64 results about "Noise correction" patented technology

The invention discloses an online self-calibration filtering method for inertial navigation, and relates to the technical field of inertial navigation, and the method comprises the following steps: obtaining multi-modal data, and establishing an inertial device error mathematical model and an inertial navigation system error mathematical model; carrying out recursive estimation on the state variables by the extended Kalman filtering model to obtain error parameters of the inertial device and the navigation system; constructing a fuzzy adaptive model based on the error parameters, generating an observation noise correction factor through a fuzzy logic rule, and correcting an observation noise covariance matrix in the extended Kalman filtering model in real time; carrying out online estimation and compensation on error parameters of the inertial navigation system by utilizing an extended Kalman filtering model, and outputting compensation parameters containing zero offset and scale factors in real time; correcting the original data based on the compensation parameters, injecting the original data into the navigation solution, and updating the fuzzy adaptive model according to the compensated error data.

The invention discloses a remote sensing image aircraft fine-grained classification method based on multi-dimensional part feature modulation, and relates to the technical field of target detection and recognition. A self-supervised pre-trained ViT model is adopted as a backbone network, and remote sensing image set noise correction and dimension adjustment are input to generate a basic feature tensor containing part information; defining a key part prototype and a background prototype, calculating a similarity distance and adding noise, generating an attention map through convolution, calculating a local feature vector and a global feature vector, and splicing and reducing dimensions to obtain fusion features; and constructing a composite loss function, calculating a category score in combination with fusion features, converting the category score into probability distribution through Softmax, and selecting a category with the maximum probability as a final fine-grained classification result. Through multi-scale feature fusion and feature modulation of an aircraft target, the fine-grained classification precision can be effectively improved, and the limitation of poor performance under a complex background and a weak feature target is overcome.

The invention relates to the technical field of hyperspectral alteration, in particular to a satellite hyperspectral alteration mineral remote sensing quantitative recognition method based on deep learning. The method comprises the following steps: acquiring high-resolution 5B satellite hyperspectral data, DEM data and atmospheric parameter data; performing multi-source noise suppression processing on the high-resolution 5B satellite hyperspectral data to obtain a noise correction data set; performing terrain shadow correction based on the noise correction data set and the DEM data, and performing atmospheric correction in combination with the atmospheric parameter data to obtain a pre-corrected hyperspectral data set; through multi-source data fusion, noise and environment interference correction, key wave band screening and multi-dimensional depth feature extraction and fusion, the quality and characterization capability of hyperspectral data are effectively improved, and a solid data basis is provided for subsequent refined mineral recognition and analysis.

In one example, an imaging device includes a light-receiving pixel, a first coupling terminal, a first voltage generation circuit, a drive circuit, a reference signal generation circuit, a noise correction circuit, a comparison circuit, and a processing circuit. The light-receiving pixel generates a pixel signal. The drive circuit is configured to drive the light-receiving pixel on the basis of a voltage from the first voltage generation circuit at the first coupling terminal. The reference signal generation circuit generates a reference signal having a ramp waveform. The noise correction circuit generates a noise correction signal corresponding to the voltage at the first coupling terminal, and superimposes the noise correction signal on the reference signal. The comparison circuit compares the pixel signal and the reference signal with the superimposed noise correction signal. The processing circuit calculates a pixel value based on a result of the comparison.

A method for correcting fixed-pattern noise in at least one image is disclosed. The method includes the following steps:—acquiring at least N substantially distinct and regular images by means of at least one rotating image sensor such that the combination of the N images forms a panorama of a surrounding scene, the fixed-pattern noise in each of the N acquired images being substantially the same;—determining at least one correction parameter for correcting the fixed-pattern noise by means of a processing module from the at least N acquired images, the at least one correction parameter minimizing a functional;—correcting the fixed-pattern noise in each of the N images acquired by the processing module using the at least one determined fixed-pattern noise correction parameter.

