384 results about "Deconvolution" patented technology

The invention provides a deep learning-based oral cavity image recognition method, a storage medium and a deep learning-based oral cavity image recognition system. The method comprises the steps of deploying a federated learning framework and collecting a multi-modal oral cavity image data set; extracting local features to obtain image features, and generating a modal adaptive weight map; a multi-head self-attention mechanism is used for fusing the cross-modal features to generate a fused feature map, and deconvolution up-sampling is carried out to form high-resolution multi-modal feature representation. A tooth segmentation mask is generated based on this representation, and an initial diagnostic report is generated. And aggregating the attention weight of each client through an encryption protocol, and generating interpretable decision support data. And finally, generating a structured clinical report by using a natural language. According to the method, the Grad-CAM thermodynamic diagram is combined with the encrypted and aggregated attention weight, so that the privacy security is guaranteed, the model interpretability is enhanced, the clinical credibility and the diagnosis decision efficiency are improved, and the problems of insufficient diagnosis precision of complex lesions and insufficient utilization of multi-modal information in the prior art are solved.

The invention discloses a living cell super-resolution imaging method and system based on self-supervised learning. The method comprises the following steps: S1, obtaining a fluorescence microscopic image through a standard fluorescence microscopic system; s2, for the acquired single noise image, generating positive and negative sample pairs of self-supervised training data through an autonomously designed and optimized self-supervised strategy; s3, constructing a front denoising neural network and a rear deconvolution network, and performing network training; and S4, inputting a newly obtained noise image into the trained front denoising network to obtain a denoised image, and inputting the denoised image into the rear deconvolution network to obtain a final super-resolution reconstructed image. According to the invention, the standard fluorescence microscopic system and the self-supervised denoising processing module are combined, so that high-quality denoising and super-resolution reconstruction of the fluorescence microscopic image can be realized in a low-photon signal scene.

Methods for determining a location of a feature on a spatial array include (a) providing an array of features on a substrate, where a feature of the array includes a barcoded oligonucleotide having, in a 5′ to 3′ direction, a spatial barcode, a cleavage domain, and a constant sequence; (b) hybridizing a priming oligonucleotide to the constant sequence; (c) extending the priming oligonucleotide using the barcoded oligonucleotide as a template; and (d) determining all or a portion of a sequence of the extended priming oligonucleotide corresponding to the spatial barcode, or a complement thereof, and a location of the extended priming oligonucleotide, and using the location of the extended priming oligonucleotide to determine the location of the feature on the spatial array.

The invention discloses an AI-assisted protein purification result analysis method and system. The method comprises the following steps: S1, collecting chromatogram data, electrophoresis image data and mass spectrum peak table data; s2, performing data preprocessing after the chromatogram data, the electrophoresis image data and the mass spectrum peak table data are obtained; s3, carrying out AI identification analysis on the preprocessed data to realize electrophoretic band identification, chromatographic peak identification and mass spectrum deconvolution; s4, performing multi-modal fusion on the chromatography, gel electrophoresis and mass spectrum information sources; and S5, automatically generating a report, and producing a visual chart. The deep learning and multi-modal data fusion technology is introduced, comprehensive automatic analysis of chromatographic data, electrophoresis images and mass spectrum results is achieved, then automatic result analysis of the protein purification process is achieved, an artificial intelligence algorithm is used for recognizing a peak structure and an electrophoresis band, multi-modal data fusion is achieved, and the detection accuracy is improved. And generating a visual analysis report according to a data result.

The invention relates to the technical field of nondestructive testing, and discloses a wind power casting nondestructive testing method and system, and the method comprises the steps: obtaining three-dimensional surface point cloud data of a to-be-tested wind power casting, employing a sound ray tracking algorithm to calculate a sound path distance and a propagation angle between each imaging point and each array element in an imaging grid, and generating a focusing rule; performing full-matrix data acquisition to obtain an original time domain signal, performing wavelet packet decomposition on the original time domain signal, and determining the peak kurtosis of the original time domain signal; carrying out weighted coherent superposition on any imaging point in the imaging grid based on a focusing rule and an original time domain signal, and reconstructing an image; and identifying a defect indication area in the reconstructed image, calculating an average spatial frequency gradient, determining the number of iterations of a deconvolution algorithm, performing deconvolution processing on the image of the defect indication area, and determining the contour and size of the defect. According to the method, the defect signal can be effectively enhanced, the noise interference is inhibited, and the accuracy of the defect detection result of the wind power casting is improved.

