Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
Hiro

51 results about "Matrix iteration" patented technology

Method for constructing quasi-cyclic low-density parity-check code (QC-LDPC code)

The invention provides a method for constructing a quasi-cyclic low-density parity-check code (QC-LDPC code), which comprises the following steps: determining a parameter of an encoding matrix; constructing an exponential matrix M (H) of a QC-LDPC code parity check matrix; using a progressive edge growth (PEG) algorithm of the QC-LDPC code parity check matrix to initialize the exponential matrix M (H); endowing the exponential matrix each nonnegative element to form a cyclic shift value with the longest minimum ring and the least ring number; carrying out iteration replacement on the elements in the exponential matrix; and completing the iteration replacement process of the exponential matrix by an exponential matrix iteration assignment algorithm of the QC-LDPC code parity check matrix. The method for constructing the QC-LDPC code of the invention can guarantee the girth and the average minimum girth of the code maximized so as to obtain excellent error code correcting performance.
Owner:ACAD OF BROADCASTING SCI SARFT

Image encryption method based on chaotic mapping and bit-level replacement

The invention provides an image encryption method based on chaotic mapping and bit-level replacement. The method comprises the following steps: converting a grayscale image into an image matrix; iterating 2D-LSCM mapping to generate a chaotic sequence and mapping the chaotic sequence into a matrix, and returning a position index to perform pixel position exchange on the image matrix after descending sorting of each row; sequentially exchanging the lowest three bits in every two pixel bits to obtain an information fusion sequence of the image; performing iteration on the Logistic mapping to generate a chaotic sequence, and performing operation on the information fusion sequence and the chaotic sequence to obtain a pixel replacement column of the image; and performing pixel diffusion on thepixel replacement sequence of the image by using a chaotic sequence and a ciphertext diffusion technology to obtain an encrypted image. The chaotic mapping and scrambled images are used for replacement operation, the differential attack resistance is enhanced through bit information fusion operation and an image pixel-based diffusion method, the secret key space is large enough, violent attack canbe better resisted, and good safety is achieved.
Owner:ZHENGZHOU UNIVERSITY OF LIGHT INDUSTRY

Load characteristics comprehensive classification method based on Markov Monte Carlo

The invention discloses a load characteristics comprehensive classification method based on Markov Monte Carlo. The method includes the following steps: finding the voltage drop time point, carrying out load dynamic characteristics extraction and classification at the disturbance moment corresponding to the voltage drop time point; judging whether the change between the load classifications has a Markov property or not; dividing all data into uniform segments by time; establishing Markov chain's probability transfer matrix based on the maximum likelihood thought for each data segment; judging whether the numerical characteristics are changed or not: if no, go to step V; if yes, carrying out clustering on the load data in the time segment according to the numerical characteristics corresponding to the matrix, and obtaining the probability transfer matrix of the load data with changed numerical characteristics in each time segment ; carrying out Markov Monte Carlo simulation and describing the load change situation; processing the sequence reflecting the load classification conversion using the Hidden Markov Model (HMM). The method provided by the invention improves the Markov chain Monte Carlo simulation and effectively reduces the possibility of the matrix entering the stable state after iteration.
Owner:SHANDONG UNIV

Emotional orientation analytical method of cross-domain texts

The invention provides an emotional orientation analytical method of cross-domain texts, which comprises the following steps: 1) determining initial emotional points of texts and words in a source domain and a target domain; 2) respectively establishing one or more of a similarity matrix M of a test text set DU and a test word set WU and a training word set WL, a similarity matrix N of the test word set WU and the test text set DU and a training word set DL and a similarity matrix V of the test word set WU and the all word sets with relations of all word sets and all text sets according to the test text set and the test word set, and calculating a normalization matrix and a neighborhood matrix of one or more of similarity matrixes; and 3) calculating the emotional points of the test textsand the test words iteratively with the neighborhood matrix of one or more of similarity matrixes by utilizing the initial emotional points of the texts and the words in the source domain and the target domain and performing normalization. The method can improve the accuracy of the classification of the texts in a new domain substantially.
Owner:INST OF COMPUTING TECH CHINESE ACAD OF SCI

