305 results about "Spectral curve" patented technology

The invention relates to the technical field of spectral signal processing, in particular to a high-precision spectral signal peak detection method and system. According to the technical scheme, the method comprises the following steps: denoising and smoothing spectral data; dividing a spectrum curve by adopting a dynamic partitioning strategy; generating candidate partitioning points based on various local features, and generating feature blocks through optimization processing; according to the method, the influence of feature differences of different regions on a detection algorithm is reduced through a dynamic partitioning strategy, a multi-scale morphological feature system is constructed to comprehensively describe spectral curve morphological characteristics, potential peaks are gradually mined by applying a layered iteration peak detection algorithm, and local characteristics of spectral signals are adapted by means of an adaptive parameter adjustment mechanism, so that the spectral curve morphological characteristics are accurately detected. The peak value correction and fusion mechanism is utilized to improve the quality and reliability of a detection result, spectral data with a complex structure and a high noise level can be effectively dealt with, an efficient and reliable solution is provided for the field of spectral analysis, and the accuracy and robustness of peak value detection and the conciseness and reliability of the result are improved.

The invention relates to the technical field of quality detection, in particular to a medical refrigerator body manufacturing detection system and method, and the system comprises a coating reflection detection module, a sound wave response measurement module, an optical signal fluctuation analysis module, a surface defect recognition module and a structural feature state evaluation module. According to the method, the reflection amplitude gradient change of the adjacent scanning areas is compared with the spectral curve continuity, so that the sensing precision of the surface reflection distribution continuity is improved, the recognition capability of weak anomalies is enhanced, and the internal potential defects are accurately captured in combination with the sound wave echo time and structural stability analysis of a specific part; a light intensity gradient model is constructed according to a sound wave analysis result, a splicing seam and a light intensity jump area of a sunken area are identified, abnormal forms such as scratches and bubbles are further extracted, coordinates are marked, the whole process depends on dynamic feedback and multi-signal linkage calibration, the identification precision and the coverage rate are improved, and medical box body quality global efficient detection is achieved.

The invention relates to the technical field of spectral data processing, and provides a soil heavy metal content detection and correction method based on an X-ray fluorescence spectrum, which comprises the following steps: acquiring a plurality of XRF spectral signals of a to-be-detected soil sample as an original spectral curve; obtaining the overall similarity of each original spectrum curve and the initial interference weight of each peak point in the original spectrum curves; decomposing to obtain a plurality of component signals; constructing a plurality of local windows of each component signal; obtaining a noise interference factor of each local window in each original spectrum curve; obtaining an adjustment reference coefficient of each local window in each original spectrum curve; obtaining a filtering adjustment coefficient of each wavelength in each original spectrum curve; dynamically filtering and smoothing the original spectrum curve to obtain a to-be-measured spectrum signal; and detecting the soil heavy metal content based on the to-be-detected spectral signal. The invention aims to solve the problem of heavy metal content detection deviation caused by noise in an X-ray fluorescence spectrum.

The invention discloses a method for constructing a true and adulterated wine discrimination model, and particularly relates to the technical field of wine detection. The method comprises the following steps: extracting component data, a spectrum curve, smell fingerprints and texture information of wine liquid through nondestructive testing, carrying out standardization processing and feature fusion on multi-modal data, extracting distribution similarity features and domain offset robustness features, constructing a comprehensive feature vector, and evaluating a generalization ability value of a true and adulterated wine discrimination model by utilizing a machine learning model; aiming at a low generalization ability model, adopting a transfer learning method to carry out multi-stage training and optimization, and evaluating the detection accuracy of the model on unseen brands and batches in combination with cross validation; and finally, inputting the multi-modal characteristics of the to-be-detected wine on the optimized model, and outputting a true and false wine judgment result and a confidence score. According to the invention, the stability and adaptability of the true and adulterated wine discrimination model can be significantly improved, and it is ensured that the method has high accuracy, high generalization ability and wide market applicability in detection of wine of different brands and batches.