The invention relates to the technical field of image reconstruction, in particular to an intelligent trolley image compressed sensing reconstruction method and system based on deformable convolution. The method comprises the following steps: an intelligent trolley obtains multiple frames of original images in a driving scene through a vehicle-mounted image acquisition module and carries out preprocessing to generate a preprocessed image set; performing feature extraction, feature matching, hypergraph transformation and image synthesis operation on the basis of the preprocessed image set, performing compressed sensing sampling processing at the same time to obtain compressed sensing sampling data, and inputting the compressed sensing sampling data into a preset compressed sensing reconstruction network; a deformable convolution module is embedded in the front end of the compressed sensing reconstruction network for feature extraction and image detail recovery calculation, and a preliminary reconstruction image is generated; and embedding a deformable deconvolution module at the rear end of the compressed sensing reconstruction network to carry out pixel-level adjustment and quality evaluation until a final reconstruction image meeting the reconstruction effect requirement is generated. According to the invention, the precision of vehicle-mounted image compressed sensing reconstruction can be improved.

The invention relates to the technical field of image processing and the technical field of industrial quality detection, and discloses an aluminum material defect detection method based on machine vision, and the method comprises the steps: obtaining and preprocessing an aluminum material defect image data set; an improved YOLOv13 aluminum material defect recognition and detection model is adopted to train a data set, and the model improvement content is as follows: a designed frequency domain deconvolution reconstruction up-sampling module FDRU is used as an up-sampling module in a neck network; a designed multi-scale receptive field convolution module MRF-Conv is used in the backbone network and the neck network; a designed feature extraction module A2C2fMRFA is used in a backbone network; and in combination with an image acquisition device and a computer algorithm, carrying out real-time detection on the surface defects of the aluminum material by utilizing the trained model. Meanwhile, the invention provides a detection system for implementing the method, electronic equipment and a readable storage medium. According to the invention, high-precision real-time detection of the surface defects of the aluminum material can be realized, and support is provided for aluminum material production quality control.

The invention discloses a mechanical fault diagnosis method based on maximum correlation kurtosis depth unrolling, and relates to the technical field of mechanical state monitoring, and the method specifically comprises the following steps: obtaining a vibration signal sliding window segmentation calculation kurtosis, and generating a window priority sequence through cross-correlation correction sorting; a high-quality window is selected to be combined with a rotating speed integral and a fast Fourier transform detection angle period to form an angle domain period boundary set, a segmentation signal calculation error is subjected to quasi-Newton optimization de-convolution to obtain a deep de-convolution fault feature signal, a Toeplitz matrix is constructed, and a multi-layer fault feature vector is weighted and extracted; according to the method, vibration signal sorting is segmented through a sliding window, so that impact positioning is enhanced, noise is reduced, definition is improved, an angle domain period is dynamically recognized, feature synchronization is achieved, error optimization unrolling is performed, impact recovery is enhanced, and feature dimensions are expanded through matrix weighted mapping. And the recognition precision and adaptability are improved through multi-scale aggregation.

The invention discloses a wide-temperature-range solid insulation space charge measurement method, and belongs to the technical field of electrical insulation test and dielectric physical measurement. A solid insulation sample is mounted between double-layer composite electrode modules in a closed high-low temperature test cavity, and a flexible compensation gasket is adopted to absorb thermal expansion difference; a PID closed-loop control system is utilized to realize accurate and stable control of the temperature of the sample in a wide temperature range; applying a direct-current bias voltage and pulse voltage superposition signal, and collecting an acoustic signal generated by space charge displacement through a wide-temperature-range PVDF piezoelectric sensor module; and based on a sound velocity and sensitivity temperature calibration curve established by a pre-experiment, carrying out deconvolution processing on a sound wave signal to correct a temperature dependence error, and finally obtaining accurate space charge density distribution at different temperatures. According to the invention, the precision and reliability of space charge measurement of the solid insulating material in a wide temperature range are effectively improved.