IQ calibration and compensation method and device

Disclosed are an IQ calibration compensation method and apparatus. In an IQ signal output by an IQ calibration compensation module, an accumulated value of an absolute value of an I path and an accumulated value of an absolute value of a Q path are filtered separately to obtain a magnitude gain of the Q path; an accumulated value of a product of the I path and the Q path and an accumulated value of a square of the I path are filtered separately to obtain a phase estimation of the Q path and a magnitude gain of the I path; and the IQ calibration compensation module performs IQ calibration compensation that does not comprise a matrix iteration process based on the magnitude gain and the phase estimation of the Q path and the magnitude gain of the I path, so as to output an IQ signal that has been compensated for. The present invention performs the filtering processing in the process of generating a compensation parameter, so that the present invention no longer needs to perform a matrix iteration process in an IQ calibration compensation method in the prior art, influence of non-uniform distribution of sampled data is reduced, a data statistical amount is increased, and the accuracy and stability of IQ calibration are ensured.
Owner:SANECHIPS TECH CO LTD

Multi-regional fully distributed active distribution network economic dispatching method

The invention relates to a multi-regional fully distributed active distribution network economic dispatching method and belongs to the field of electric power system operation and control technology, and the method comprises the steps: 1) establishing an active distribution network economic dispatching model for considering linearizing transmission losses and minimum light discarding cost; 2) realizing localization quasi-Newton matrix iteration initial value construction, localization quasi-Newton direction iterative computation and localization step size searching calculation for regions through neighbor boundary node communication; 3) judging whether global convergence is achieved through the neighbor communication after regional convergence. According to the invention, the method aims at operating characteristics of an active distribution network, point-to-point communication technology are fully utilized, on the basis of fully distributed quasi-Newton method, the method complete active distribution network economic dispatching multi-regional fully distributed superlinear convergence computation without a coordination center, the reliability is improved, privacy of the regions is protected, and plug-and-play is realized.
Owner:TSINGHUA BERKELEY SHENZHEN INST

Distributed computing system and method and storage medium

The invention discloses a distributed computing system and method and a storage medium. The distributed computing system includes at least two computing nodes and at least two parameter service nodes.A computing node initializes vectors of corresponding users in a user matrix according to the users, which are included by a subset of training data, to obtain a user sub-matrix composed of the initialized vectors; the computing node carries out iterative computing on the user sub-matrix and an item sub-matrix according to the subset of the training data and the item sub-matrix obtained from a parameter service node, and transmits an item sub-matrix, which is obtained after each time of iterative computing, to the corresponding parameter service node; the parameter service node initializes vectors, which correspond to partial items, to obtain the item sub-matrix composed of the initialized vectors; and the parameter service node updates an item sub-matrix, which is stored by the parameterservice node, according to the item sub-matrix transmitted by the computing node. By implementing the system, computing tasks can be completed in an intensive manner.
Owner:TENCENT TECH (SHENZHEN) CO LTD

Fractal cloud model cellular automation based art pattern generation method

A fractal cloud model cellular automation based art pattern generation method is used for generating art patterns with existing integral fractal effect and detail uncertainty. According to the technical scheme, the method includes performing fractal matrix iteration for q times to an input picture subjected to initialization processing, and performing cloud model cellular automation evolution for p times before fractal matrix iteration each time to generate a large number of art patterns with existing integral fractal effect and detail uncertainty. Cloud model cellular automation is combined with a fractal matrix method, and fuzziness and randomness are taken into consideration in a pattern generation process, so that the generated art patterns are integrally fractals and have differences in detail to give a new visual perception to people, and personality needs of people on beauty can be met by the detail differences.
Owner:NORTH CHINA ELECTRIC POWER UNIV (BAODING)