A terahertz electromagnetic metamaterial multi-frequency wave-absorbing characteristic analysis design method based on deep learning and structural parameters comprises the steps that (1) a structural prototype and a parameterized model of a terahertz electromagnetic metamaterial are established, and an electromagnetic metamaterial data set containing terahertz multi-frequency wave-absorbing characteristic information is constructed through iterative simulation calculation; 2) segmenting the data set into a 80% training data set and a 20% test data set; 3) establishing bidirectional spectrum networks SN-1 and SN-2, sending the training data set into the SN-1 and SN-2 for training, and verifying the performance through the test data set; 4) inputting the training data set into SN-1 and SN-2 network training, and testing the training effect by using the test data set; 5) utilizing the multivariate Gaussian distribution curve to randomly generate double-peak, three-peak and four-peak spectrum curve data as a multi-peak wave absorbing curve needing to be designed; and 6) inputting spectral curve data into an SN-2 network for retrieval, reconstructing a terahertz electromagnetic metamaterial multi-frequency wave-absorbing structure in CST and performing simulation calculation, and designing an electromagnetic metamaterial structure with a multi-frequency wave-absorbing characteristic.

The invention belongs to the technical field of evaluation, development and research of oil and gas reservoirs, and discloses a T2 cutoff value calculation method based on a saturated nuclear magnetic resonance experiment. The method comprises the following steps: acquiring a saturated / bound water nuclear magnetic resonance T2 spectrum curve of a compact sandstone sample; determining a T2 cut-off value based on a saturated / bound water nuclear magnetic resonance accumulative porosity curve, and recording the T2 cut-off value as T2-Cuff; acquiring a logT2-logphi intersection curve of the saturated nuclear magnetic resonance T2 spectrum, and recording a T2 value corresponding to an inflection point of the curve as T2-Log; and determining a mathematical relationship between the T2-Log and the T2-Cuff, and establishing a T2 cut-off value calculation model based on the saturated nuclear magnetic resonance experiment. According to the method, the T2 cutoff value can be calculated only by performing the nuclear magnetic resonance experiment on the saturated water sample and performing mathematical treatment on the T2 spectrum, and the method has the advantages of being economical, simple, efficient and high in practicability.

The invention discloses a method and a system for reflecting a growth state by a spectral curve of a plant leaf under salt mist stress and a storage medium, and relates to the field of plant growth monitoring. Processing the collected leaf image, extracting a leaf main body area, removing an interference part, and matching the filtered original spectral data to obtain target spectral data of the leaf main body area; the method comprises the following steps: combining target spectral data of a leaf main body area with leaf moisture content data, constructing a spectral analysis model by adopting a partial least square method, taking the leaf moisture content as a covariable into the spectral analysis model, peeling off the interference of leaf moisture change on the spectral data, and calculating the leaf moisture content. And finally obtaining a target spectrum curve of the plant leaf under the salt mist stress. A characteristic wave band corresponding to the target spectrum curve is extracted, a plant growth index is constructed according to the characteristic wave band, and the plant growth index is used for reflecting the growth health state of the plant. The problems that traditional monitoring is high in subjectivity, insufficient in precision and low in efficiency are solved.

The invention relates to the technical field of soil detection, and particularly discloses a rapid soil heavy metal pollution detection method based on a spectrum technology, which comprises the following steps: obtaining dual-band spectral reflectivity data and temperature and humidity parameters of to-be-detected soil, carrying out spectral characteristic compensation, generating a spectrum curve, and obtaining environment compensation parameters; extracting spectral characteristic parameters from the spectral curve, combining the spectral characteristic parameters to construct a multi-dimensional spectral characteristic matrix, and inputting the multi-dimensional spectral characteristic matrix into a CNN-RF hybrid model for soil heavy metal pollution detection to obtain a heavy metal pollution detection result. According to the method, in-situ lossless rapid detection of the soil heavy metals is achieved through the dual-band spectrum and the environment dynamic compensation technology, interference of environmental factors on spectral measurement is effectively reduced, the detection efficiency is improved, the pollution condition of the heavy metals in the soil is accurately recognized through the CNN-RF mixed model, the reliability and precision of soil heavy metal pollution are improved, and the method is suitable for popularization and application. Corresponding treatment measures can be taken in time, and the safety of the soil environment is guaranteed.

