland scene is improved.

This application relates to a feature extraction apparatus, device, and storage medium for time series data, belonging to the field of artificial intelligence technology. The method includes: acquiring a time-domain signal corresponding to a user's time-series data, wherein the time-series data represents the user's economic behavior data; performing N-level frequency domain decomposition on the time-domain signal using an N-level filter bank to obtain multiple time-series sub-band signals of different frequency bands, wherein N is a positive integer greater than or equal to 1; and performing feature extraction on each of the time-series sub-band signals to obtain the target features of the time-series data. This application can acquire the performance of time-series data in different frequency bands and can extract more features from the time-series data.

This invention discloses a method and system for generating digital watermarked images. The method includes: acquiring an original image to be watermarked; generating an image coding mask based on the image complexity of the original image; performing frequency domain decomposition on the original image to obtain a low-frequency component image; embedding an initial watermark image into the low-frequency component image according to the embedding strength values ​​corresponding to the masks in the image coding mask to obtain a low-frequency watermark image; different masks in the image coding mask correspond to different embedding strength values; and performing an inverse frequency domain transform on the low-frequency watermark image to obtain a synthesized watermark image. This invention can improve the success rate and accuracy of watermark extraction.

This invention relates to the field of defect detection technology in image recognition, specifically disclosing a method and system for detecting defects in mobile phone SIM card tray manufacturing based on image recognition. By acquiring multiple frames of time-series images and performing sub-pixel-level spatiotemporal registration, the influence of mechanical vibration and displacement deviation is eliminated. The optical flow field is calculated, and local strain tensors, optical flow vorticity, and motion consistency entropy are extracted to construct a time-varying feature matrix. Frequency domain decomposition and motion amplification are performed on the feature matrix to enhance the signal response of minute defects. Dynamic time-domain features and static image features are fused to establish a deep learning-based defect detection model. Finally, a defect probability cloud map is generated, and the defect type and location are output using a dynamic threshold mechanism. This invention achieves high-precision and high-efficiency automatic detection of minute defects in mobile phone SIM card trays, exhibiting good stability and adaptability.

This invention specifically relates to a method for comprehensive stability analysis of three-phase current in range extender controllers, comprising the following steps: First, constructing a multi-condition gradient acquisition set covering the entire operating range of the range extender; acquiring time-domain data of three-phase current under all operating conditions and performing segmentation and frequency domain decomposition; then, determining the fundamental and harmonic characteristic frequencies based on motor parameters, and extracting the frequency domain amplitude sets of each order under all operating conditions; subsequently, calculating three types of stability coefficients: fundamental order, conventional harmonics, and unconventional harmonics; finally, calculating the comprehensive stability coefficient based on weighting rules, and completing the comprehensive current performance rating. This method enables pre-emptive, full-condition, quantitative analysis and comprehensive performance rating of three-phase current stability in range extender controllers, identifying control defects in advance, shortening product development cycles, and reducing R&D costs.

The application discloses a tunnel water mist environment under inspection image de-scattering enhancement and reconstruction method, and belongs to the technical field of image processing. The method comprises the following steps: acquiring a polarized image sequence and pose information collected synchronously; decoupling a scene by using a fusion neural network to output a fog-free image, a geometric gradient map and a water mist concentration map; calculating a frequency domain fusion weight map by using a frequency domain attention generation network; performing multi-scale frequency domain decomposition on the fog-free image to obtain high-frequency components and low-frequency components; performing pixel-level weighted fusion and inverse transformation according to the weight map to generate a preliminary enhanced image and perform color correction. The application adopts a physical constraint deep learning and a frequency domain adaptive fusion mechanism, can effectively eliminate the interference of tunnel water mist scattering, and improves the definition, detail performance and color authenticity of the inspection image, thereby providing high-quality visual support for tunnel safety monitoring.

The application provides a distributed optical fiber vibration event detection method and system, relates to the technical field of detection, and comprises the following steps: collecting vibration signals with time domain waveforms and spatial identifiers through a distributed optical fiber, performing time-frequency domain decomposition to extract a characteristic vector, constructing a vibration energy transmission matrix to perform traceability analysis, establishing an inter-regional vibration response causal dependence graph to identify a linkage intrusion event and determine an intrusion source position and a trajectory, and further generating a regional cascade response instruction. The application realizes accurate identification and tracking of cross-regional intrusion behaviors, and improves the sensing capability and linkage response efficiency of a protection system.