Inverse matrix iteration deconvolution method for spectrum resolution enhancement

ActiveCN110162740ATo achieve the purpose of peak width narrowingRealize deconvolution calculation effectImage enhancementGeometric image transformationNMR - Nuclear magnetic resonanceMatrix iteration
The invention discloses an inverse matrix iteration deconvolution method for spectrum resolution enhancement. The method comprises the following steps of 1, sequence convolution and a convolution square matrix; 2, convolution square matrix cumulative multiplication and convolution kernel function peak broadening; and step 3, peak resolution enhancement. An original function is multiplied by a deconvolution matrix to achieve the goal of narrowing the peak width; furthermore, a deconvolution unit matrix construction method is provided, the deconvolution calculation effect with expected precisionis realized; the method has the advantages that the calculation process is fast and controllable, the result is stable and accurate, the application range is wide, the method can be used for Raman, infrared and other molecular spectrums and can also be used for resolution enhancement of other spectrums with symmetrical peak types such as mass spectrums, nuclear magnetic resonance spectrums, XRD and XRF, and the Raman spectrums are taken as examples, so that implementation steps and implementation effects are given.
Owner:GUANGXI UNIVERSITY OF TECHNOLOGY

Target tracking method under DoS attack

InactiveCN111107070AHigh precisionGuaranteed Computational ComplexityTransmissionComputation complexityOptimal weight
The invention discloses a target tracking method under DoS attack. The method comprises the following steps: establishing a state model and a measurement model of a system, and initializing a system state, sampling time and control parameters; modeling a signal received by the monitoring center under the DoS attack; defining a local state estimate (LSE) and a distributed fusion estimate (DFE); designing two optimization problems for the networked nonlinear estimation system, and respectively calculating to obtain a local estimation gain and an optimal weighting matrix; iterative updating is carried out to obtain the optimal estimation of the target position; the distributed fusion estimation algorithm is provided for a networked nonlinear system under DoS attack, and compared with an existing target tracking method, the distributed fusion estimation algorithm effectively improves the target tracking precision on the premise of ensuring the calculation complexity.
Owner:ZHEJIANG UNIV OF TECH

Translation matrix model selection system based on OpenKiWi

The invention provides a translation matrix model selection system based on OpenKiWi, a method for evaluating a plurality of translation results output by a plurality of translation engines based on an evaluation engine of an OpenKiWi core, and a translation terminal. The selection system comprises a to-be-translated source text input component, a GMM clustering engine, an evaluation engine basedon an OpenKiWi core and a selection output interface. The evaluation engine based on the OpenKiWi core performs sentence-level evaluation and word-level evaluation on a translation result, obtains a sentence-level evaluation matrix and a word-level evaluation matrix, combines the sentence-level evaluation matrix and the word-level evaluation matrix to obtain an N-order evaluation matrix PN, performs iterative computation on the N-order evaluation matrix PN, and outputs an optimal translation result based on an iterative computation result.
Owner:IOL WUHAN INFORMATION TECH CO LTD

Three-dimensional thermal imaging method based on OpenCV and OpenGL

InactiveCN101930625ASolve the problem that the 3D infrared heat map of the target cannot be obtainedRadiation pyrometry3D-image renderingGraphicsData file
The invention provides a three-dimensional (3D) thermal imaging method based on OpenCV (open source computer vision library) and OpenGL (open graphics library), which is characterized by comprising the following steps: 1) in computer software programming, reading camera data by virtue of a CvCam module in an OpenCV, and converting the camera data into a spectral intensity digital image data file; 2) by means of a CxCore module in the OpenCV and a CvMat class matrix, realizing spectral chromatography matrix iterative reconstruction operation to obtain a 3D emission coefficient and temperature distribution; and 3) calling coordinate points of a target space and temperature data reconstructed based on the OpenCV by the OpenGL to realize the final 3D thermal image display.
Owner:NANCHANG HANGKONG UNIVERSITY

