61 results about "Resolution improvement" patented technology

The invention relates to the technical field of underwater detection, and discloses a seabed sonar image feature enhancement processing method and a storage medium. The method comprises the following steps: acquiring multi-source seabed sonar image data, and performing space-time alignment and signal denoising processing to obtain a standardized sonar data sequence; performing multi-scale feature extraction on the standardized sequence to generate a feature map containing texture, shape and strength information; inputting the characteristic spectrum into a space-time attention network for characteristic weight distribution and fusion, and outputting enhanced characteristic representation; based on the enhanced features, a deep learning model is adopted to carry out resolution improvement and detail recovery, and a high-resolution sonar image is generated; and performing region division on the high-resolution image through an adaptive threshold segmentation algorithm, verifying feature consistency in combination with historical data, and finally completing feature enhancement processing. According to the method, the definition, the detail expressive force and the feature separability of the submarine sonar image are improved, and accurate recognition and analysis of submarine topography and targets are effectively supported.

This invention discloses an online super-resolution imaging method for scanning radar based on the minimization criterion. By establishing an azimuth echo convolution model and discretizing it, the super-resolution problem is transformed. Under a regularization framework, the minimization criterion is used to construct an analytically optimizeable cost function containing sparse prior information, transforming the forward-looking imaging resolution improvement problem into an optimization estimation problem. By dividing the antenna measurement matrix into rows and blocks, matrix operations are degraded into the continuous accumulation of vector operations, enabling real-time online updates of targets in the forward-looking region. Finally, the conjugate gradient method is used to degrade and invert the matrix, outputting the super-resolution result. This method can achieve rapid imaging with limited system memory through an online conjugate gradient super-resolution algorithm based on the minimization criterion. Compared with existing batch processing regularization methods, it eliminates false targets, significantly reduces computational complexity, and is beneficial for rapid imaging of scanning radar.

The application relates to the field of semiconductor optoelectronics, in particular to a self-powered up-conversion imaging array and a preparation method thereof. The technical key points are as follows: array pixel units are arranged on an insulating substrate and are mutually isolated; each unit contains a central up-conversion detection subunit and a peripheral photovoltaic power supply subunit; the photovoltaic unit is formed by connecting micro photovoltaic cells in series, and the anode and the cathode are connected with the cathode and the anode of the central up-conversion detection subunit respectively, so as to form an independent closed loop series driving circuit in the pixel; the central up-conversion detection subunit is used for receiving incident near-infrared light signals and converting the signals into visible light signals; the peripheral photovoltaic power supply subunit is used for absorbing incident light and converting the light energy into electric energy to provide a driving voltage and a driving current; the peripheral photovoltaic power supply subunit and the central up-conversion detection subunit meet the current matching condition; the effects of self-power supply at the pixel level, complete elimination of electrical crosstalk between pixels, improvement of imaging contrast and resolution, improvement of photon utilization rate and device fault tolerance are achieved.

The application discloses a time-frequency electromagnetic exploration method and system for improving exploration depth and resolution, and relates to the technical field of exploration. The method comprises the following steps: arranging a multi-directional excitation source and a mobile remote end array observation system, and sparsely collecting time-frequency electromagnetic signals; reconstructing data and reducing noise based on compression sensing; constructing a time-frequency dual-domain joint inversion target function, introducing cross gradient constraints and physical constraints, and realizing intelligent collaborative inversion through depth reinforcement learning; outputting a high-precision three-dimensional geological model through Bayesian uncertainty analysis and multi-source data fusion verification; and the system comprises excitation, observation, data processing, intelligent inversion and interpretation visualization modules. The scheme improves exploration depth and resolution, improves data acquisition and inversion efficiency, and guarantees the accuracy of geological interpretation.

The invention relates to a flow field super-resolution reconstruction method, computer equipment and a readable storage medium. Obtaining a to-be-reconstructed multiphase flow simulation result corresponding to the to-be-simulated multiphase flow case; performing image mapping processing on the to-be-reconstructed multiphase flow simulation result to obtain to-be-reconstructed flow field image data which is in space-time alignment with the to-be-reconstructed multiphase flow simulation result; and through the trained flow field super-resolution reconstruction model, performing super-resolution prediction on the to-be-reconstructed flow field image data to obtain target reconstructed flow field image data of which the image resolution is higher than that of the to-be-reconstructed flow field image data. By converting low-resolution simulation data into image representation adaptive to deep learning and by means of a trained flow field super-resolution reconstruction model, a complex mapping relation between a high-resolution flow field and a low-resolution flow field is learned to realize rapid resolution improvement, and while the calculation precision of high-resolution flow field data is ensured, the calculation efficiency of the high-resolution flow field is improved. And the calculation efficiency of the high-resolution flow field data is effectively improved.