The invention provides a rare earth element content quantitative estimation method and device, equipment and a storage medium. The method relates to the technical field of hyperspectrum, machine learning and quantitative estimation. The method comprises the following steps: acquiring hyperspectral data of each sample, eliminating gross error points, performing data dimension reduction and K-means clustering analysis to obtain effective spectral data of each sample, performing denoising processing, averaging to obtain an average spectral curve, and extracting spectral characteristics; analyzing the correlation between the characteristic data of each wave band and the element content by utilizing Pearson correlation, selecting a high-correlation wave band range, carrying out importance test on the high-correlation wave band range by utilizing a random forest, and screening out a characteristic wave band corresponding to each element; and constructing a data set by using the characteristic wave bands corresponding to the elements and the chemical test contents of the elements, and training and evaluating the machine learning model based on the data set to obtain a quantitative estimation model. The rare earth element content can be quickly scanned and evaluated in time.

The invention discloses a hyperspectral satellite target rock and mineral identification method and system based on spectral characteristic parameters. The method comprises the following steps: S1, collecting and preprocessing a standard rock and mineral reflection spectrum curve; s2, collecting a regional hyperspectral satellite image and carrying out image preprocessing; s3, identifying diagnostic spectral reflection and absorption characteristics of the target rock mine based on the standard curve; s4, screening and extracting a diagnostic spectral range of the hyperspectral satellite image; s5, calculating diagnostic spectral characteristic parameters of the hyperspectral image target rock mine; and S6, on the basis of the S5, processing the spectral characteristic parameters obtained by calculation, and realizing target rock and mine identification based on multi-dimensional diagnosis spectral characteristics. According to the method, the application potential of the hyperspectral satellite image in the field of remote sensing geology is effectively improved and the scientificity and the accuracy of the hyperspectral satellite image in large-range rock and ore recognition drawing application are improved through the target rock and ore standard spectral curve characteristic analysis, hyperspectral satellite image processing and spectral characteristic parameter estimation methods.

The invention discloses a method for quantifying the fusion degree of new asphalt and old asphalt, and relates to the technical field of road thermal regeneration. An asphalt film sample on the surface of an RAP material is subjected to layered extraction, and a uniform film is formed in combination with a potassium bromide sheet technology; secondly, acquiring spectral characteristic data and characteristic peak values of each layer of sample by adopting a Fourier transform infrared spectroscopy FTIR technology, normalizing the data by adopting methods such as baseline correction and standardization, extracting characteristic peak positions and intensity by utilizing a first-order derivative, and constructing a standardized curve; a dynamic time warping (DTW) algorithm is adopted to quantify the spectrum curve matching degree of the new asphalt and the old asphalt, a fusion similarity index FSI is defined, and the similarity of the new asphalt and the old asphalt in chemical component distribution is effectively quantified; meanwhile, a microscopic infrared spectrum technology is adopted to further perform local scanning on RAP particle surface spectrum distribution, a fusion heat map is generated in combination with an FSI result, and the local fusion difference of the new asphalt and the old asphalt is visually displayed.

The invention provides a spectral image demodulation method, spectral image demodulation equipment and a computer readable storage medium. The spectral image demodulation method comprises the following steps: S1, acquiring a first spectral image; s2, processing the first spectral image to obtain a second spectral image; s3, performing spectrum correlation region division on the second spectrum image to obtain a plurality of spectrum correlation regions; s4, obtaining a plurality of spectrum curves according to the plurality of spectrum correlation areas; and S5, obtaining a spectral image of a target wave band according to the target wave band, the plurality of spectral curves and the first spectral image. The spectral image demodulation method provided by the invention can break through the limitation of the prior art, can realize spectral image demodulation and spectral image inversion, has the advantages of high precision, high accuracy, high reliability and the like, and can be widely applied to various fields and various scenes.