Co-prime array underdetermined direction finding method based on zero denoising technology

The invention provides a co-prime array underdetermined direction finding method based on a zero denoising technology. The method specifically comprises the following steps of: arranging a co-prime array and receiving array data; preprocessing the data; constructing a uniform virtual array data vector and a model noise covariance matrix; iteratively solving a zeroing coefficient; solving the rootof the equation and measuring the direction. Compared with the prior art, the method has the advantages that firstly, estimation of array noise variance is introduced in the underdetermined solving process, interpolation of holes in a co-prime array virtual array is achieved through adoption of the zeroing denoising technology, and all array element information and the array freedom degree are fully utilized; secondly, modeling is carried out on model noise caused by the limited snapshot number, and the direction finding precision can be improved; finally, the method provided by the inventionbelongs to a meshless algorithm, does not have a grid effect, and does not need to carry out airspace grid scanning.
Owner:HANGZHOU DIANZI UNIV

Method for computing feature kernels for optical model simulation

A method for computing feature kernels for optical model simulation is provided. In the method, a feature matrix mathematically describing a plurality of properties of an optical imaging system is identified. A sampling matrix comprising at least one vector serving as input to form a low-rank basis for the feature matrix is generated. The sampling matrix is iteratively multiplied by the feature matrix and a multiplication result is adaptively rescaled according to numerical stability until a convergence condition is met. The iteration results are used to form a reduced feature matrix. Decomposition values of the reduced feature matrix are computed and a plurality of feature kernels are extracted from the computed decomposition values.
Owner:TAIWAN SEMICON MFG CO LTD

Multi-view image clustering method based on clustering adaptive canonical correlation analysis

The invention discloses a multi-view image clustering method based on clustering self-adaptive canonical correlation analysis, and the method mainly constructs a self-adaptive optimization model of multi-view canonical correlation analysis and clustering and solves the problem of clustering inadaptability of multi-view correlation learning in a clustering task, thereby improving the clustering performance of multi-view images. The method comprises the following steps: (1) initializing a class label indication matrix of an original high-dimensional sample; (2) iteratively solving a related projection matrix, a class centroid matrix and a class label indication matrix; (3) directly obtaining a clustering result based on the solved class label indication matrix. Compared with the prior art, the multi-view image clustering method provided by the invention is more effective and robust.
Owner:ANHUI UNIV OF SCI & TECH

Image recognition method based on adaptive matrix iterative extreme learning machine

The invention discloses an image recognition method based on an adaptive matrix iterative extreme learning machine. The method comprises the following steps: combining a network structure of a traditional extreme learning machine with a self-adaptive matrix iteration mode, and performing convergence by using a convergence factor self-adaptive matrix iteration mode to obtain an output weight matrix when solving the output weight matrix; therefore, while the characteristics of simple network structure, random parameter generation and the like of a traditional extreme learning machine are kept; the method has better training precision in the aspect of image recognition, occupies less computing resources and consumes less time. A new thought and a new approach are provided for improvement and optimization of a machine learning algorithm and image recognition.
Owner:BEIJING INSTITUTE OF TECHNOLOGYGY

Information system security situation assessment method based on fuzzy Petri network

The invention discloses an information system security situation assessment method based on a fuzzy Petri network. The method comprises the following steps: constructing an information system securitysituation evaluation index system; establishing a fuzzy Petri network model; obtaining an expert evaluation result; calculating the credibility of a library; obtaining an initial state matrix; iteratively calculating a state matrix; and obtaining the security level of the qualitative information system. The information system security situation evaluation method based on the fuzzy Petri network has the beneficial effects that the method has a visualized graphic structure; uses a Pythagorean fuzzy set for describing the evaluation indexes so as to achieve a wider expressible range and better description capacity for the evaluation indexes compared with a traditional fuzzy set and an intuitionistic fuzzy set; and has reasonability, can avoid repeated triggering of the same transition by setting judgment on transition equivalent input, and thus guarantees that a security situation evaluation result is obtained in limited iteration times.
Owner:CIVIL AVIATION UNIV OF CHINA