This invention discloses a time series completion method, device, and medium based on diffusion sampling correction and frequency domain optimization. The method includes: constructing an observation mask; inputting the complete time series and the corresponding simulated missing mask into a diffusion model, generating an original noise sequence through forward diffusion, and generating a perturbation noise sequence based on step-dependent Gaussian perturbation, using the perturbation noise sequence as model input, and using the original noise sequence as a supervisory signal to train a diffusion noise predictor; in the inference phase, inputting the time series containing partial observations and the corresponding real missing mask into the trained diffusion noise predictor, iteratively generating intermediate samples according to a preset diffusion step size, performing frequency domain decomposition and correction on the intermediate samples during the iteration process, and continuing subsequent iterations based on the corrected signal until all preset diffusion steps are completed; and outputting the completed time series data. This invention achieves the completion of missing data in multivariate time series.

This invention relates to the field of wireless charging technology, and in particular to a smart alignment method and system for wireless charging in a waterproof socket. The method involves acquiring spatial position data of the charging receiver within the charging area of ​​the waterproof socket, establishing a spatial coordinate relationship model to calculate a three-dimensional offset vector, generating a layered alignment strategy, and adjusting the excitation distribution and phase amplitude of the transmitting coil array. During strategy execution, coupled voltage waveform data is collected and frequency-domain decomposed to obtain harmonic components. The influence characteristics of the waterproof medium are identified as compensation factors to correct the offset vector. When the corrected offset vector converges, a locked state is established to create an energy transmission channel. This achieves efficient and precise alignment for wireless charging in a waterproof environment, improving charging efficiency and stability.

This invention belongs to the field of computer vision and digital image processing technology, specifically relating to a semantically aware image enhancement method based on amplitude-phase dual-path collaboration. The invention includes: acquiring a degraded image to be processed; performing frequency domain decomposition on the degraded image to obtain amplitude and phase components; designing a semantically aware Mamba algorithm to enhance the amplitude component, and enhancing the phase component; using bidirectional cross-attention to collaboratively enhance the enhanced amplitude and phase features, obtaining optimized amplitude and phase features; recombining the optimized amplitude and phase features and transforming them to the spatial domain to obtain the enhanced image. This invention captures global dependencies by introducing a state-space model, decouples and enhances the amplitude and phase information of the image separately in the Fourier domain, and utilizes cross-attention to achieve collaborative optimization of the two, thereby achieving effective enhancement for various types of image degradation.

The application discloses a kind of power coordination control methods of wind storage combined power generation system, it is related to wind power generation and energy storage system collaborative control field, including: the present application is by obtaining wind speed prediction data of wind farm and calculating probabilistic wind power prediction, constructs the multi-branch scene tree of power uncertainty, and formulates the energy storage charge-discharge power plan considering multi-scenario probability distribution based on this;After solving the optimal power instruction sequence by rolling optimization, the total power demand is scientifically distributed to the fan and energy storage for execution by using frequency domain decomposition technology;By real-time calculation of the deviation between the actual total power of wind storage system and predicted power, and based on the deviation, the total power demand reference value of the subsequent regulation link is dynamically corrected, effectively correcting the energy accumulation deviation of the total output energy of the system relative to the planned energy in the long time scale, and improving the tracking accuracy of the power grid dispatching plan.

The application discloses a kind of robot cross-domain visual scene identification method and system, method includes: obtaining current domain sample and the feature thereof is frequency domain decomposition, separate low-frequency component and high-frequency component;Two kinds of components are modeled and fused respectively using mapping function, obtain frequency domain regulation feature, mapping is pseudo sample data, with current domain sample common input scene identification model training, constrain the discrimination ability of model to historical domain and current domain scene in training;The model parameters before and after training are gradually fused, and a cross-domain visual scene recognition model is obtained for the identification of the current observation scene. By frequency domain decomposition and regulation, the structural consistency and texture expression of cross-domain style transfer are enhanced, combined with pseudo sample generation and distribution replay training strategy, without relying on historical real data, effectively alleviating catastrophic forgetting, significantly improving the long-term cross-domain scene recognition robustness of robot under the drastic appearance change of season, illumination and the like.

This invention relates to the field of wind speed prediction technology, and discloses a wind speed prediction method and apparatus based on FFT-Rankpooling-LSTM. The method includes: acquiring multi-source time-series wind speed data; generating frequency domain features through normalization and FFT decomposition; selecting core features using a ranking pooling algorithm and constructing a spatiotemporal correlation data matrix; fusing features and reducing dimensionality through a multi-channel attention mechanism to generate a time-series feature encoding vector; building a prediction evolution model based on LSTM network modeling and optimizing hidden state propagation; and finally generating the result through multi-step rolling prediction. The apparatus includes modules such as frequency domain decomposition, feature pooling, spatiotemporal modeling, and a prediction engine, supporting data smoothing, visualization, and adaptive parameter optimization. This invention improves the accuracy and generalization ability of wind speed prediction by integrating frequency domain analysis, hierarchical feature pooling, and spatiotemporal correlation modeling, and is applicable to multi-step wind speed prediction scenarios in fields such as wind power and meteorology.