The invention relates to the technical field of heavy metal detection, in particular to a method and a system for rapidly detecting heavy metals in traditional Chinese medicinal materials, and the method comprises the following steps: respectively collecting X-ray fluorescence spectrum curves of to-be-detected honeysuckle and a honeysuckle sample, and sequentially recording the X-ray fluorescence spectrum curves as a to-be-detected spectrum curve and a standard spectrum curve; determining a peak position difference and a contour difference of each characteristic peak in the spectrum curve to be measured; obtaining the peak shape variation degree of each characteristic peak in the spectrum curve to be measured; determining the peak bottom energy difference of each characteristic peak in the spectrum curve to be measured; obtaining an abnormal correction coefficient of each characteristic peak in the spectrum curve to be measured; obtaining the difference between the measured value of the heavy metal content in the honeysuckle with the known heavy metal content and the real value, recording the difference as the measurement difference, and determining the actual content of each heavy metal element in the honeysuckle to be measured based on the correlation between the measurement difference and the abnormal correction coefficient. Therefore, the accuracy of measuring the content of the heavy metal elements in the to-be-measured honeysuckle is improved.

The invention relates to the field of remote sensing image processing, in particular to a farmland soil type classification-oriented remote sensing image data acquisition method, which comprises the following steps of: regularly acquiring remote sensing image data of a target area according to a preset period; obtaining a distance measure between any two pixel points in each remote sensing image, and segmenting each remote sensing image to obtain a ground feature region; analyzing the trend and the amplitude of each pixel point on the spectral curve, and determining the red edge expression degree of each pixel point; determining the vegetation growth performance degree of each pixel point; determining the farmland soil segmentation quality of each remote sensing image; and screening a to-be-classified image from all remote sensing images of the target area, screening a farmland soil area to construct a bare soil mask image, and performing farmland soil classification on a ground feature area in the to-be-classified image in combination with a neural network. According to the method, the farmland soil area in the remote sensing image is accurately divided, and the accuracy of farmland soil type classification is improved.

The invention relates to the technical field of optical systems, and provides a full-automatic collimation method and system for an optical system, a medium and equipment, and the method comprises the steps: extracting pulse features of a pulse sequence, calculating pulse interval jitter, an energy fluctuation rate and a deviation between a pulse repetition rate and a theoretical value, and judging a mode locking state; for the output laser, reconstructing a spectral curve of the output laser, and extracting spectral features; obtaining a mode locking stability score through a deep learning model; the basic parameters of the laser cavity are optimized by maximizing the power and the collimation degree of the output laser; basic parameters of the laser cavity are used as priori, and parameters of all optical elements of the resonant cavity are adjusted by maximizing a mode locking stability score and a target pulse width score; and on the premise of the adjusted parameters of the optical elements of the resonant cavity, the pose of the lens is adjusted by maximizing the power and the collimation of the output laser. High-efficiency and progressive optimization of optical system parameters is realized, and the collimation degree is improved.

The invention relates to the technical field of carbon content measurement, in particular to a soil organic carbon content measurement method and system.The method comprises the steps that soil to be measured and a preset number of soil samples with known soil PH and soil organic carbon content are obtained; collecting a spectral curve of each position in each soil sample; recording each wavelength on the spectrum curve as each sampling point; determining a spectrum mutation interference coefficient of each sampling point; acquiring a spectral anomaly correction coefficient of each sampling point on each spectral curve; performing weighted fusion on all the spectral curves of each soil sample based on the spectral anomaly correction coefficient to obtain a corrected spectral curve of each soil sample; forming a training set by utilizing the correction spectrum curves, the organic carbon contents and the PH values of all the soil samples, and training the prediction model; and measuring the organic carbon content of the to-be-measured soil through the trained prediction model. Therefore, the accuracy of measuring the organic carbon content of the soil is improved.

The invention relates to the technical field of microbiological detection, in particular to a public area microbiological pollution risk assessment system based on spectral analysis, which comprises a spectrum acquisition module used for acquiring a Raman spectrum curve of microorganisms contained in the air of a public area; the spectrum correction module is used for carrying out baseline correction on the Raman spectrum curve, specifically, baseline values of the Raman spectrum curve at all wavelengths are determined, and the corrected Raman spectrum curve is obtained; the feature analysis module is used for identifying the types of the microorganisms, specifically, the feature evaluation value of each wave crest is calculated, all the wave crests are distinguished, and a feature main peak and a feature shoulder peak are obtained; extracting characteristic peaks contained in the acromion; identifying the types of microorganisms in the air; and the risk assessment module assesses the microbial pollution risk of the public area. According to the method, misjudgment and missing detection caused by characteristic peak overlapping are reduced, and the accuracy of microorganism variety identification is improved, so that the pollution risk of microorganisms in a public area is accurately evaluated.