The invention relates to the technical field of biomedical imaging, and discloses a photoacoustic-ultrasonic volume imaging method and system based on an annular ultrasonic array, and the method comprises the steps: reconstructing a two-dimensional photoacoustic image and a two-dimensional ultrasonic image; after photoacoustic imaging and ultrasonic synthetic aperture imaging of one imaging plane are completed, the annular ultrasonic array is moved in the direction perpendicular to the imaging plane, photoacoustic images and ultrasonic images of all the imaging planes are repeatedly executed on the next imaging plane, rebuilt and stacked, and a three-dimensional photoacoustic volume image and a three-dimensional ultrasonic volume image are formed; performing spatial pulse response deconvolution processing on the three-dimensional photoacoustic volume image; and fusing the processed three-dimensional photoacoustic volume image and the three-dimensional ultrasonic volume image to generate a three-dimensional bimodal volume image. According to the invention, high-time-resolution opto-acoustic-ultrasonic bimodal three-dimensional imaging is realized by using an annular ultrasonic array and motion scanning, and the high spatial resolution of the image is further ensured through a series of technologies.

The invention discloses a magnetic source target positioning method, which belongs to the technical field of geophysical exploration, is used for positioning a magnetic source target, and comprises the following steps: collecting and preprocessing plane grid magnetic anomaly data, constructing a simplified Euler deconvolution model under the constraint of derivative features, segmenting observation data by using a local sliding window, and resolving to generate an Euler solution point cloud. The method comprises the following steps: converting a selection problem of an equivalent scale factor into a one-dimensional global optimization problem, adaptively searching an optimal solution through a Bayesian optimization method, identifying a magnetic source high-probability solution point and a noise solution point by using a density-based spatial clustering method, introducing a density penalty term to strengthen a high-probability solution constraint, and constructing a final positioning target function. And outputting the optimal position solution of the magnetic source. According to the method, a traditional Euler deconvolution model is simplified, noise interference is effectively suppressed, the stability and robustness of the inversion process are improved, uncertainty caused by manual parameter selection is avoided, and the positioning precision is improved.

The invention discloses a light field microscopic self-supervised denoising system and method fusing a physical process. The system comprises a light field imaging subsystem and a self-supervised denoising subsystem. The light field imaging subsystem is used for capturing multi-view light field original data of a sample; the self-supervised denoising subsystem is used for processing original data of a light field, preprocessing the original data of the light field in a training stage, and decomposing the original data into multi-view sub-images; the method comprises the following steps: constructing original data of a training set through two different self-supervision strategies, and then generating a self-supervision training data pair through RL deconvolution according to a physical optical process; and finally, a denoising model is obtained through 3D neural network training. In the reasoning stage, light field original data to be reconstructed are preprocessed and then subjected to RL deconvolution to generate input data to be denoised, the input data are input into the trained denoising model, and a reconstruction result is output. According to the invention, the problems of noise amplification, data dependence and poor universality of the light field fluorescence microscope are effectively overcome.

Methods for determining a location of a feature in a spatial array with features include: (a) providing an array with a first set of one or more features immobilized on a substrate, a first feature of the first set having a first barcoded oligonucleotide with a first spatial barcode and a first constant sequence, and a second set of one or more features immobilized on the substrate, a second feature of the second set having a second barcoded oligonucleotide with a second spatial barcode and a second constant sequence; (b) attaching the first constant sequence to the second constant sequence to generate a nucleic acid product; (c) determining all or a portion of a sequence of the nucleic acid product or a complement thereof; and (d) associating the second barcoded oligonucleotide with the first barcoded oligonucleotide in the nucleic acid product.

The application provides a method and system for extracting a road from a remote sensing image and a storage medium, the method comprising: inputting original remote sensing image data to an encoder to extract feature information of the remote sensing image by the encoder; inputting an output of a second residual module to an inflation convolution module to perform inflation convolution processing, and inputting an output of the inflation convolution module to a decoder; wherein the encoder comprises a convolution pooling module, a plurality of residual modules and a down-sampling module; processing the output of the inflation convolution module through an up-sampling module, then through a plurality of up-sampling modules corresponding to the plurality of first residual modules, adding the output of each first residual module to the input of the corresponding up-sampling module in a manner of jump connection, and obtaining a binary graph of the road extracted from the original remote sensing image through a first deconvolution layer and a second convolution layer respectively from the output of the last up-sampling module. The application can more accurately extract road content from the original remote sensing image.

