57 results about "Orthogonal transformation" patented technology

The invention belongs to the technical field of optical fiber sensing, and relates to an optical time domain reflection monitoring method and device based on forward and reverse signal joint optimization, and the method comprises the steps: obtaining reverse and forward OTDR observation signals of a to-be-measured optical fiber; constructing a reverse OTDR linear signal reconstruction item based on a product of the orthogonal transformation matrix and a to-be-solved first sparse coefficient vector; constructing a forward OTDR linear signal reconstruction item based on the product of the orthogonal transformation matrix and a to-be-solved second sparse coefficient vector; constructing a data fidelity item based on the OTDR observation signal and the linear signal reconstruction item; constructing a joint sparse regular term based on the structural sparsity of the first sparse coefficient vector and the second sparse coefficient vector to be solved; and constructing a spatial constraint regular term based on the continuous smoothness of the reconstructed OTDR linear signal in the space, thereby constructing a joint optimization function and solving the joint optimization function to obtain a first sparse coefficient vector and a second sparse coefficient vector, and further reconstructing the OTDR linear signal to detect the optical fiber to be detected.

The invention discloses a method for calculating secondary ripple current of a single-phase PWM (Pulse Width Modulation) rectifier, which comprises the following steps of: A, obtaining input voltage us and input current is of the rectifier through sampling, and respectively obtaining orthogonal components us alpha and us beta of the voltage and orthogonal components is alpha and is beta of the current through orthogonal transformation or delay of a quarter period; respectively calculating a voltage amplitude Us and a current amplitude Is; the included angle between the input voltage and the current is phi, namely a power factor angle; b, the active power P and the reactive power Q of the input power of the rectifier are calculated, and then the power factor angle phi of the grid-side current and the voltage is calculated; obtaining the angular velocity omega of the input voltage by using a phase-locked loop; c, according to the inductance value L, the voltage amplitude Us, the current amplitude Is, the power factor angle phi and the angular velocity omega of the PWM rectifier, the alternating current side secondary pulsation power Pshc of the rectifier is calculated; and D, calculating the secondary ripple current of the DC side of the PWM rectifier in real time by using the PWM DC side given voltage Vdcref and the AC side secondary pulse power Pshc. The method is different from a method for obtaining the reference value of the secondary ripple current on the DC side of the PWM rectifier by adopting current sampling and filtering and the like, and the reference value of the secondary ripple current is directly calculated by directly utilizing the existing parameters of the PWM rectifier; and the precision of the secondary ripple current reference value is further improved on the basis of eliminating a current sensor and a circuit thereof.

The present application relates to the technical field of feature vector merging in large model inference acceleration, in particular to a feature vector similarity measurement method and device based on transformation domain, by constructing a new type of calculation path of "transformation-measurement", using an orthogonal transformation matrix to project the feature vector to the transformation domain and then calculating the absolute value error sum, effectively extracting the internal structured features, solving the defect of insufficient accuracy of traditional lightweight measurement indicators, designing a multiplierless hardware acceleration architecture supporting the algorithm, by strictly constraining the element values of the orthogonal transformation matrix within the integer set of {-1, 0, 1}, completely equivalent to the matrix multiplication into simple addition and subtraction operation, completely avoiding multiplication, division and square root operation, greatly reducing the chip area and power consumption overhead, in addition, the pareto optimality of model accuracy and hardware efficiency is also realized, and the configuration capability of searching and customizing the transformation matrix for specific models is also provided.

This invention relates to the field of natural language processing technology and discloses a multilingual word vector weighted alignment method based on orthogonal Protodyakonov analysis. The method includes the following steps: S1: Randomly initialize an orthogonal matrix R, and randomly rotate the Chinese word vectors, i.e., X < -XR; substitute the weight matrix A, calculate the weighted orthogonal transformation matrix W according to the provided weighted Protodyakonov analysis, and obtain the weighted aligned Chinese word vector XW2. The Chinese word vectors are downloaded through a given word vector link to obtain Chinese and English word vectors X and Y. A given alignment dictionary L is used to weight sentiment words, with the goal of weighted alignment from X to Y. The method provided by this invention can perform weighted alignment according to the needs of downstream tasks, further improving the performance of downstream tasks, realizing a shared word vector space, and providing a mathematical proof of the weighted orthogonal Protodyakonov analysis.