The invention discloses a speech enhancement and noise reduction system based on a generative adversarial network, and the system comprises a time-frequency alignment module which is used for collecting and preprocessing original noise speech; the three-component coding module is used for extracting voice content features, voiceprint features and noise characterization; the dual-domain collaborative generation module is used for generating candidate enhanced waveform frequency spectrums and performing collaborative correction; the multi-head discrimination module is used for generating a noise correction factor through a time domain discriminator, a frequency domain discriminator and an index discrimination head; the noise re-projection module is used for correcting noise representation, generating heavy noise voice and obtaining a noise self-adaptive enhancement result set; and the consistency reconstruction module is used for spectrum consistency correction and phase reconstruction. According to the method, the double-domain three-decoupling VoiceGAN is used for voice enhancement and noise reduction, and the method has the advantages of being high in naturalness, high in intelligibility and good in generalization.

To provide an X-ray phase imaging apparatus and an image processing method for improving contrast and visibility of an image photographed by a phase contrast method.SOLUTION: According to an aspect of the present invention, there is provided an X-ray phase imaging apparatus comprising an X-ray source, a detector configured to detect X-rays emitted from the X-ray source to a subject, and an image processing unit configured to generate an image of the subject, the image processing unit includes a storage unit in which a plurality of noise correction images for correcting a luminance variation between pixels are stored separately for brightness and darkness, and an image generation unit that generates a luminance image of the subject by irradiating the subject with X-rays, generates a plurality of corrected images of the subject by correcting the luminance image according to the noise correction image, and generates one subject image by selecting the plurality of corrected images according to a threshold value.SELECTED DRAWING: Figure 5

The invention discloses a multi-camera correction method and system and a medium, and the method comprises the steps: controlling a sampling condition, and analyzing the system gain of each camera; based on the ratio of the standard gain to the system gain, the correction coefficient of each camera is determined to correct each camera, so that the responsivity of each camera after correction is consistent; the method comprises the following steps of: extracting multiple frames of images acquired by the same camera, calculating a time domain noise value of each camera, determining a noise correction coefficient of each camera by adopting a ratio of a standard time domain noise value to the time domain noise value of each camera, and performing noise filtering on a gray value of each pixel point in the images acquired by the cameras according to the noise correction coefficient of each camera, therefore, the noise consistency adjustment of each camera is realized. According to the method, the effect that the noise level is consistent on the premise that the responsivity of the cameras is consistent is achieved, and the universality of the cameras of the same model or different models is improved.

This invention provides a fan speed control method, device, electronic device, and computer-readable storage medium, applied to a range hood controller, and relating to the field of household appliance technology. The method includes: first, responding to user operation and acquiring the initial speed setting and its adjustment value; then, determining the noise variation range; furthermore, real-time monitoring of the range hood's back pressure, determining the noise correction coefficient corresponding to the back pressure using a preset noise correction coefficient table, and adjusting the noise variation range based on the noise correction coefficient; if the noise variation range is large, dividing it into multiple speed control intervals; adjusting the fan's operating state based on the rotational speed corresponding to the endpoints of the speed control intervals and a preset speed control time; by determining the noise variation range, dividing it, and performing linear speed control within the divided speed control intervals, the overall noise variation during the entire speed control process remains within an acceptable range for the user, improving the user experience.

The application relates to a fixed-pattern noise correction method of an infrared detector and belongs to the technical field of infrared imaging. The application can eliminate the fixed-pattern noise of the infrared detector in a real-time image by taking the fixed-pattern noise phenomenon of the infrared detector as a template and calculating the fixed-pattern noise intensity in the real-time image through correlation.

The invention relates to the technical field of noise prediction, in particular to a vibration noise correction method and device based on a structural rigid body mode. The method comprises the following steps: constructing a primary function of elastic displacement to form a first vibration mode matrix, and introducing a rigid body mode into the first vibration mode matrix to form a second vibration mode matrix; and obtaining a structural modal mass matrix, a modal stiffness matrix and a modal force by using the second vibration mode matrix, and deducing a modal participation coefficient and a modal frequency. Performing a simulation test in combination with the deduced modal parameter coefficient and the complete vibration mode matrix to obtain vibration responses of the structure under different frequencies, and predicting the structure according to the vibration responses to obtain simulation data; and dynamically correcting the modal matrix according to the difference between the actual test data and the simulation data of the structure. By introducing the degree of freedom of rigid body motion, the displacement function of the structure can more accurately describe the actual behavior of the structure, so that a more accurate basis is provided for simulating the actual vibration behavior.