The invention relates to the field of edge calculation, in particular to a sprinkling machine variable irrigation control method based on target water depth cutting and geometric volume-preserving mapping, and the method is universal in two operation modes of a central pivot (circle) and translation. According to the method, a closed-loop process of target cutting-connected partition and discrete gear selection-AMR cell refinement-CMI mapping-cell-level constraint deconvolution-60s period greedy backtracking-execution and monitoring is followed, and cooperative control of 4-5 management regions and a plurality of cells in the whole field can be realized; the community-level and whole-field double-layer (> = 90%) satisfaction degree is preferentially ensured, and the uniformity and water saving are both considered.

The application discloses a near-field strong interference suppression method based on spherical wave deconvolution beam forming positioning, which comprises the following steps: performing fine grid division on a near-field scanning area of a linear array; performing spherical wave focusing beam forming on the near-field scanning area and two-dimensional deconvolution beam forming about distance-angle under the condition of unknown sound source position; calculating the distribution entropy characteristics of the deconvolution beam forming result in the angle dimension, and using a preset threshold to correlate the interference positioning of a target; constructing an array flow pattern matrix A of each near-field target l and generating a covariance matrix A; calculating the orthogonal complement space projection matrix of the covariance matrix A, and applying the orthogonal complement space projection matrix to the array element domain covariance matrix C to perform near-field interference suppression; using a far-field plane wave model to perform deconvolution beam forming calculation on the covariance matrix for suppressing near-field interference, so as to obtain passive detection results of the near-field strong interference suppression. The application can realize near-field interference positioning and interference suppression under the condition that the interference position is unknown.

The present application relates to the technical field of gas sensor calibration, and particularly relates to an intelligent calibration method for a methane sensor in a coal mine, which comprises the following steps: establishing a near-field session and obtaining a permeation tube metrological certificate parameter and a calibration instruction set; obtaining a zero point data set and a sealing score in a closed microcavity, generating a standard concentration trajectory according to the metrological certificate, and collecting a standard section data set; forming original collected data according to joint target optimization of a temperature time profile and an H-infinity norm information matrix; implementing robust deconvolution and variational Bayesian inference based on a first-order inertial dynamic model to obtain a set of robust point estimation parameters, a posterior distribution and a combined uncertainty; calculating a Wasserstein barycenter on a section adjacency graph and solving a one-norm uniform optimization release correction parameter set to realize traceable, robust and section-consistent online correction and re-calibration triggering.

The application relates to the technical field of gear production, and particularly provides a press-fitting verification system and method for gear production, which comprises a dynamic stress field capturing subsystem, a micro deformation field reconstruction subsystem and an integrated resonance verification subsystem. The method comprises the following steps: forming a piezoelectric nanowire network through a nanometer piezoelectric material deposition process, capturing a dynamic stress field in a press-fitting process, and outputting a stress field space-time sequence; receiving the stress field space-time sequence, reconstructing a micro deformation field of the gear through space-time deconvolution processing, realizing space-time deconvolution processing based on wave propagation principle, and outputting a deformation field distribution diagram; receiving the deformation field distribution diagram, synthesizing theoretical resonance characteristics through a structure resonance model, matching the structure resonance model with measured resonance characteristics through a discretized structure dynamics method, obtaining the measured resonance characteristics through an acoustic resonance scanner, and outputting a press-fitting quality verification value. The application improves the controllability of a gear press-fitting process and the reliability of a product.

This invention discloses a visual enhancement method and system for underground monorail cranes based on dynamic modeling, specifically relating to the field of image processing technology. It involves constructing exposure window time-series data by acquiring video frames of the underground monorail crane's operation, corresponding timestamp data, and exposure time parameters; detecting track joint impact features and identifying transient impact intervals on the exposure window time-series data; establishing a coupled dynamic model of vehicle body vibration and crane sway; calculating the time-varying pose response within the exposure window; and constructing a corresponding composite fuzzy kernel model; using the composite fuzzy kernel model to perform deconvolution constraint solving on the operating video frames to generate structural consistency enhancement images; and finally, outputting the visually enhanced image of the underground monorail crane after cross-frame consistency verification, thereby significantly improving the clarity and stability of the underground monorail crane's operating images.