Sparse underwater acoustic channel estimation method based on generalized approximate message passing-sparse Bayesian learning

The invention provides a sparse underwater acoustic channel estimation method based on generalized approximate message passing sparse Bayesian learning, and belongs to the field of underwater acousticsignal processing. The invention relates to a channel estimation method combining generalized approximate message passing (GAMP) and sparse Bayesian learning (SBL). The method comprises the followingsteps: (1) inputting a frequency domain baseband receiving signal, a dictionary matrix, an iteration termination condition and a related parameter initial value; (2) calculating a channel impulse response by utilizing GAMP under the framework of the SBL; (3) updating a noise variance and a channel hyper-parameter; and (4) judging an iteration termination condition, and outputting a channel estimation result if the iteration termination condition is met. The method has the advantages that the underwater acoustic channel impulse response is calculated in an approximate message passing mode under the framework of the SBL, the calculation complexity of the SBL is reduced under the condition of basically no performance loss, and the operation time of the algorithm is shortened.
Owner:CNOOC TIANJIN BRANCH +1

A path planning method, device and control system for an ultra-redundant manipulator

The invention discloses an ultra-redundant mechanical arm path planning method and device and a control system. The method comprises the following steps: after arrangement of a mechanical arm, arm shape description parameters are obtained, and an anticipant path with multiple anticipant nodes is planned; with the first anticipant node as a target anticipant node, parameters of the target anticipant node are obtained; terminal present parameters and Jacobi matrixes of equivalent arm shapes are obtained through the arm shape description parameters; a terminal anticipant speed is obtained throughcombination of the terminal present parameters and the target anticipant node; anticipant arm shape parameters of each arm section are obtained based on iteration of the Jacobi matrixes and the terminal anticipant speed; joint driving parameters are obtained according to the obtained anticipant arm shape parameters; the terminal of the mechanical arm is driven to the target anticipant node according to the joint driving parameters; and the path planning of the mechanical arm is finished with each anticipant node traversed as the target anticipant node. The efficient ultra-redundant mechanicalarm path planning method solves the technical problem of complex calculation in an ultra-redundant mechanical arm path planning method in the prior art.
Owner:HARBIN INST OF TECH SHENZHEN GRADUATE SCHOOL

Sliding window type multivariate time sequence missing value filling method based on 5G network

The invention provides a sliding window type multivariate time sequence missing value filling method and device based on a 5G network, and the method comprises the steps: collecting the sensor data of all measurement points of equipment; combining the sensor data of each measuring point; according to a preset window value, for the time series data segment with the fixed window size, filling a missing value by using a matrix iterative optimization mode based on the reconstruction error; and performing missing value filling on the whole time series data by using a sliding window according to a preset stride value. The problems of low network bandwidth and time delay are solved, the relation between data of all measuring points of historical operation of equipment is fully utilized by using an alternating matrix algorithm and a sliding window method in proportion, and compared with mean value filling, maximum value filling, mode filling and model prediction filling, the data missing value filling accuracy is higher. Meanwhile, the method provided by the invention is simple to implement, high in portability and suitable for most equipment.
Owner:朗坤智慧科技股份有限公司