The present disclosure relates to an image processing device and method capable of reducing an increase in processing delay.A data path of a one-dimensional input signal array of residual data is switched in a forward direction or a reverse direction, and matrix operation of one-dimensional orthogonal transform is performed on the one-dimensional input signal array input via the switched data path. The present disclosure can be applied to, for example, an image processing device, an image coding device, an image decoding device, a transmission device, a reception device, a transmission / reception device, an information processing device, an imaging device, a regeneration device, an electronic apparatus, an image processing method, an information processing method, or the like.

The invention discloses a displacement measurement method and device based on four-way interference signals, a medium and equipment, and relates to the technical field of semiconductor detection, and the method comprises the steps: carrying out the differential processing of four-way interference signal data with a preset phase relation, and obtaining two-way orthogonal signal data; respectively carrying out extreme value detection on the two paths of orthogonal signal data through a self-adaptive updating strategy to obtain extreme value detection results; performing first normalization processing, orthogonal transformation processing, second normalization processing and phase resolving processing on the two paths of orthogonal signal data in sequence based on an extreme value detection result to obtain phase data; calculating the phase difference between any two adjacent elements in the phase data to obtain phase difference data; and calculating a displacement increment corresponding to each phase difference in the phase difference data through an intelligent direction identification algorithm, and accumulating the displacement increments to obtain a displacement measurement result corresponding to the four paths of interference signal data. According to the invention, a displacement measurement result with higher precision and higher stability can be provided.

This invention relates to the field of biometric recognition technology and discloses a visual recognition-based aquaculture classification system and method, including: a feature preprocessing module, an index mapping module, a feature projection module, and a classification determination module. The preprocessing module acquires the visual feature array of the individual to be identified. The mapping module generates a mask matrix representing the distribution of dirt based on the gradient distribution law. The projection module extracts effective feature components by multiplying element by element and performs dimensionality truncation on the orthogonal transformation basis according to the failure bit index in the mask matrix to construct the projection submatrix. The determination module performs isomorphic dimensionality truncation on the reference feature vector according to the failure bit index, calculates the distance metric value in the dimensionality-reduced feature subspace, and determines the target category of the individual. This invention eliminates the dimensionality collapse phenomenon caused by local dirt by transforming the body surface occlusion state into topological constraint parameters, ensuring the scale consistency of feature comparison under different recognition conditions.

The present application relates to a kind of tree species identification method based on spectral orthogonal transformation and multivariate filtering, comprising the following steps: S1, the high spatial resolution remote sensing image of forest area is collected, and pretreatment is carried out;S2, preliminary interpretation is carried out to the image after pretreatment, and tree species sample is obtained in combination with field investigation, tree species sample is marked on the image after pretreatment, and is divided into training sample and verification sample;S3, the spectral band of remote sensing image is orthogonally transformed, and a set of two two orthogonal feature layers are generated;S4, multivariate filtering processing is carried out to the layer after orthogonal transformation, and filter feature is extracted, and hybrid feature set is constructed;S5, based on hybrid feature set, tree species is identified using machine learning classification method, and classification result and precision statistics are output.The present application reduces the correlation between layers by orthogonal transformation of multispectral data, then highlights the difference of tree species by four filters of texture, low-pass, median and enhanced Lee, to promote the identification of tree species.

The present application provides an image decoding apparatus and method, an image encoding apparatus and method, and a storage medium. A decoding unit decodes data corresponding to a first array of quantized coefficients including an N x M array of quantized coefficients corresponding to a first block and data corresponding to a second array of quantized coefficients including an N x M array of quantized coefficients corresponding to a second block. An inverse quantization unit derives a first array of orthogonal transform coefficients from the first array of quantized coefficients using at least a first quantization matrix of elements of the N x M array and derives a second array of orthogonal transform coefficients from the second array of quantized coefficients using at least a second quantization matrix of elements of the N x M array. An inverse orthogonal transform unit inverse orthogonally transforms the first array of orthogonal transform coefficients to generate a first prediction residual of pixels of a P x Q array and inverse orthogonally transforms the second array of orthogonal transform coefficients to generate a second prediction residual of pixels of the N x M array.