This invention relates to the field of pet food spectral detection technology, and discloses a pet food spectral detection device and method. The method involves sampling at equal intervals within a preset wavelength range, eliminating noise interference through dark current and standard white plate calibration; taking the natural logarithm of reflectance to balance the dynamic range; constructing a linear spline basis at equal intervals, and adaptively calculating the regularization coefficient based on the energy ratio of the signal and basis functions; constructing an augmented matrix based on the inner product of the basis functions and the signal, and solving for the spline coefficients using Gaussian elimination; reconstructing the spectrum using the coefficients and quantifying the residuals; establishing a reference model based on the mean and standard deviation of multiple qualified sample coefficients; comparing the Euclidean distance between the coefficients of the sample to be tested and the model mean with a threshold, and generating residuals, distances, thresholds, and a judgment report; the entire process requires no empirical parameters or manual tuning, enabling batch online adaptive detection, achieving rapid and accurate detection with noise correction, feature enhancement, overfitting suppression, and traceability.

The application relates to the technical field of video data processing, and relates to an audio-video synchronous noise reduction method and system based on AI visual perception, which comprises the following steps: pre-processing audio data to obtain pre-processed audio data; performing time-frequency analysis on the pre-processed audio data to obtain a speech feature set and a background sound feature set; performing noise reduction on video data to obtain primary noise reduction video data, performing visual perception on the primary noise reduction video data to obtain a video feature set; performing time axis correction operation on the speech feature set based on a mouth shape feature according to the video feature set to obtain an updated time axis; performing active noise reduction operation on the progress correction audio data according to the updated time axis and a pre-constructed background sound adaptation degree sequence to obtain noise correction audio data; and performing merging operation on the noise correction audio data and the primary noise reduction video data to obtain synchronous noise reduction audio-video. The application can improve the clarity of images and sounds in a video.

This invention discloses a direct snow depth inversion method based on a spaceborne photon-counting lidar, comprising: extracting the target snow-covered area, removing areas with thick cloud cover and non-target surface point clouds, aligning the surface, and performing noise correction; performing deconvolution to obtain a corrected snow attenuation backscattering profile; setting an initial snow absorption coefficient, inverting the backscattering path length distribution of non-absorbing snow, and calculating the first and second moments of this distribution; based on the first and second moments, simultaneously inverting snow depth, snow albedo, and diffuse scattering coefficient; combining the snow albedo and diffuse scattering coefficient to calculate an updated snow absorption coefficient; if the updated snow absorption coefficient does not meet the set convergence threshold, using the updated snow absorption coefficient as the new initial snow absorption coefficient, and repeating the iteration; if it meets the threshold, outputting the final snow depth. Using this invention, along-track snow depth inversion can be achieved directly without relying on any other data.

A kind of hot infrared non-uniform noise correction method based on motion scene, comprising: recording multiple continuous motion scene images with hot infrared camera;Motion scene image is divided into multiple non-overlapping regions;Estimate the difference of non-uniform noise corresponding to different pixels at a time;Establish noise mathematical model, solve actual non-uniform noise using least square method;Design mask, remove non-uniform noise estimation error part;After removing outliers, non-uniform noise graph is filled, original image is subtracted from noise graph, and non-noise image after correction can be obtained.The application proposes a new non-uniform noise estimation modeling idea, and designs improved median extraction algorithm, which can effectively extract the difference relationship of non-uniform noise between different pixels.All processing processes only need a few original dynamic video frames, avoid the strict requirement of traditional non-uniform correction algorithm to scene and video frame number, so that it can be widely applied to non-uniform correction in different environments.