The invention discloses a construction method of a fresh tobacco leaf harvesting maturity discrimination system. Comprising the following steps: a, collecting fresh tobacco leaf samples with different maturity degrees; the method comprises the following steps: acquiring a hyperspectral image of a tobacco leaf sample, performing metabolite determination on the tobacco leaf sample, expanding tobacco leaf metabolite test data to form a tobacco leaf metabolite test data set, combining the hyperspectral image set with the tobacco leaf metabolite test data set, and constructing a tobacco leaf maturity identification model and a tobacco leaf metabolite prediction model by adopting a deep learning method. And a fresh tobacco leaf harvesting maturity discrimination system is designed by using PyQt5. According to the method, a fresh tobacco leaf maturity identification system is formed by using the tobacco leaf maturity identification model, the tobacco leaf metabolite prediction model and the relevance of the two models, so that the models required by the system can be accurately established by using fewer tobacco leaf samples, and the final identification system can be used for identifying the maturity of the fresh tobacco leaves by inputting a spectral curve of unknown tobacco leaves. And acquiring comprehensive data about tobacco leaf maturity, metabolites and quality information.

The invention relates to the technical field of chemical or physical testing, in particular to a differential absorption spectrum-based respiratory inflammation marker detection system, which comprises the following steps of: determining an effective spectrum interval of an original spectrum curve by using a marker wave band of a marker to be detected; splitting the effective spectral interval by using a wavelet basis to obtain a plurality of spectral decomposition hierarchies; determining the separation effectiveness of the spectral decomposition hierarchy waveform by using the difference of adjacent data points in the spectral decomposition hierarchy waveform and the difference of frequency peak values of adjacent spectral decomposition hierarchies; determining an effective hierarchy in the spectral decomposition hierarchy by using a differential value of the separation effectiveness sequence; and determining a target filtering frequency threshold value by utilizing respective separation effectiveness of the effective hierarchy and the current spectral decomposition hierarchy corresponding to the effective hierarchy, and further filtering the spectral decomposition hierarchy to obtain the content of the to-be-detected marker. According to the technical scheme, the accuracy of detecting multiple gas markers caused by inflammation of the respiratory tract is improved.

The invention relates to the technical field of blood concentration detection by optical means, in particular to a clinical blood concentration detection method. The method comprises the following steps: acquiring the delay responsivity of spectral data to drug concentration and the change complexity of other wavebands except a preset characteristic waveband; further obtaining the interfered change credibility of the drug concentration; according to the morphological characteristics of the spectrum curve corresponding to each drug concentration in the preset characteristic wave band and the distribution difference characteristics of the data points between the corresponding spectrum curve and the preset spectrum curve in the preset characteristic wave band, obtaining the spectrum shape distortion of the spectrum curve corresponding to each drug concentration in the preset characteristic wave band; obtaining the processing beneficial confidence of the spectral data; filtering the spectral data to obtain filtered spectral data; and detecting the blood concentration. According to the method, interference components in spectrum detection are removed, so that the quality of spectrum data and the detection precision of the blood concentration are improved.

The invention discloses a soil profile occurrence layer division method based on an indoor imaging spectrum technology. The method comprises the following steps: S1, acquiring a complete soil profile hyperspectral image by adopting a method based on pixel embedding; s2, carrying out geometric correction on the image, removing a wooden frame, and reserving a soil profile original image with a fixed width; s3, using a random forest algorithm to remove invalid pixels such as cracks and shadows; s4, carrying out smooth denoising processing on the reserved spectral data; s5, extracting a row average spectrum curve, and obtaining a fixed-depth row average spectrum curve as spectrum data of the layered strip soil sample; s6, generating a spectrum curved surface; and S7, dividing a generation layer based on the semantic segmentation network model. According to the method, accurate division of various soil profile generation layers can be realized, and powerful technical support is provided for large-scale soil classification and sustainable management of soil resources.