The invention belongs to the technical field of chemical engineering, artificial intelligence and model compression, and discloses a uniform quantification method for a large model in the chemical engineering field. On the premise that the expression capability of a model function is not changed, orthogonal equivalent transformation is introduced, combined smoothing processing is carried out on the weight and activation in a linear layer, and activation distribution approximately obeys uniform distribution in a quantization interval by constructing an orthogonal transformation matrix and systematically remodeling the activation distribution, so that the low-bit uniform quantization error is remarkably reduced. And carrying out joint quantization on the smoothed weight and activation by adopting a uniform low-bit uniform quantizer, further introducing a small number of learnable global orthogonal parameters, and carrying out lightweight optimization on module output errors before and after quantization. According to the method, quantitative perception training or large-scale fine tuning is not needed, calculation and storage overhead is low, engineering implementation is simple, inference performance close to full precision can be kept under the 4-bit or 6-bit quantization condition, and video memory occupation and energy consumption in the model deployment process are remarkably reduced.

An image encoding device 1 of the present disclosure includes a neighboring pixel non-reference predictor 11 configured to generate a predicted image by a predetermined neighboring pixel non-reference prediction for each pixel signal of an original image in a block unit, a filter processor 12 configured to perform a low-pass filter process on a prediction signal located at a boundary of a block of the predicted image by using a decoded neighboring signal neighboring to the predicted image under a predetermined control, a prediction residual signal generator 55 configured to generate a prediction residual signal of the block unit by using the predicted image, an orthogonal transformer 14 configured to perform an orthogonal transformation process on the prediction residual signal of the block unit under the predetermined control, and an orthogonal transformation selection controller 25 configured to control the filter processor 12 and the orthogonal transformer 14 and generate a predetermined transformation type identification signal. An image decoding device 5 of the present disclosure performs a decoding process based on a transformation type identification signal at the time of the predetermined neighboring pixel non-reference prediction.

This disclosure provides a method, apparatus, storage medium, and electronic device for quantizing feature data in a model. The method includes: acquiring feature data to be quantized from a target model deployed on a target device, wherein the feature data includes at least two data blocks; performing an inter-block transformation on the feature data using a pre-constructed inter-block orthogonal transformation matrix to obtain inter-block transformed data; performing an intra-block transformation on each data block included in the inter-block transformed data using a pre-constructed intra-block orthogonal transformation matrix to obtain intra-block transformed data; and quantizing the intra-block transformed data to obtain quantized feature data. This disclosure can evenly distribute the energy of the feature data across all blocks, eliminating the influence of high-variance blocks on the quantization sharing index, and redistributing the intra-block data so that the intra-block data is evenly mapped to each codeword interval of the codebook, solving the codebook collapse problem and thus significantly improving quantization accuracy.

Decoding corresponding to a first array of quantized coefficients including an N×M array corresponding to a first block and data corresponding to a second array including an N×M array corresponding to a second block. Deriving a first array of orthogonal transform coefficients from the first array of quantized coefficients by using at least a first quantization matrix of an N×M array of elements, and derives a second array of orthogonal transform coefficients from the second array of by using at least a second quantization matrix of an N×M array of elements. Performing inverse orthogonal transform on the first array of orthogonal transform coefficients to generate a P×Q array of pixels of first prediction residuals, and performs inverse orthogonal transform on the second array of orthogonal transform coefficients to generate an N×M array of pixels of second prediction residuals.