The present application provides a kind of enterprise labor efficiency evaluation model construction method, medium and system, belong to artificial intelligence model technical field, the present application is by gathering staff behavior data to construct high-dimensional sparse feature matrix, after coding by hash embedding layer, gaussian mixture distribution modeling and double network cross screening noise correction are carried out to labeled label, employee collaboration relationship graph is constructed based on training sample set, after hierarchical random neighbor sampling and graph clustering presegmentation, employee performance evaluation vector is obtained by inputting causal inference enhancement double-flow comparative evaluation model, then the efficiency distribution is time series evolution by optimal transport gradient flow continuous flow algorithm, finally, according to the deviation of prediction result and equilibrium threshold, external incentive term is adjusted and intervention strategy is output, the technical problem that employee performance evaluation model cannot accurately estimate causal effect in organization network and realize group efficiency time series evolution prediction is solved.

The invention discloses a detection-guided noctilucent remote sensing image irregular stripe noise removal method and device, and the method comprises the steps: carrying out the intense light source, non-illumination and to-be-processed region division and edge enhancement of an original image, and obtaining an enhanced image; performing stripe noise iterative detection according to the enhanced image in combination with frequency domain analysis and multi-constraint Hough transform to obtain a stripe mask; aligning the strip mask with the binary image of the original image based on morphological operation, and performing inter-band filtering to obtain a noise-free mask; obtaining a noise target area according to the noise-free mask and the original image; performing progressive noise correction on the noise target area based on a dynamic window and an adaptive correction algorithm to obtain an initial result; and performing residual error correction according to the initial result in combination with the strong light source region to obtain a stripe-removed image.

The invention discloses a label propagation text classification method and device generated by fusing weak supervision information, and relates to the technical field of text classification. The method comprises the steps of selecting an initial category word set, pre-training an initial multi-label text classification model, inputting an original text into an encoder layer to obtain deep potential features, and inputting the deep potential features into a prediction layer to obtain an initial classification prediction result; determining a pseudo label set; gradually updating the pseudo label set, and further training the pre-trained multi-label text classification model; an integrated pseudo label set is obtained; determining an adjacent matrix of the k-neighbor graph; determining a label correlation matrix; performing noise correction on the integrated pseudo label set, and performing final training on the multi-label text classification model; and according to the trained multi-label text classification model, obtaining a label corresponding to the to-be-classified text. According to the invention, the noise supervision information is corrected by using the text neighbor relation and the label correlation, so that the classification accuracy is improved.

The invention discloses a potentiometer noise correction method. The method comprises the following steps: performing noise suppression processing on an analog signal which is output by a potentiometer and maps the working state of the potentiometer; converting the analog signal subjected to noise suppression into a digital signal, and performing preprocessing operation on the digital signal; calculating a predicted value by combining a voltage change rate parameter of the potentiometer and a confidence estimation value generated in the previous period, wherein the voltage change rate parameter comprises a positive change rate when the voltage of the potentiometer rises and a negative change rate when the voltage of the potentiometer drops; calculating the deviation between the preprocessed digital signal and the confidence estimation value of the previous period, and judging the credibility of the preprocessed digital signal based on a preset curing confidence interval; determining a current confidence estimation value according to the credibility judgment result; compared with the prior art, the method covers the vast majority of potentiometer noise correction demand scenes.

The application relates to an oil-out grade analysis method based on an oil measurement surface paper, which comprises the following steps: obtaining an image to be analyzed of the oil measurement surface paper; performing a pretreatment operation on the image to be analyzed to obtain a pretreated image; performing edge detection on the pretreated image to obtain an edge image, and performing inclination correction on the edge image to obtain an image to be area-analyzed; determining an oil-out range of the image to be area-analyzed to obtain a preliminary oil-out range; performing noise correction on the preliminary oil-out range based on a preset floating value to obtain an accurate oil-out range; marking the accurate oil-out range in the image to be area-analyzed, and pushing the marked image to be area-analyzed to a user terminal interface after visualizing the marked image. The application can make the finally obtained accurate oil-out range as accurate as possible, and finally visually display the accurate oil-out range to the user, so that the user can understand the product use condition of the user, and the user experience is improved.