Optical layer decomposition method based on spatial transformation model

The invention provides an optical layer decomposition method based on a spatial transformation model, aiming to solve the technical problems of bad decomposition effect and low efficiency in a conventional non-supervision single mixed image decomposition method. The realization steps include reading in an original mixed image; fuzzing by means of a blind deconvolution algorithm; defining iterative variables and a transmission layer matrix iteration formula; creating a spatial transformation model and deriving the function expression of the transmission layer matrix and the reflection layer matrix; constructing target functions, adding constraint conditions by means of layer prior, performing regularization with L1 norms, and realizing the target function convex optimization; solving the optimal solution of convex target function by means of a Lagrangian multiplier method; calculating the transmission layer matrix and the reflection layer matrix by means of the function expression; and determining whether the iteration frequency is greater than the set threshold, if yes, the iteration is completed, otherwise, the convex target function is updated. The optical layer decomposition method improves the decomposition effect of the mixed image and improves the calculating efficiency.
Owner:XIDIAN UNIV

Impulse noise suppression underwater acoustic channel estimation method

The invention provides an impulse noise suppression underwater acoustic channel estimation method, and belongs to the field of underwater acoustic signal processing. The invention relates to a method for suppressing impulse noise and realizing underwater acoustic channel estimation by using GAMP-SBL. The method comprises the following steps: (1) inputting a baseband receiving signal, a dictionary matrix, an iteration termination condition and related parameter initial values; (2) estimating impulse noise by using a GAMP-SBL (Generalized Amplitude Modulation Spatial Block Library); (3) subtracting the impulse noise estimation result from the baseband receiving signal; and (4) performing underwater acoustic channel estimation by using the GAMP-SBL. The method has the advantages that the sparsity of the impulse noise in the time domain is utilized, the impulse noise is estimated and subtracted from the baseband signal to suppress the impulse noise, damage to the signal structure can be reduced, and the impulse noise suppression performance is improved.
Owner:哈尔滨工程大学青岛船舶科技有限公司 +1

Method and device for acquiring ion concentration in concrete structure based on finite element method

The invention discloses a method and a device for acquiring ion concentration in a concrete structure based on a finite element method. The method comprises the following steps: acquiring a partial differential equation of a diffusion model of any ion in the concrete structure; obtaining a weak form of the partial differential equation based on the partial differential equation; carrying out unitdivision on a solving space domain, and obtaining a discrete unit finite element control equation based on the weak form of the partial differential equation; discretizing the solving time domain, andobtaining a matrix iterative solving form of the discrete unit finite element control equation based on the discrete unit finite element control equation; and obtaining the concentration of the ionsin the concrete structure based on time step iterative computation and the matrix iterative solution form. Parameter units such as eight nodes of a quadrangle are adopted to divide and solve a spatialdomain, the structure of the bent boundary can be well simulated, and the beneficial effect of accurately obtaining the ion concentration in the concrete structure containing the bent boundary shapeis achieved.
Owner:NINGBO UNIVERSITY OF TECHNOLOGY

Online measurement method for generalized short-circuit ratio of multi-infeed new energy grid-connected system

ActiveCN111239511AResistance to noise interferenceRealize on-line measurement of generalized short circuit ratioResistance/reactance/impedenceElectrical testingAlgorithmNew energy
The invention discloses an online measurement method for a generalized short-circuit ratio of a multi-infeed new energy grid-connected system. System nodes are divided into a feed-in port node, an intra-network node and an infinite power grid equivalent node, discrete time point data of current and voltage of the feed-in port node are acquired from the nodes, and then a current difference matrix and a voltage difference matrix are acquired; and a relation model between the current difference matrix and the voltage difference matrix is established through a rearranged Crocco order reduction admittance matrix, then a rearrangement matrix is solved according to a least square method matrix iterative solution, and the generalized short-circuit ratio is obtained by utilizing the rearrangement matrix. According to the generalized short-circuit ratio real-time measurement algorithm, real-time online measurement of the generalized short-circuit ratio can be achieved, and the calculation resulthas high resistance to measurement noise interference through the matrix iterative algorithm based on the least square method.
Owner:ZHEJIANG UNIV +1

A Noise Reduction Method for Fetal Heart Sound Signal Based on Non-negative Matrix Decomposition