The invention relates to the technical field of adaptive filters, and discloses a robust gradient adaptive lattice filter, which is characterized in that input vectors are constructed by using input signals at current and historical moments, forward and backward prediction errors of each order are calculated through lattice structure recursion, and a reflection coefficient is updated by adopting a normalized gradient descent method; performing linear orthogonal transformation on an input vector based on a transformation matrix formed by the obtained reflection coefficients to generate a backward prediction error vector, and taking the backward prediction error vector as the input of a transverse filter; performing inner product operation in combination with the current adaptive weight vector to obtain filtering output, and comparing the filtering output with an expected signal to generate an estimation error; constructing a nonlinear cost function in a rational fraction form based on the error, further calculating the gradient of the cost function about the weight, introducing a normalization factor of a backward prediction error vector, and performing robust iteration updating on the adaptive weight; and finally, outputting an anti-impulse noise filtering result according to the inner product of the updated weight and the backward prediction error vector.

To suppress an increase in load.SOLUTION: A predetermined context variable is assigned to the first bin of a bin string obtained by binarizing an adaptive orthogonal transformation identifier indicating the mode of adaptive orthogonal transformation in image encoding to perform context encoding. Further, a predetermined context variable is assigned to a first bin of a bin string obtained by binarizing an adaptive orthogonal transform identifier indicating a mode of inverse adaptive orthogonal transform in image decoding, and context decoding is performed. The present disclosure can be applied to, for example, an image processing device, an image encoding device, an image decoding device, an information processing device, an electronic apparatus, an image processing method, an information processing method, or the like.SELECTED DRAWING: Figure 4

This invention discloses a spatial target semantic segmentation method and apparatus based on ISAR echoes, belonging to the field of radar target recognition technology. The method first uses a trainable domain alignment module (DAM) to directly perform a learnable orthogonal transformation on the original ISAR echo without imaging transformation. While maintaining the same amount of information, the echo domain of the ISAR echo is mapped to a feature domain compatible with the mask domain of the semantic segmentation generation mask, obtaining a first feature. This first feature is then fed into a complex domain encoder for scattering feature extraction. The scattering feature is further processed by a complex domain decoder to achieve the recognition of the semantic segmentation mask. This invention differs from the traditional "image first, segment later" process, avoiding potential information loss and time consumption during the imaging stage, and achieving a direct mapping from the original ISAR echo to pixel-level segmentation.

The application belongs to the technical field of optical fiber sensing, and relates to an optical time domain reflection monitoring method and device based on joint optimization of forward and reverse signals. Reverse and forward OTDR observation signals of a to-be-detected optical fiber are acquired. A reverse OTDR linear signal reconstruction term is constructed based on the product of a orthogonal transformation matrix and a first sparse coefficient vector to be solved. A forward OTDR linear signal reconstruction term is constructed based on the product of the orthogonal transformation matrix and a second sparse coefficient vector to be solved. A data fidelity term is constructed based on the OTDR observation signals and the linear signal reconstruction terms. A joint sparse regularization term is constructed based on the structured sparsity of the first sparse coefficient vector and the second sparse coefficient vector to be solved. A spatial constraint regularization term is constructed based on the continuous smoothness of the reconstructed OTDR linear signal in space. Thus, a joint optimization function is constructed and solved to obtain the first sparse coefficient vector and the second sparse coefficient vector, and the OTDR linear signal is reconstructed to detect the to-be-detected optical fiber.

This invention provides a seismic image processing method and apparatus, relating to the field of image processing technology. The method includes: decomposing a seismic image based on its sub-bands and pixel count to obtain multiple image block matrices; performing an orthogonal transformation on each image block matrix to obtain a transformed image block; extracting spectral coefficients from each transformed image block; performing an arc-orthogonal transformation on each set of spectral coefficients to obtain an approximate matrix corresponding to each image block matrix; subtracting each image block matrix from each approximate matrix to obtain a difference image matrix; and decomposing the difference image matrix based on sub-band update values ​​and pixel count update values ​​until all sub-bands and pixels are traversed to obtain all transformed image blocks. The apparatus executes the above method. The method and apparatus provided by this invention can accurately and efficiently perform seismic image processing.