This invention relates to a hyperspectral image demixing method based on a graph wavelet transform convolutional network, belonging to the field of hyperspectral image demixing technology. The method includes: using a Mexican Hat graph wavelet filter to perform frequency domain decomposition on the undirected graph of the original hyperspectral image, obtaining low-frequency and high-frequency components, which are then input into a parallel dual-branch feature encoder network for feature learning, extraction, and adaptive recalibration. Enhanced low-frequency and high-frequency features are obtained and fused element-wise to obtain multi-frequency comprehensive features, which are then input into an endmember constraint and reconstruction module for abundance estimation. This yields a reconstructed hyperspectral image, which is compared with the original hyperspectral image. A multi-dimensional joint loss function is constructed and optimized to obtain a hyperspectral demixing model for hyperspectral image demixing. This method aims to solve the technical problem of existing technologies' inability to effectively distinguish the mixing characteristics of high-frequency boundary regions and low-frequency homogeneous regions, leading to boundary blurring and spectral distortion.

This invention relates to the field of image processing technology in industrial visual inspection, specifically an image processing algorithm based on texture fusion analysis. The algorithm includes: performing multi-resolution frequency domain decomposition on a surface grayscale image to generate a sub-band image group; generating background suppression frequency domain mask data for a process background reference image; inputting the sub-band image group into a non-local texture similarity calculation engine to output a texture structure consistency metric value, and performing point-by-point logical operations with the background suppression frequency domain mask data to generate a preliminary defect response map; calling a dynamic weight allocation network to perform regional stability evaluation on the preliminary defect response map, and generating a texture fusion weight matrix based on the local gradient distribution characteristics of the surface grayscale image; and performing weighted fusion and threshold segmentation on the preliminary defect response map based on the weight matrix to output the defect region localization result. This invention combines a dynamic weight allocation network to optimize defect response fusion, improving the accuracy of defect localization in industrial visual inspection.

This invention discloses a cervical image classification method and system based on wavelet transform multimodal fusion, comprising: 1) constructing a multi-branch feature encoding network to process cervical OCT images, colposcopy saline images, iodine smear images, and acetic acid whitening images respectively; 2) introducing wavelet transform to perform frequency domain decomposition on the multimodal images, obtaining high-frequency and low-frequency components, and designing a cross-frequency interaction mechanism to more effectively capture the structural and detailed information of the images, and extracting the features of each modality through a ResNet-18 backbone network; 3) designing a feature fusion module to stack and compress the three colposcopy features, and then concatenate them with the OCT features to form fused features; 4) achieving feature dimensionality reduction through a double fully connected layer, adopting a weighted voting multimodal decision classification strategy, and employing a five-fold cross-validation strategy, combined with a weighted cross-entropy loss function to solve the data imbalance problem; 5) finally classifying using Softmax. This invention can achieve accurate classification of cervical images.

This invention provides a multi-sensor-based method for early warning and dredging scheduling of silt in urban road sedimentation wells, relating to the field of urban drainage system management technology. The method includes collecting silt layer thickness, fluid flow characteristics, and well structure response signals from sensors within the sedimentation well; extracting features through time-frequency domain decomposition to identify the siltation state; constructing a risk propagation graph model for dynamic diffusion calculation to generate regional early warning information; and matching dredging process parameters based on siltation characteristics to generate a dredging scheduling plan. This invention enables real-time monitoring and accurate early warning of silt status in sedimentation wells, improving drainage system safety and dredging operation efficiency.

The application provides a medical image cross-modality generation method based on double-contrast frequency domain decomposition and adaptive diffusion, which comprises the following steps: acquiring MRI samples and preprocessing, constructing a k-space decomposition and enhancement process, constructing a k-space double-flow adaptive degradation network, and constructing a k-space double-flow adaptive prediction network. Through data enhancement technology and an adaptive Gaussian filter set, different frequency components are separated and enhanced, common edge features are highlighted, and the quality of generated information is improved. By using a residual noise diffusion model, high-resolution images with rich details and high diversity are generated while maintaining smooth areas and large-scale structures of images. By introducing anatomical structure consistency constraints and multi-modality fusion optimization strategies, the complementary information of T1 and T2 weighted MRI is fully utilized, and the generation effect and visual authenticity of the medical image cross-modality generation in a complex scene are significantly improved.