The invention relates to the technical field of spectral analysis, in particular to a crop growth state real-time monitoring method, which comprises the following steps of: acquiring hyperspectral data and temperature data consisting of all wavebands and reflection intensities at each sampling moment; acquiring a spectrum curve at each sampling moment and a progressive line of each wave band; calculating an asymptotic line noise evaluation coefficient, and further obtaining an asymptotic line difference coefficient; calculating a spectrum progressive temperature drift coefficient at each sampling moment, forming a temperature drift verification sequence at each sampling moment according to the hyperspectral data and the temperature data, and further obtaining a temperature drift verification factor at each sampling moment; calculating a noise temperature drift verification coefficient, and obtaining a self-adaptive wavelet threshold value of each decomposition layer number at each sampling moment; and denoising the hyperspectral data at each sampling moment according to a self-adaptive wavelet threshold to obtain the growth condition of the crops. The invention aims to solve the problems of incomplete noise reduction and excessive smoothness of signals caused by a fixed threshold in traditional wavelet threshold filtering.

The invention discloses a solid waste classification method based on a spectrometer, and relates to the technical field of solid waste analysis, and the method comprises the following steps: S1, preliminary classification: obtaining a sample of unknown solid waste, and observing the characteristics of the sample to obtain characteristic information; s2, selecting a spectrograph: selecting the spectrograph according to the feature information; analyzing the sample by using a spectrograph to obtain a spectrogram; the spectrogram comprises a spectrum curve; s3, spectrogram preprocessing: carrying out normalization processing on the data of the spectrum curve; s4, similarity analysis: for the spectrum curves of the multiple samples, randomly selecting two samples for comparison to obtain a similarity value; if the similarity value is within a preset threshold range, the two samples to be compared belong to the same kind of solid waste. The method is low in operation cost and low in professional degree requirement on workers, and can realize rapid and efficient analysis and classification of solids.

The invention discloses an unsupervised multivariate feature extraction method for hyperspectral image classification, and relates to the technical field of spatial-spectral joint feature extraction, and the method comprises the steps: carrying out the smooth denoising of a spectral curve through employing an MSMC, and capturing multi-scale spectral features through employing a cascade residual mode; gaussian smooth features of the hyperspectral remote sensing image are obtained by using MRCD, deep and shallow layer spatial features are extracted, and the deep and shallow layer spatial features are fused to obtain final spatial features; using M3FuNet to fuse the multi-scale spectral features and the final spatial features to obtain spatial-spectral joint features; the three features are input into a classifier to be summarized for decision fusion, and a spectrum decision, a space decision and a fusion decision are obtained; and obtaining a final classification decision of a spectrum decision, a space decision and a fusion decision through a majority voting mode. According to the method, spectral and spatial multiple scales of the hyperspectral remote sensing image are mined, characterization contributions of spectral features and spatial features are reserved, and single-class feature confrontation is relieved.

The invention discloses a dimension reduction processing method for space-time joint foresight imaging of an airborne array radar, and the method comprises the steps: constructing a foresight imaging geometric model of the airborne array radar, receiving a distance, direction and channel three-dimensional echo matrix, converting data in a single coherent processing time interval into a frequency domain for the echo matrix, and compensating a Doppler phase, and then, truncating the effective information of the frequency spectrum to obtain snapshot data after dimension reduction. And then applying the newly obtained space-time snapshots to space-time adaptive iterative super-resolution imaging, estimating space-time super-resolution spectrum curves corresponding to the space-time snapshots of the distance-azimuth units, and performing spectrum accumulation on the super-resolution spectrum curves of the distance-azimuth units according to the motion speed of an aerial carrier and antenna scanning parameters to finally obtain a two-dimensional forward view image. According to the method, effective echo information is intercepted in the Doppler domain, space-time data are effectively compressed, the operation complexity is remarkably reduced, and meanwhile the imaging quality is kept.