The present application belongs to the field of harmonic analysis, and provides a harmonic disturbance analysis method and system based on harmonic active power distribution, which carries out pretreatment on the obtained three-phase voltage and current data of each node, separates the fundamental wave and harmonic components through orthogonal transformation, calculates the instantaneous harmonic active power of each node under the reference working condition, calculates and counts the average harmonic active power under the steady state, calculates the harmonic active power of each node under the working condition of single or multiple harmonic sources, carries out difference processing on the harmonic active power under the working condition and the harmonic active power under the reference working condition, obtains the node harmonic active power increment distribution of the to-be-tested working condition relative to the reference working condition, and preliminarily determines the harmonic source position; according to the positive and negative directions, amplitude size and change law along the topology path of the harmonic active power of each node, a harmonic active power distribution propagation heat map is drawn, and the dominant harmonic order, the frequency prone to harmonic amplification and the harmonic disturbance sensitive area are identified. The present application realizes accurate analysis of harmonics.

Decoding corresponding to a first array of quantized coefficients including an N×M array corresponding to a first block and data corresponding to a second array including an N×M array corresponding to a second block. Deriving a first array of orthogonal transform coefficients from the first array of quantized coefficients by using at least a first quantization matrix of an N×M array of elements, and derives a second array of orthogonal transform coefficients from the second array of by using at least a second quantization matrix of an N×M array of elements. Performing inverse orthogonal transform on the first array of orthogonal transform coefficients to generate a P×Q array of pixels of first prediction residuals, and performs inverse orthogonal transform on the second array of orthogonal transform coefficients to generate an N×M array of pixels of second prediction residuals.

The present application provides an image decoding apparatus and method, an image encoding apparatus and method, and a storage medium. A decoding unit decodes data corresponding to a first array of quantized coefficients including an N x M array of quantized coefficients corresponding to a first block and data corresponding to a second array of quantized coefficients including an N x M array of quantized coefficients corresponding to a second block. An inverse quantization unit derives a first array of orthogonal transform coefficients from the first array of quantized coefficients using at least a first quantization matrix of elements of the N x M array and derives a second array of orthogonal transform coefficients from the second array of quantized coefficients using at least a second quantization matrix of elements of the N x M array. An inverse orthogonal transform unit inverse orthogonally transforms the first array of orthogonal transform coefficients to generate a first prediction residual of pixels of a P x Q array and inverse orthogonally transforms the second array of orthogonal transform coefficients to generate a second prediction residual of pixels of the N x M array.

In the present invention, an orthogonal transformation means performs an orthogonal transformation of a prediction error for a block of P×Q pixels and thereby generates an N×M number (where N is an integer satisfying N < P and M is an integer satisfying M < Q) of orthogonal transformation coefficients. A quantization means uses at least one quantization matrix with N×M elements to quantize the N×M orthogonal transformation coefficients and generates N×M quantization coefficients.

An image encoding device encodes a block-based target image. The image encoding device comprises, in a transform skip mode in which orthogonal transform processing of the target image is skipped, a motion compensation predictor configured to generate a prediction image corresponding to the target image by performing motion compensation prediction using a plurality of reference images, an evaluator configured to evaluate a degree of similarity between the plurality of reference images on a pixel-by-pixel basis, a subtractor configured to calculate prediction residuals each indicating a difference between the target image and the prediction image on a pixel-by-pixel basis, a rearranger configured to rearrange the prediction residuals based on a result of evaluation by the evaluator and an encoder configured to encode the prediction residuals rearranged by the rearranger.

The application relates to the technical field of adaptive filter, and discloses a robust gradient adaptive lattice filter, which constructs an input vector by using input signals at current and historical moments, calculates forward and backward prediction errors of each order through a lattice structure recursion, and updates reflection coefficients by using a normalized gradient descent method; a transformation matrix formed based on the obtained reflection coefficients is used to perform linear orthogonal transformation on the input vector to generate a backward prediction error vector, and the backward prediction error vector is used as the input of a transverse filter; inner product operation is performed on the current adaptive weight vector to obtain a filtering output, and the filtering output is compared with an expected signal to generate an estimation error; a nonlinear cost function in the form of a rational fraction is constructed based on the error, a gradient of the cost function with respect to the weight is further calculated, a normalization factor of the backward prediction error vector is introduced to robustly and iteratively update the adaptive weight; and finally, the updated weight and the backward prediction error vector are subjected to inner product operation to output a filtering result resistant to impulse noise.