The invention discloses a resolution improving method and device, electronic equipment and a readable storage medium, and relates to the technical field of image processing, and the method comprises the steps: obtaining a current frame image of a first resolution, a historical frame image of a second resolution adjacent to the current frame image, and an image change parameter; performing sampling operation on the current frame image at a second resolution to obtain a first color parameter of each first pixel point in the current frame image after the sampling operation; a second color parameter of each second pixel point in the historical frame image is corrected based on a color clamping parameter and the image change parameter, a third color parameter of each second pixel point is obtained, and the color clamping parameter is determined based on the first color parameter; and performing color mixing operation on the first color parameter of each first pixel point in the current frame image based on the displacement parameter and the third color parameter of each second pixel point in the historical frame image to obtain the current frame image with the second resolution. The color effect of an image after super-resolution can be improved.

The application discloses a low-contrast image joint enhancement and super-resolution reconstruction method, which comprises the following steps: firstly, performing multi-scale Gaussian filtering on the low-contrast image; secondly, expanding the imaging dynamic range of the quotient image by using an S function; thirdly, performing color correction on the expansion result to obtain a contrast-improved image; and finally, optimizing the resolution and detail information of the contrast-improved image by a blind super-resolution reconstruction model constrained by a total variation to obtain a detail information and resolution improvement result. The low-contrast image joint enhancement and super-resolution reconstruction can help to overcome the influence of image adverse factors, such as low contrast / resolution and fuzzy details, and lay a foundation for image feature extraction and image interpretation.

The present invention relates to the technical field of data processing. Disclosed are a data-analysis-based resolution improvement method and system for a mass spectrometer. The method comprises: training a resolution factor model; on the basis of an original mass spectrum and an original fragment map, determining basic information of a compound to be tested, determining a target signal-to-noise ratio of the original mass spectrum on the basis of a mass spectrum of a standard compound, and performing denoising processing on the original mass spectrum by means of a denoising method; on the basis of the denoised original mass spectrum and the mass spectrum of the standard compound, determining a baseline reference region; identifying features of peaks in the mass spectrum, separating overlapped peaks by means of the features of the peaks, and performing clustering analysis on all the peaks on the basis of a resolution factor; and performing dimension reduction processing on a class of peaks having data dimensions exceeding a dimension threshold value. In this way, data inconsistency caused by baseline drifts is reduced or avoided. Finally, the resolution of a mass spectrometer is improved by means of the separation of overlapped peaks, the clustering of peaks, and dimension reduction processing of the peaks, thereby ensuring the efficiency of resolution improvement.

The invention discloses a seismic data resolution improving and weak signal enhancing method, and belongs to the technical field of geophysical seismic data processing. According to the method, a high-precision seismic single channel synthesized by logging information is introduced as a physical constraint to be input into the conditional generative adversarial network, constraint is provided for a conditional generative adversarial network generator in the training and reasoning process, the physical constraint is used as auxiliary supervision of a conditional generative adversarial network discriminator, and the authenticity of a generation result and the geological rationality are improved. Meanwhile, a composite loss function combining image loss, adversarial loss, perception loss, a total variation regular term and well control matching loss is designed, and unification of pixel-level error constraint, detail feature keeping and geological physical consistency is achieved. According to the method, weak reflection signals can be enhanced under the condition of low signal-to-noise ratio, high-frequency details are recovered, structural illusion is suppressed, and the resolution and the signal-to-noise ratio of seismic data are effectively improved. The method provided by the invention still has good actual data generalization ability after synthetic data training.