The invention discloses a method for reducing noise of fetal heart sound signal based on non-negative matrix decomposition. The existing filtering methods have limitations in improving the signal-to-noise ratio of fetal heart sound signals. The invention performs short-time Fourier transform on the signal to obtain the amplitude information matrix and phase angle information matrix; performs singular value decomposition and non-negative matrix decomposition on the amplitude information matrix, selects useful features according to the characteristic matrix and coefficient matrix, and combines them to obtain The amplitude information matrix after noise reduction; phase compensation is performed on the phase angle information matrix to obtain the compensated phase matrix; short-time Fourier inverse transform is performed on the amplitude information matrix and the compensated phase matrix to obtain the time domain signal. The singular value decomposition method adopted in the present invention avoids the influence of the randomness of the initial value on the matrix iteration result, and the adopted phase compensation method replaces the original band noise signal phase with the compensated phase, which increases the reliability of the phase signal after noise reduction. Reliability, significantly improved the noise reduction effect of fetal heart sound signal.
Owner:CHINA JILIANG UNIV

Flow transmission system of double-wall cooling structure

The invention discloses a double-wall cooling structure flow transmission system, relates to the technical field of double-wall cooling structure flow design, and solves the problems that in the prior art, designers need to have rich experience, human intervention is needed, an automatic optimization system separated from people cannot be constructed, and the like. The system comprises a flow automatic identification module, a flow priority matrix D calculation module and a flow transmission matrix iteration correction module. According to the method, the flow distributed to each air film hole by each impact hole can be calculated under the condition that the flow of the impact hole inlet and the flow of the air film hole outlet of the double-wall cooling structure are known. The flow path of the double-wall cooling structure is not restrained by the structure, but the fluid is automatically distributed under the condition that the inlet and outlet pressure is given, so that the flow distribution of the double-wall cooling structure cannot be calculated by a common fluid network method. The flow matrix clearly points out the directions of all fluids in the double-wall cooling structure, and the result can be used for effectively guiding the design of the cooling structure.
Owner:BEIHANG UNIV

Method and device for obtaining ion concentration in concrete structure based on finite element method

The invention discloses a method and device for obtaining ion concentration in a concrete structure based on a finite element method. The method includes: obtaining a partial differential equation of a diffusion model of any ion in a concrete structure; obtaining the ion concentration based on the partial differential equation. The weak form of the partial differential equation; the solution space domain is divided into units, and the discrete element finite element control equation is obtained based on the weak form of the partial differential equation; the solution time domain is discretized, and based on the discrete unit The finite element control equation obtains the matrix iterative solution form of the discrete element finite element control equation; based on the time step iterative calculation and the matrix iterative solution form, the ion concentration in the concrete structure is obtained. The invention adopts quadrilateral eight-node and other parameter units to divide and solve the space domain, can better simulate the structure of the curved boundary, and realizes the beneficial effect of accurately obtaining the ion concentration in the concrete structure with the shape of the curved boundary.
Owner:NINGBO UNIVERSITY OF TECHNOLOGY

Method, device and system for accurately locating characters in an image

The invention specifically relates to a method, device and system for accurately locating characters in an image. The method provided by the present invention includes: step S1, initializing the affine transformation matrix and character area distribution probability; step S2, performing histogram statistics on the pixel values ​​of the image according to the character area distribution probability; step S3, according to the histogram statistical results and The character region distribution probability iteratively updates the affine transformation matrix; step S4, updates the character region distribution probability according to the updated affine transformation matrix; step S5, repeats steps S2-step S4, until the first termination condition is met; step S6, The position of the character in the image is obtained according to the finally determined distribution probability of the character area. The method, device and system for accurately locating characters in an image provided by the present invention can achieve accurate locating results only by using histograms and regional locating matrix iterations similar to the optical flow method. The processing speed is fast and the locating effect is good, which can be directly applied to Low-cost module-side identification.
Owner:杭州泽火科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products