According to the model training method and device and the model application method and device, irrelevant information is filtered from the text feature of the annotation text and the second image feature of the redundant image, cross-modal alignment is improved, seamless fusion of the text, the background and the texture in scene text editing is enhanced, and the user experience is improved. A noise correction mask of the first image feature of the to-be-edited image is obtained through a noise correction unit, and prediction noise of the first image feature is adjusted based on the noise correction mask, so that the accuracy of noise estimation is improved; and through the depth feature of the first image feature, guiding the second depth feature extraction network to edit the first image feature through a second fusion feature obtained after the text feature and the second image feature are fused to obtain an editing result so as to enhance the consistency of the edited fonts.

The invention relates to an attitude-driven video editing method based on diffusion model time sequence consistency modeling, and belongs to the technical field of video editing.The method comprises the steps that initial noise is obtained through DDIM inversion, and initialization of a local editing area is achieved in combination with a binary mask and random noise; utilizing KL divergence to measure characteristic difference, constructing a guide function and a joint scoring function, and optimizing a generation process; further introducing a posterior score function and a noise correction mechanism to ensure that a non-edited region is consistent with an original video; and through cross-frame feature fusion, the time sequence coherence is enhanced. According to the method, high-precision and high-consistency video posture editing can be realized, the generation quality and generalization ability are remarkably improved, and the method is suitable for intelligent video processing of complex dynamic scenes.

The application provides a non-uniform noise correction method, an electronic device and a computer program product. In the application, after the output of the infrared imaging device at the current temperature is corrected each time, the non-uniform noise fpn_M is learned based on the correction result, and the background data corresponding to the highest temperature value and the background data corresponding to the lowest temperature value in the temperature interval where the temperature is located are adjusted according to the learned non-uniform noise fpn_M. In this way, the non-uniform noise data dynamically learned is updated to the background data in real time, so that the background data in the stored background-temperature data can be updated in real time based on the non-uniform noise data. Since the stored background-temperature data is used to correct the output of the infrared imaging device at a temperature, it is equivalent to realizing the non-uniform noise correction of the output of the infrared imaging device at a temperature.

This invention provides a source domain noise correction method for unsupervised domain adaptation, belonging to the field of machine vision domain adaptation. The method first obtains all samples X in the source domain. S The category is N. The feature extractor and classifier are pre-trained in the source domain. The pre-trained feature extractor is then used to extract the feature set Z from all samples in the source domain. S The extracted feature set Z was processed using a Gaussian mixture model (GMM) with K=N cluster categories. S The method involves modeling, calculating and comparing the probability of each sample x belonging to each cluster, and updating the label of the current sample by taking the cluster label of the cluster with the highest probability. Then, the number of samples is counted using the label transformation matrix A, and the original label of the sample with the highest number of samples corresponding to the new label is used as the corrected label for all samples corresponding to the new label, thus completing the correction of source domain noise. This invention reclassifies the noisy source domain dataset according to its maximum probability value, reducing the noise ratio of the dataset and enhancing the robustness of the unsupervised domain adaptation method.

The invention relates to an image noise reduction method and device, equipment and a storage medium, and relates to the technical field of image noise reduction. The method comprises the following steps: acquiring a to-be-processed first image; determining depth information of the first image; calculating a noise correction parameter based on the depth information; and based on the noise correction parameter, performing noise reduction processing on the first image to obtain a noise-reduced image. The method comprises the steps of obtaining a to-be-processed first image; determining depth information of the first image; calculating a noise correction parameter based on the depth information; and based on the noise correction parameter, performing noise reduction processing on the first image to obtain a noise-reduced image. The noise correction parameter is calculated by using the image depth information, and three-dimensional space information is considered during noise reduction processing, so that noise and real details can be distinguished more accurately, and the image noise reduction effect is improved.