The invention relates to an unmanned aerial vehicle hyperspectrum-based rice bacterial blight severity inversion method, which comprises the following steps of: analyzing a spectrum curve of different types of sample point reflectivity along with wavelength change, determining sensitive wavebands of different bacterial blight harm degrees, and selecting the sensitive wavebands to construct a bacterial blight harm rice inversion index NBLBRI; and determining NBLBRI thresholds of inversion of the rice harmed by the four different bacterial blight diseases. The method has the beneficial effects that the sensitive wavebands with different bacterial blight damage degrees are determined by analyzing the spectral curve difference of the training samples, and the sensitive wavebands are selected to construct the bacterial blight damage rice inversion index NBLBRI; analyzing differences between statistical characteristics of NBLBRI between rice with different disease damage degrees and healthy rice, and determining threshold intervals of NBLBRI for estimating the damage degrees of different types of rice diseases so as to accurately extract areas with different disease degrees; and a flexible and wide application prospect is provided for agricultural disease management.

The invention relates to a wave band self-adaptive hyperspectral image compression method based on semantic guidance. According to the method, a compression task and a classification task are unified under the same framework, and a natural relation between a spectral curve and a semantic category is utilized to apply category constraint to a spectral structure, guide generation of spectral information of a reconstructed image, and relieve adverse effects of spectral structure mutation on a compression process and subsequent hyperspectral image analysis and interpretation. According to the method, semantic information of ground features under a high compression ratio is effectively reserved, and the adverse effect of lossy compression on downstream task analysis and interpretation is relieved while bandwidth and storage resources are saved. Besides, the compression framework can process the hyperspectral data of any wave band, and does not need to train an independent model according to the hyperspectral images of different wave bands, so that the storage cost of the compression model is reduced.

The invention discloses a material classification method and device based on an ultraviolet electron spectrum, a medium and a product, and relates to the technical field of material surface science and data analysis. The method comprises the following steps: acquiring ultraviolet light electron energy spectrum data of a target material to be analyzed, and extracting a first spectrum curve from the energy spectrum data; obtaining a conductor characteristic identification result of the target material according to the first spectrum curve; under the condition of determining that the target material does not have the conductor characteristics based on the conductor characteristic identification result, extracting a second spectrum curve in a preset valence band top energy interval extending towards the negative energy direction by taking the Fermi level as a starting point from the ultraviolet light electron spectrum data, and performing multi-dimensional characteristic fusion on the second spectrum curve, generating a semiconductor characteristic identification result of the target material based on the fusion result; and determining a classification result of the target material according to the semiconductor characteristic identification result. According to the scheme, the conductive attributes of the target material can be quickly and accurately classified.

The invention discloses a high-quality tobacco leaf automatic slitting method which comprises the following steps: data acquisition: scanning tobacco leaves through a hyperspectral camera system to obtain hyperspectral image data; region extraction: positioning and extracting main veins, branch veins and tobacco lamina parts by utilizing computer vision; extracting spectral information to obtain spectral curves in different wavelength ranges; analyzing chemical components; analyzing moisture characteristics; mechanical arm positioning is conducted, specifically, the to-be-cut tobacco leaf part is accurately positioned; cutting operation: cutting the main vein, the branch vein and the high-quality part by using cutting equipment; and quality selection: evaluating the quality of the cut tobacco leaf part, and selecting a high-quality tobacco leaf part suitable for high-grade cigarette development. According to the automatic slitting method for the high-quality tobacco leaves, a hyperspectral camera mode in machine vision and automatic cutting equipment are adopted, and a precise positioning technology is combined, so that all parts of the tobacco leaves are rapidly and accurately cut, and the high-quality tobacco leaves are obtained to be used for manufacturing high-grade cigarettes.

The invention discloses a hyperspectral information-based target object classification method, apparatus and device, and a medium. The method comprises the steps of obtaining hyperspectral information of sparse sampling points of a scene where a to-be-classified target object is located; obtaining visible light information of a scene where the to-be-classified target object is located; performing affine transformation alignment on the hyperspectral information and the visible light information; extracting a contour region of a to-be-classified target object from the visible light information by using an instance segmentation model; acquiring a spectral curve of each sparse sampling point in the contour area from the hyperspectral information according to the contour area of the target object and an affine transformation alignment result; and obtaining spectral features of the target object according to the spectral curves of all the sparse sampling points in the contour region, and performing feature classification on the target object by using a classification model based on the spectral features. The target objects can be quickly classified, the processing time is saved, and the execution efficiency is improved.