A height measurement device (100) according to the present disclosure is provided with a height measurement information acquisition unit (4) that, while moving relative to an object (Ta) to be measured in the horizontal direction, acquires height measurement information for each measurement section (Mes) that is divided in each of one direction of the horizontal direction and an orthogonal direction that is orthogonal to the one direction, and acquires height measurement information for each measurement section (Mes) that is divided in each of the one direction of the horizontal direction and the orthogonal direction that is orthogonal to the one direction. Height measurement information for measuring the height of an object (Ta) to be measured is acquired. A height measurement device (100) is provided with a control unit (5) for increasing the resolution of height measurement information by adding, to the height measurement information, a plurality of pieces of second height measurement information obtained by shifting the position of a measurement section (Mes) by a predetermined amount (Loff) within the range of the length (Lpv) of the measurement section (Mes) after the plurality of pieces of first height measurement information have been obtained.

This system uses both visible and invisible light images, and employs machine learning with appropriate training data to perform image correction such as noise reduction and resolution improvement. [Solution] The image processing device comprises a machine learning unit that performs learning and inference for processing non-visible light images using training data, an environmental information acquisition unit that acquires information about the surrounding environment, an effectiveness determination unit that determines the effectiveness of visible light images as training data based on the environmental information acquired by the environmental information acquisition unit, and a training data selection unit that determines whether a visible light image is effective as training data for a non-visible light image that is temporally corresponding to that visible light image, based on the effectiveness determined by the effectiveness determination unit, and selects training data that has been determined to be effective.

This invention discloses a nonlocal visual enhancement and resolution improvement method, belonging to the field of image processing technology. First, the degraded image to be processed is acquired, and the initial scattering map and initial scene radiance are estimated. Then, a variational scene restoration model is constructed. Finally, the model is iteratively solved using the semi-proximal alternating direction multiplier method. In each iteration, the optimized scattering map, auxiliary variables, and high-resolution scene radiance are updated sequentially until convergence, at which point the high-resolution scene radiance is output as the enhanced image. This invention employs the aforementioned nonlocal visual enhancement and resolution improvement method. By incorporating visual enhancement and resolution improvement into the variational scene restoration framework, it uses nonlocal fractional gradient regularization and depth super-resolution prior joint constraints, and utilizes the semi-proximal alternating direction multiplier method for efficient iterative solution. This approach can simultaneously suppress artifacts and color distortion, restore clear details, and significantly improve the overall visual quality of the final image.

The application discloses a method for improving the imaging resolution of a while-drilling azimuthal gamma logging device based on regularization processing, and relates to the technical field of oilfield logging, and comprises the following steps: acquiring azimuthal response sensitivity data of the while-drilling azimuthal gamma logging device; collecting multi-sector gamma imaging data of the azimuthal gamma logging at each depth according to the azimuthal response sensitivity data; and performing regularization processing on the multi-sector gamma imaging data to obtain new gamma data close to the real radioactivity intensity distribution of a formation, so that the imaging resolution of the while-drilling azimuthal gamma logging is improved. The application solves the problems of low imaging resolution and large measurement error of the existing logging.

The resolution of the pixel array structure is increased, further improving display quality. [Solution] The pixel array structure comprises a plurality of pixel groups (01), each pixel group (01) including a first subpixel (111), a second subpixel (112), a third subpixel (113), and a fourth subpixel (114). In the pixel group (01), the connecting line between the center (C2) of the second subpixel (112) and the center (C3) of the third subpixel (113) is the first line segment (LS1). The first subpixel (111) and the fourth subpixel (114) are located between the second subpixel (112) and the third subpixel (113), and are provided on both sides of the first line segment (LS1). The connecting line between the center (C1) of the first subpixel (111) and the center (C4) of the fourth subpixel (114) is the second line segment (LS2), whose length is shorter than the length of the first line segment (LS1).

The invention discloses a super-resolution image model optimization method and an image super-resolution model. The method comprises the steps of inputting a received low-resolution image into an image super-resolution model, extracting shallow-layer feature information and deep-layer feature information by using a feature extraction module, obtaining frequency domain feature information of the image through an adaptive frequency spectrum optimization module, carrying out feature information superposition through a feature fusion module, generating fusion feature information, and outputting the fusion feature information to the image super-resolution model. And performing up-sampling operation on the fused feature information through an up-sampling module, performing super-resolution image enhancement processing on the low-resolution image based on the fused feature map to obtain a target image with improved resolution, calculating a loss value according to the amplitude spectrum of the target image and the optimized amplitude spectrum, and constraining the image super-resolution model based on the loss value. According to the method and the device, the technical problem that feature missing of a reconstructed image is easily caused due to relatively low model constraint when a low-resolution image is mapped to a high-resolution image in the prior art is solved.