The application discloses a small-area BDS-3 real-time normal high intelligent measurement method and an intelligent measurement system based on deep learning, and comprises the following steps: 1, preprocessing and standardizing space remodeling of sampling point data measured by BDS-3 RTK; 2, stripping a wide-area background field component from original observation; 3, introducing a spatial correlation measure model with a second-order continuous characteristic, introducing an adaptive observation noise correction operator, and constructing a full-field covariance matrix to absorb random observation noise and constrain a fitting process; 4, introducing a super parameter group and constructing a target loss function of deep learning; 5, establishing a back propagation path under a deep learning architecture; performing deep optimization on an evidence function of the constructed model to obtain an optimal super parameter group; 6, performing multi-scale linear superposition on the wide-area background field component and a local refinement correction component to realize reconstruction; and 7, outputting a system result. The model constructed based on the Beidou positioning technology effectively improves the robustness of the model through deep learning.

A temperature noise correction method for CMOS space cameras based on attention mechanisms and LSTM is presented. This method relates to the field of optical remote sensing technology, specifically to the field of temperature noise correction for CMOS space cameras. The core of the method is the use of a multi-level Long Short-Term Memory (LSTM) network with an attention mechanism to explore how temperature changes affect the noise performance of CMOS space cameras under two different operating conditions: dark and bright fields. This deep learning model possesses powerful autonomous learning capabilities, capable of mining and understanding complex noise patterns hidden within massive amounts of data, thereby improving the accuracy of noise identification and calibration. Its key advantage lies in its end-to-end learning approach, automatically learning and establishing a deep, nonlinear relationship between temperature and noise directly from the raw data, without requiring manual intervention for feature extraction or designing complex correction algorithms. This significantly enhances the model's generalization ability and adaptability to unseen data scenarios.

The application discloses a phase scanning based road surface micro-texture depth evaluation method and system, and particularly relates to the field of road surface micro-texture depth evaluation.The scheme comprises the following steps: a laser of a preset wavelength is emitted to a road surface by a laser phase scanning device; an original phase difference signal of the road surface micro-texture is collected by a phase detection unit; a laser incidence angle, an illumination angle and road surface image data are synchronously obtained by an environment sensing unit; an initial micro-texture depth of the road surface micro-texture is calculated; texture feature extraction is performed on the obtained road surface image data; the initial micro-texture depth is dynamically corrected by combining an illumination correction factor obtained based on the illumination angle; a final micro-texture depth is obtained; and the final micro-texture depth is converged until the final micro-texture depth converges.The application realizes high-precision and high-adaptability evaluation of the road surface micro-texture depth by means of illumination correction, texture density correction, noise correction and dynamic iteration optimization, and by combining image data processing, three-dimensional image generation and other technologies.

The application provides a correction system and method of a shock absorber simulation model and a road noise correction system. The correction system comprises a shock absorber simulation model, a modal test component and a modal analysis device. The modal analysis device is configured to obtain standardized first modal data and second modal data, perform matching analysis on the second modal data and the first modal data, and correct multiple simulation parameter values of the shock absorber simulation model until a matched shock absorber simulation model is determined when it is determined that the second modal data and the first modal data do not match. A two-stage progressive correction system framework of "test-parameter coupling correction" is proposed, which reduces the complexity of the system. A multi-dimensional parameter synchronous correction mechanism of the shock absorber simulation model is proposed, which integrates three types of parameter collaborative correction of damping nonlinearity, bushing frequency variable stiffness and main structure modal. Performance correction is performed through modal fitting, which improves the modeling accuracy of the shock absorber simulation model.

In one example, an imaging device includes a light-receiving pixel, a first coupling terminal, a first voltage generation circuit, a drive circuit, a reference signal generation circuit, a noise correction circuit, a comparison circuit, and a processing circuit. The light-receiving pixel generates a pixel signal. The drive circuit is configured to drive the light-receiving pixel on the basis of a voltage from the first voltage generation circuit at the first coupling terminal. The reference signal generation circuit generates a reference signal having a ramp waveform. The noise correction circuit generates a noise correction signal corresponding to the voltage at the first coupling terminal, and superimposes the noise correction signal on the reference signal. The comparison circuit compares the pixel signal and the reference signal with the superimposed noise correction signal. The processing circuit calculates a pixel value based on a result of the comparison.