The invention provides a laser point cloud semantic segmentation method and device, electronic equipment and a storage medium, and the method comprises the steps: inputting three-dimensional laser point cloud data into a convolutional neural network model, carrying out the orthogonal transformation of original space coordinates of the three-dimensional laser point cloud data, and determining aligned space coordinates; local features of all point clouds in the three-dimensional laser point cloud data are extracted through a dynamic graph convolutional network layer by using the current space coordinates and the current features of the three-point clouds; carrying out interaction force-based point cloud displacement calculation processing based on the plurality of local features and the current space coordinate, and updating the space coordinate of the point cloud to obtain a displaced point cloud three-dimensional space coordinate; and continuing to repeat local feature extraction and space coordinate updating of the point cloud, and outputting a segmentation result of each point cloud through a multi-layer perceptron after splicing the local features and global features extracted on the basis of each layer, so as to improve the accuracy of semantic segmentation of the point cloud data.

An encoding device 1 includes: a transformer 13 configured to calculate an orthogonal transform coefficient by performing an orthogonal transformation process on a residual image indicating a difference between the input image and a predicted image of the input image; a quantizer 14 configured to generate quantization coefficient by quantizing the orthogonal transform coefficient based on a quantization parameter; an entropy encoder 24 configured to generate encoded data by encoding the quantization coefficient; an image decoder 10 configured to restore an orthogonal transform coefficient from the quantization coefficient based on the quantization parameter and generate a pre-filtering image by adding the predicted image to a residual image restored by performing inverse orthogonal transformation on the orthogonal transform coefficient; and a deblocking filter 18 configured to perform a filtering process on the pre-filtering image and control a filtering strength depending on a result of comparison between a luminance signal level of the pre-filtering image and a luminance threshold value.

Decoding corresponding to a first array of quantized coefficients including an N×M array corresponding to a first block and data corresponding to a second array including an N×M array corresponding to a second block. Deriving a first array of orthogonal transform coefficients from the first array of quantized coefficients by using at least a first quantization matrix of an N×M array of elements, and derives a second array of orthogonal transform coefficients from the second array of by using at least a second quantization matrix of an N×M array of elements. Performing inverse orthogonal transform on the first array of orthogonal transform coefficients to generate a P×Q array of pixels of first prediction residuals, and performs inverse orthogonal transform on the second array of orthogonal transform coefficients to generate an N×M array of pixels of second prediction residuals.

The application discloses a kind of graphic processor, key value cache read-write method, equipment and storage medium, graphic processor includes: instruction analysis unit, vector processing unit, key value cache unit, data management unit and tensor core unit.Instruction analysis unit determines instruction parameter by parsing instruction.Data management unit retrieves target vector compression data corresponding to target vector.Vector processing unit carries out normalization processing or carries out anti-quantization.Tensor core unit carries out orthogonal transformation or orthogonal inverse transformation.Vector processing unit also carries out quantization processing and denormalization.Key value cache unit stores vector compression data.In the graphic processor provided in the application, a matrix multiplication calculation array does not need to be built in again, can be cooperated with tensor core unit to complete the efficient read-write of key value vector, effectively reduce the on-chip cache capacity and on-chip interconnection bandwidth pressure, and effectively reduce hardware area occupation and functional redundancy, improve key value cache read-write efficiency and resource utilization efficiency.

An encoder including circuitry and memory coupled to the circuitry. In a second type of residual coding among a first type of residual coding where an orthogonal transform is applied to a current block and the second type of residual coding where the orthogonal transform is skipped for the current block, wherein a first syntax used for the first type of residual coding is different from a second syntax used for the second type of residual coding, the circuitry: in a first loop process, derives a context index by using at least one of a plurality of surrounding coefficients; and encodes a plurality of coefficient information flags by CABAC with the derived context index; and in a second loop process of the plurality of loop processes, encodes a plurality of absolute value flags by CABAC with another context index.