This invention discloses a method and system for assessing the grid adaptability of offshore wind power connected to the grid via flexible DC in a multi-access-point scenario, specifically relating to the field of power system technology. The method includes: real-time acquisition of electrical quantity data from multiple common connection points; time-frequency domain decomposition of the electrical quantity data; extraction of disturbance feature vectors; inputting the disturbance feature vectors into a grid adaptability assessment model; calculating the node adaptability index and disturbance propagation impact index for each node; and generating adaptability assessment results. The system includes: multiple data acquisition terminals, a digital twin assessment server, and a collaborative optimization controller. This invention enables proactive perception and quantitative assessment of disturbances in the onshore power grid. Through multi-objective collaborative optimization, it adaptively adjusts the operating parameters of each converter station, forming a perception-assessment-decision-control closed loop, significantly improving the grid adaptability of offshore wind power connected to the grid via flexible DC in a multi-access-point scenario and effectively reducing the risk of large-scale grid disconnection.

This invention provides a hybrid vibration reduction method for large-span steel space frames in wind tunnels based on TMD and viscous dampers, belonging to the field of wind tunnel technology. This invention involves installing magnetorheological elastomer variable stiffness devices and temperature-compensated viscous dampers within each tuned mass damper unit. It collects signals using triaxial accelerometers and performs frequency domain decomposition algorithm mode identification, establishing a two-layer game optimization model to output the optimal frequency distribution vector and optimal damping force distribution vector. Based on the optimization results, the stiffness coefficient and damping coefficient are adjusted using a current controller. The vibration reduction effect is evaluated using a sliding time-window power spectral density analysis method, and the optimization model is re-executed to update the control parameters based on the evaluation results. This solves the technical problem of vibration reduction failure in large-span steel space frame structures due to frequency detuning of tuned mass dampers caused by multimodal coupled vibration under wind load excitation.

This invention provides a corneal disease diagnosis system, method, and terminal based on anterior segment photography. The acquired anterior segment photograph is input into a corneal disease diagnosis model based on the fusion of learnable wavelet and Morse topological features to obtain diagnostic results. In the model, a frequency-aware modulation module performs learnable frequency domain decomposition and sub-band modulation on the input image to obtain frequency domain features and reliability scores; a topology-aware module globally models the long-range dependencies of the topological response at different scales and locations of the image to obtain topological embedding features and response scores. A confidence-aware topology and frequency fusion module uses a bidirectional cross-attention mechanism to align, modulate, and perceive consistency across modalities to obtain diagnostic results. This invention introduces Morse theory modeling into the topological structure branch to construct stable corneal morphological representations at multiple scales, and combines this with the Mamba model to efficiently model long-range dependencies; the frequency domain representation branch highlights discriminative frequency components, suppresses interference, and improves the stability of texture features. Furthermore, a fusion mechanism is designed to achieve adaptive correction and complementarity of heterogeneous features, and a reliability voting strategy is introduced to improve the robustness and credibility of the diagnostic results.

The application relates to a bridge structure modal parameter identification method, which is based on a bridge structure micro-vibration video collected by a camera, fuses a variational mode decomposition (VMD), a phase-based video motion magnification (PBVM), a Gabor filter-based phase extraction method (GBP), an improved clustering algorithm (ICA) and an integrated covariance-driven stochastic subspace identification (SSI-Cov) and frequency domain decomposition (FDD) algorithm, and constructs a complete non-contact structure modal parameter identification system. In the absence of prior knowledge of the target frequency bandwidth, environmental noise interference and without obvious characteristic targets, the effective decoupling of multiple vibration modes of the bridge structure, the frequency domain feature extraction of small-amplitude vibration and the intelligent and automatic modal parameter extraction are realized, and the application belongs to the technical field of bridge structure health monitoring.

This invention belongs to the field of time series forecasting technology, specifically a method and system for effective wave height prediction based on dynamic constraints and interactive learning. The method includes: acquiring wave time series data, segmenting it according to the parity of the time index to obtain odd and even sequences; determining the boundary frequency based on the spectral distribution, and performing frequency domain decomposition on the two sub-sequences to obtain swell components and wind wave components; performing evolutionary prediction using a linear Koopman predictor and a nonlinear Koopman predictor respectively; and then performing recursive interactive fusion and time series reorganization based on the prediction results to obtain fused features and outputting an effective wave height prediction sequence for a preset future time period. This invention can improve the accuracy, stability, and generalization ability of medium- and long-term predictions.