The invention provides a water body extraction method fusing CYGNSS data and SAR data, and belongs to the technical field of remote sensing disaster monitoring. According to the method, the advantages of high sensitivity of CYGNSS data to surface roughness change and high spatial resolution of SAR data are utilized, deep fusion and resolution improvement of two types of data are realized through a random forest model, the resolution adaptation problem is solved, and the separability of a water body and a non-water body can be enhanced, so that the water body recognition accuracy is improved. The method further introduces a super-pixel improved algorithm of DEM perception, is not limited to single slope threshold rejection, and constructs an effective water body elevation range by analyzing elevation distribution of a full-graph high-confidence water body. The superpixels are constrained by using the range, so that the high-altitude mountain shadow can be effectively eliminated, the leak detection weak signal water body can be recalled through the elevation characteristics, and the water body extraction precision under the complex terrain is remarkably improved.

The application provides a detection signal high-frequency continuation method and related equipment, constructs a local dip angle estimation network to process target deep seismic data to obtain local dip angle field data, captures signal edge time delay characteristics through a dynamic spatial attention network, constructs a seismic facies hidden space representation model to extract geological ages corresponding to seismic facies hidden space representation results; the geological ages and the seismic profile data are input into an inversion physical model for processing to obtain elastic parameter data, and the elastic parameter data, well logging data and well side seismic data are input into a forward physical model for reconstruction to obtain time-varying wavelet spectrum characteristics to construct a high-frequency continuation inversion target function, complete inversion solving, obtain a high-resolution reflection coefficient sequence, and after convolution of the high-resolution reflection coefficient sequence and the time-varying wavelet spectrum characteristics, obtain deep seismic data after high-frequency continuation; and the technical difficulties of high-fidelity, high-robustness band widening and resolution improvement of deep seismic data are solved.

The application discloses a kind of spatiotemporal spectrum joint super-resolution reconstruction methods based on giant remote sensing star group, including from multiple isomerism satellites acquisition same area time series, multi-angle and multispectral data and carry out radiation calibration and atmospheric correction;Subpixel level alignment of multi-source data is realized using joint registration model;Three-dimensional convolution is used to extract time-varying characteristics, two-dimensional convolution is used to extract spatial structure and texture characteristics, and one-dimensional convolution is used to extract and reduce dimension spectral characteristics;Through attention mechanism, time, space and spectral characteristics are adaptively weighted and fused to generate joint feature tensor;Super-resolution reconstruction is carried out, and high spatial resolution, high time resolution and high spectral fidelity target image are obtained.The method considers resolution improvement and spectral authenticity, and is suitable for fine city mapping, agricultural monitoring, ecological environment assessment, disaster emergency and battlefield situation awareness, long-range reconnaissance, target change detection and damage assessment and other high-precision remote sensing application scenarios.

The invention provides an optical flow-driven high-quality space-time cascade optical field video generation method. The method comprises five processes of acquisition of a low-frame-rate and low-resolution micro-image array sequence, depth perception optical flow estimation, adaptive dynamic compensation time interpolation, structure perception super-resolution and three-dimensional reconstruction of a high-frame-rate and high-resolution micro-image array sequence. According to the method, light field information is collected through a lens array, a low-frame-rate and low-resolution micro-image array sequence is generated to serve as input, a depth perception light flow estimation method is utilized to be fused with depth information to optimize a motion field, and local self-adaptive interpolation is executed to generate an intermediate frame on the premise that the integrity of a micro-image structure is kept based on the optimized light flow field. Spatial resolution is improved through multi-frame motion compensation and feature fusion, a high-frame-rate and high-resolution dynamic three-dimensional video is reconstructed by means of an integrated imaging display device, and the space-time quality and the three-dimensional display effect of the light field video are effectively improved.