One embodiment is a method of deconvolving overlapping first and second pulses in a photon-counting CT scanning system, the method comprising detecting a first pulse event having a first detected level; detecting a second pulse event having a second detected level; determining an amount of time that elapses between the detected first pulse event and the detected second pulse event; and reconstructing the first pulse and the second pulse using the first and second detected levels, the duration of time between the first and second pulse events, and a known pulse shape.

The application provides a kind of dynamic monitoring method and system of olive leaf polyphenol components based on principal component analysis, it is related to computer technology field, comprising: obtaining original data;According to the original data, the spectrum interval is separated, and the purified polyphenol characteristic spectrum sequence is obtained;According to the purified polyphenol characteristic spectrum sequence, spectral deconvolution processing is carried out, and the polyphenol subcomponent concentration evolution matrix is obtained;According to the polyphenol subcomponent concentration evolution matrix, the principal component is reconstructed, and the reconstruction result is obtained;According to the reconstruction result, track fitting is carried out, by fitting the principal component score track into the kinetic network model containing hydroxytyrosol glycoside hydrolysis and oxidative condensation step, the fitting result is obtained;According to the fitting result, prediction is carried out, the evolution end point of hydroxytyrosol free state content and its derivative relative proportion is predicted, and the dynamic monitoring result is obtained.The application realizes the accurate monitoring and prediction of the dynamic conversion process of specific polyphenol components such as hydroxytyrosol glycoside and oleuropein in olive leaves.

A processor calculates a system function by a first deconvolution operation based on a set of a first detection signal obtained while a calibration sample is disposed in an examination region and a numerical model of a spatial distribution of magnetic particles included in the calibration sample. The processor obtains a spatial distribution of magnetic particles included in an examination sample by a second deconvolution operation based on a set of a second detection signal obtained while the examination sample is disposed in the examination region and the system function.

This invention discloses a method for field-of-view deconvolution image restoration of an imaging system based on wavefront aberration reconstruction, belonging to the field of wavefront aberration measurement and computational imaging technology. The method includes: measuring the wavefront aberration of the optical system under test in an imaging system containing detection markers; setting the measured wavefront aberration as the wavefront aberration of the optical system under test in an imaging system containing pinhole arrays, and simulating and generating point spread functions corresponding to different field-of-view blocks; replacing the detection markers with imaging samples, and acquiring the actual image of the imaging samples through the optical system under test as the image to be restored; dividing the image to be restored into field-of-view partitions, performing deconvolution restoration on each block using the corresponding point spread function, and then stitching the restored block images together to obtain a complete restored image. This invention decouples wavefront aberration measurement from image restoration, achieving effective correction of spatially varying aberrations in a large field of view without increasing hardware complexity, significantly improving the restoration accuracy, efficiency, and consistency of the entire field-of-view image.

The application discloses a kind of based on graph embedding microwave filter fault diagnosis's performance debugging method, first, the admittance parameter feature extraction network of complex convolution coding subnetwork and complex deconvolution decoding subnetwork cascade is built, then combined with the microwave filter structure to be diagnosed and debugged graph convolution inversion network is built, generates admittance parameter training set and scattering parameter training set, finally alternately trains admittance parameter feature extraction network and graph convolution inversion network, carries out fault diagnosis and debugging to microwave filter.The present application is an unsupervised model, which reduces the difficulty of collecting a large number of data sets;And in the process of fault diagnosis and debugging, without any iterative optimization process, a large amount of diagnosis and debugging time can be saved;The model constructed in combination with the physical structure of the filter can be used for fault diagnosis and debugging of microwave filters of any structure, and has higher applicability.

Traditionally, feature extraction and thickness quantification of Pulsed eddy current (PEC) composite signals from multiple concentric pipes in oil and gas wells or similar, have been approached via pre-built frequency domain inversion models that do not utilize in-situ data and are subject to many unknown variables. We apply independent component analysis (ICA) with known applications in other areas including facial feature extraction, in a novel fashion to determine from in-situ data, the original independent PEC signals corresponding to each pipe, from their composite signal, hence allowing the thickness of each pipe to be reliably quantified. Since ICA is unsupervised, no prior modelling with synthetic or out-of-sample data is required, rather in-situ data is utilized. The results are consistent across logging tool manufacturers as prior knowledge of each tool sensor / physics is not required.