The present application relates to the technical field of tunnel blasting parameter design, in particular to a stratum tunnel blasting parameter intelligent design optimization method, the method comprises: based on the design axis and the geological exploration profile layout space synchronous monitoring measuring point, obtaining the surrounding rock wave velocity time sequence change, the fracture density spatial distribution and the ground stress field measured component; the frequency domain decomposition of the surrounding rock wave velocity time sequence change generates the rock mass integrity index atlas, the fracture density and the ground stress field are coupled and analyzed to generate the three-dimensional rock mass anti-explosion sensitivity surface; superimpose two kinds of results to demarcate the candidate section and check, identify the blasting disturbance propagation mode, locate the dangerous area and map to the candidate section, generate the preliminary blasting parameter configuration with risk mark. The present application realizes the accurate adaptation of blasting parameters and the actual characteristics of rock mass, optimizes the selection of blasting test section and the parameter configuration process.

The application relates to an uncertain spatio-temporal knowledge graph query method based on fuzzy semantics and long-short-term dynamic semantics. The method comprises the following steps: embedding each element of an uncertain spatio-temporal knowledge graph to be queried into a feature vector; the elements include a head entity, a relation, a tail entity, time and space, and each element corresponds to a membership degree; inputting each feature vector and the membership degree into a pre-trained query model to obtain an entity query result; the training sample of the query model is a plurality of known uncertain spatio-temporal knowledge graphs, each graph includes elements and membership degrees; the query model comprises a feature enhancement module, a frequency domain decomposition module, a spatio-temporal semantic embedding module and a scoring module; the feature enhancement module generates a fuzzy vector of each element based on the feature vector and the membership degree of each element; the frequency domain decomposition module decouples the long-short-term relation of the dynamic evolution of the relation fuzzy vector; the spatio-temporal semantic embedding module is used for fusing spatio-temporal information and long-short-term relations; and the scoring module is used for generating a predicted entity and calculating the similarity between the predicted entity and a candidate entity. The application can improve the query accuracy of the knowledge graph.

This invention provides a method and system for segmenting remote sensing images of mining areas, relating to the field of remote sensing image processing technology. The method includes: extracting features from remote sensing image data of a target mining area to obtain multi-scale spatial features; performing frequency domain decomposition and multi-scale context aggregation on the multi-scale spatial features to obtain low-frequency sub-band features, high-frequency sub-band features, and global context enhancement features; performing context modeling on the low-frequency sub-band features to generate low-frequency semantic information, and gating the high-frequency sub-band features based on the low-frequency semantic information to obtain frequency domain enhancement features corresponding to the multi-scale spatial features; correcting and fusing the multi-scale spatial features and frequency domain enhancement features to obtain a fused feature map of the multi-scale spatial features and frequency domain enhancement features; and generating a mining area segmentation result for the target mining area based on the fused feature map and the global context enhancement features. This invention improves the segmentation accuracy of remote sensing images of mining areas.

The present application relates to the field of power electronics, in particular to a grid-connected inverter device with adaptive impedance matching. It comprises hardware module, sensing module, prediction module and coordination module; the hardware module comprises a composite branch connected in parallel with the grid bus, the composite branch contains resistance elements, capacitance elements and inductance elements with adjustable parameters; the resistance elements are connected with the adjustable resistance path through three-phase full-bridge switches, the adjustable resistance path includes main path and standby path; the inverter and the grid bus form a grid-connected loop. Through the sensing module to calculate data, the prediction module to determine the reference, the coordination module to control the path and parameters, it not only adapts to the grid impedance in real time, but also accurately treats harmonics by frequency domain decomposition, solves the problem of impedance matching destroying harmonic suppression conditions and weakening the harmonic suppression ability of the inverter, and can also resist interference, reduce loss and improve energy efficiency.

This invention discloses a method and system for tracing the source of harmonics in a distribution network, belonging to the field of distribution network monitoring technology. The method includes: S1, constructing an initial distribution network harmonic causal graph; S2, sensing real-time changes in the distribution network topology and dynamically updating the initial harmonic causal graph to obtain an updated harmonic causal graph; S3, decomposing the updated harmonic causal graph into multiple single-frequency causal subgraphs, performing causal intervention on each subgraph to separate background harmonics from user-side harmonics, calculating the contribution of each user-side harmonic source, and outputting the tracing results. This invention uses a causal graph neural network to replace traditional statistical correlation analysis, thereby effectively distinguishing between causal relationships and spurious correlations in harmonic propagation. By decomposing the harmonic causal graph into multiple independent single-frequency causal subgraphs, accurate decoupling of broadband coupled harmonics is achieved, avoiding error accumulation caused by frequency domain decomposition.