This disclosure relates to a method, system, device, medium, and procedure for deconvolution of seismic data. The method includes: performing multi-level noise suppression on seismic data to obtain denoised seismic data; performing geometric correction and data realignment on the denoised seismic data to obtain a common centroid gather, and dividing the common centroid gather into time windows to obtain a stratigraphic time window group; performing seismic trace analysis on the denoised seismic data to obtain a seismic trace energy spectrum matrix; performing spectral factor decomposition on the seismic trace energy spectrum matrix based on the denoised seismic data to obtain a spectral component group; and performing dual-window deconvolution on the denoised seismic data based on the spectral component group and the stratigraphic time window group to obtain standard seismic data. This improves the resolution and accuracy of the seismic data after deconvolution.

The utility model discloses a CT resolution uniformity assessment die body, comprising a die body main body, the die body main body is cylindrical, the die body main body is sequentially provided with nine layers of modules from top to bottom, the nine layers of modules are divided into three modules, the first layer of modules to the third layer of modules are first modules for assessing tangential resolution, and the fourth layer of modules are second modules for assessing tangential resolution. The modules from the fourth layer to the sixth layer are second modules for evaluating the radial resolution, the modules from the seventh layer to the ninth layer are third modules for evaluating the contrast, and resolution line pairs parallel to the diameter of a module main body are uniformly arranged on the first module; resolution line pairs perpendicular to the diameter of the module body are evenly arranged on the second module, and round filler is evenly arranged on the third module. According to the utility model, on one hand, the resolution uniformity motif can be used for evaluating the spatial resolution uniformity of an image, and on the other hand, the advantages and disadvantages of a resolution improvement algorithm can be evaluated.

A device for improving the interplanar resolution of a three-dimensional medical image according to the present invention comprises an image generation module for improving the resolution of a three-dimensional medical image whose interplanar resolution is lower than the intraplanar resolution, wherein the image generation module: reconstructs an externally input first medical multi-layer image set in a preset direction to form cross-sections perpendicular to the layer planes, thereby generating a second medical multi-layer image set whose resolution in the preset direction is low;inputs the second medical multi-layer image set into a pre-trained deep learning model to generate a third medical multi-layer image set from the layers contained in the second medical multi-layer image set, the resolution of which is improved by an integer multiple in the preset direction by the deep learning model; and generates and outputs a high-resolution three-dimensional image with improved resolution based on the third medical multi-layer image set.

The application discloses a kind of super-resolution imaging method and device based on quantum correlation photon parallel collection, the present application uses optical beam splitter to divide the fluorescence detected in microscope into two beams again into parallel detection system, two-way parallel detection can realize more efficient collection to fluorescence correlation photon pair after correction alignment, to reach better image quality.Compared with single parallel detection light path, the present application is used in quantum dot and solid defect with single-photon source characteristics;The present application can realize resolution improvement without improving laser power by using simple beam splitting second-order correlation measurement.Compared with existing quantum correlation imaging technology, the two-way parallel collection mode designed in the present application can perform coincidence measurement in space-time, and more coincidence photon information is retained, the imaging time is shortened, which can be applied in different super-resolution light paths, and the application range of quantum correlation photon detection technology is widened.

The application relates to a resolution improvement control method and device of a liquid crystal device and the liquid crystal device. The method comprises the following steps: S1, acquiring a high-resolution image to be displayed, and decomposing the high-resolution image to obtain a plurality of low-resolution images; S2, judging whether a current high-resolution image display period is a positive display period or a negative display period; S31, in the positive display period, controlling an array of display units to display the plurality of low-resolution images in a first order in time, and simultaneously controlling the direction of a galvanometer to offset the position of the plurality of low-resolution images; S32, in the negative display period, controlling the array of display units to display the plurality of low-resolution images in a second order in time, and simultaneously controlling the direction of the galvanometer to offset the position of the plurality of low-resolution images; and S4, returning to step S1 to acquire a new high-resolution image. The application changes the display order of the plurality of low-resolution images, thereby reducing the risk of image sticking and screen burning of the liquid crystal device when the resolution is improved.

The application provides a multi-channel seismic deconvolution method based on two-dimensional K-SVD and convolution sparse coding, global frequency division processing is performed on seismic data according to convolution sparse coding; secondly, two-dimensional K-SVD-based seismic data low-frequency component deconvolution processing is performed to learn the horizontal characteristics from two-dimensional low-frequency seismic data; then, convolution sparse coding-based seismic data high-frequency component deconvolution processing is studied; finally, the advantages of single-channel deconvolution and multi-channel deconvolution are complementary, vertical and horizontal constraints are added in the deconvolution objective function, so that the deconvolution result has both resolution improvement and continuity enhancement. The application learns the reservoir reflectivity characteristics and the mapping relationship with the seismic data characteristics from the logging data, has the characteristics of high vertical resolution, simultaneously adds the horizontal distribution law of the seismic data to the deconvolution processing, so that the inversion result presents strong horizontal continuity, and the resolution is significantly enhanced compared with the seismic data.