The application discloses a nonlinear Fourier reweighting image resolution enhancement method and system without a point spread function. The method comprises the following steps: acquiring an original digital image; performing Fourier transform on the original digital image to separate an amplitude spectrum and a phase spectrum; constructing a nonlinear mapping function with two adjustable parameters, and performing reweighting processing on the amplitude spectrum by using the function to balance high and low frequency energy of the amplitude spectrum and generate a new amplitude spectrum; recombining the new amplitude spectrum and the original phase spectrum; and performing inverse Fourier transform on the combined spectrum to output a resolution enhanced image. The application does not depend on PSF measurement or estimation, only needs one Fourier transform process, and has high calculation efficiency. The core construction is to directly reweight the amplitude in the frequency domain by using a specific nonlinear function, so that the ill-conditioned problem and artifacts of a traditional deconvolution algorithm are avoided. The application can be widely applied to optical fluorescence microscopy, astronomical imaging and clear processing of general digital images.

The invention relates to a non-interference single-frame high-resolution phase recovery method and system based on a weak object transmission equation. The method comprises the following steps that a sample is placed on an objective table, an annular LED lamp panel, the sample, an objective lens, a barrel lens and a camera are sequentially placed in the direction of an optical axis, and light beams emitted by the LED lamp panel are received by a target surface of the camera after being transmitted by the sample, amplified by the objective lens and focused by the barrel lens; the illumination mode of the annular LED lamp panel is set to be a semi-annular asymmetric illumination mode, the annular LED lamp panel is turned on, and the camera shoots a light intensity image; calculating a phase transfer function in a semicircular asymmetric illumination mode according to a weak object light intensity transfer equation under the inclination angle illumination condition; and calculating a sample phase through a deconvolution algorithm according to the light intensity image and the phase transfer function. According to the invention, the data acquisition amount is small, the phase recovery is rapidly and stably realized, the phase information is complete, the resolution is high, and the non-interference rapid high-resolution phase recovery is effectively realized.

This application discloses a deconvolution method for gas chromatography-mass spectrometry (GC-MS) data, relating to the field of GC-MS data analysis technology. This method utilizes a non-negative matrix factorization function to initially decompose the time-mass-charge ratio two-dimensional intensity matrix, obtaining an initial elution curve matrix and an initial mass spectrometry matrix. An exponentially modified Gaussian peak shape modeling function is then used to fit the initial peak shape parameter set. For each component: based on the initial peak shape parameter set, an exponentially modified Gaussian peak shape modeling function is used to generate the original elution curve matrix. Non-negative least squares optimization and non-linear least squares optimization methods are used to iteratively optimize the mass spectrometry matrix, peak shape parameter set, and elution curve matrix. Finally, a comprehensive scoring function is used to select the optimal component set, and the optimal peak shape parameter set, mass spectrometry matrix, and elution curve matrix for the optimal component set are sorted and output. This application enables precise separation of co-eluted components, improving the accuracy and reliability of complex mixture analysis.

The invention discloses a multi-view microscopic image self-supervision joint deconvolution method based on hybrid neural representation, which is a method for realizing multi-view microscopic image joint deconvolution in a self-supervision mode. The method aims at realizing high-efficiency and high-fidelity three-dimensional image reconstruction without any external training label, and completing multi-view information fusion, deconvolution and noise suppression in a unified frame. End-to-end optimization is carried out through constraints such as internal consistency, dependence on external high-quality paired training data sets is eliminated, the continuous function expression itself forms effective image prior, the inherent smooth characteristic of the continuous function expression plays a role in implicit regularization, noise amplification can be inhibited in the iteration process, and the image quality is improved. And the de-noising process and the de-convolution process are integrated. Meanwhile, the single continuous model is jointly constrained by all observation view angles during optimization, and finally a global optimal solution which is physically self-consistent under all view angles is obtained.

Methods for a system including a detector and a separation media are disclosed. The method may include receiving input data from the detector. The input data may include an observed signal of an unknown sample and a signal of at least one known sample. The method may also include generating one or more deconvolution candidate signals based on the observed signal and using one or more deconvolution processes. The method may further include defining a domain of the detector using the one or more deconvolution processes and based on the observed signal, the calibrated separation media, the input data, or a combination thereof. The method may also include modifying the system based on the domain.