This application discloses an image processing method, apparatus, electronic device, and non-volatile storage medium. The method includes: acquiring an original core image of a target rock and determining a residual image corresponding to the original core image, wherein the residual image is used to supplement missing details in the original core image; performing a magnification operation on the original core image and the residual image; compensating and reconstructing the original core image of the target size based on the residual image of the target size to obtain a reconstructed core image; and performing threshold segmentation processing on the reconstructed core image to obtain the target core image, wherein the resolution of the target core image is higher than that of the original core image. This application solves the technical problems of insufficient resolution improvement and inaccurate detail reproduction in related technologies for core image super-resolution reconstruction when dealing with complex shale structures and micron-level cracks.

The invention belongs to the field of geophysical exploration and the technical field of artificial intelligence, and particularly relates to a seismic image self-supervision super-resolution reconstruction method based on diffusion model data augmentation, which comprises the following steps of: firstly, carrying out data augmentation, constructing a data set and initializing model parameters, and inputting a small amount of high-quality seismic images into a diffusion model for training; selecting a network model parameter with the minimum loss function value as an optimal solution; performing data augmentation by using the stored optimal network model to generate a new high-quality seismic image, and verifying the validity of the augmented data; constructing a data set of a super-division network model through downsampling and filtering, and designing the super-division network model; the method comprises the following steps: performing parameter initialization and training detail setting, inputting augmented data into SRUNet for training and storage, finally testing a stored optimal model in a test set, verifying the accuracy of the model through indexes such as structural similarity, and applying the model to seismic data resolution improvement in a larger range.

The application provides a superpixel-based resolution improvement method and a scanning controller, relates to the technical field of image display, and can improve the resolution without increasing the size of a display screen, multiplexes information of adjacent pixels, makes one light-emitting diode participate in imaging of multiple adjacent pixels, and meets the requirements of ultrahigh resolution and refresh rate of a micro display; the method comprises the following steps: S1, performing image processing on an original image; S2, determining the number of frames for separating the original image according to the spatial resolution of the original image; S3, separating the original image into a plurality of frame sub-images according to the number of frames determined in S2 and the row and column positions of each pixel in the original image, and meanwhile adjusting the gray intensity of each pixel color channel according to the separation ratio; S4, expanding the pixels in the sub-image by dividing sub-pixels; and S5, superimposing and merging the plurality of expanded frame sub-images to obtain a high-resolution image.

This invention relates to the field of geophysical exploration and discloses an automatic characterization and inversion method, apparatus, electronic device, and storage medium. The automatic characterization and inversion method includes: acquiring well logging information and observed seismic records; performing sparse characterization on the well logging information to obtain sub-feature libraries of well logging information at different scales; and, based on the observed seismic records and the sub-feature libraries of well logging information at different scales, inverting impedance parameters using a pre-constructed inversion objective function. This achieves inversion driven by a model and data jointly based on joint norm constraints, improving the stability of the inversion process and the resolution of the inversion results, as well as increasing the applicability and accuracy of the inversion method.

This invention relates to the field of CT detector technology, and particularly to a high spatial resolution, high-energy industrial CT linear array detector and its manufacturing method. The detector includes: a scintillation crystal linear array, a first photodiode array, and a second photodiode array; X-rays are incident from the side of the scintillation crystal linear array, and the scintillation light generated by the scintillation crystal linear array is emitted from the upper and second end faces to the first and second photodiode arrays, respectively. This invention proposes a "two-end readout + intermittent light emission" coupling scheme between the scintillation crystal array and photodiodes, achieving spatial resolution improvement through physical means, breaking through the resolution bottleneck of existing detectors, while ensuring high detection efficiency and high signal-to-noise ratio. The aim is to utilize existing standard pixel-sized photodiode arrays, through innovative structural design of the scintillation crystal linear array, to achieve a spatial resolution more than twice that of the physical pixel size.