88 results about "Spectral similarity" patented technology

A source signal (e.g. a speech sample) is processed or transmitted by a speech coder 1 and converted into a reception signal (coded speech signal). The source and reception signals are separately subjected to preprocessing 2 and psychoacoustic modelling 3. This is followed by a distance calculation 4, which assesses the similarity of the signals. Lastly, an MOS calculation is carried out in order to obtain a result comparable with human evaluation. According to the invention, in order to assess the transmission quality a spectral similarity value is determined which is based on calculation of the covariance of the spectra of the source signal and reception signal and division of the covariance by the standard deviations of the two said spectra.The method makes it possible to obtain an objective assessment (speech quality prediction) while taking the human auditory process into account.

A method for the detection of noise and speech segments in a digital audio input signal, the input signal being divided into a plurality of frames including a first stage in which a first classification of a frame as noise is performed if the mean energy value for this frame and the previous N frames is not greater than a first energy threshold, N>1, a second stage in which for each frame that has not been classified as noise in the first stage it is decided if the frame is classified as noise or as speech based on combining at least a first criterion of spectral similarity of the frame with acoustic noise and speech models, a second criterion of analysis of the energy of the frame and a third criterion of duration, and of using a state machine for detecting the beginning of a segment as an accumulation of a determined number of consecutive frames with acoustic similarity greater than a first threshold and for detecting the end of the segment; a third stage in which the classification as speech or as noise of the signal frames carried out in the second stage is reviewed using criteria of duration.

The signature verification methods and devices disclosed herein can be used to verify signatures signed on electronic key pads and other input devices, such as signature pens. Many different aspects of a dynamic signature can be measured in an attempt to verify the signature, including, but not limited to spatial measurements, measurements over time, and frequency. These measurements can be of points on a signature, but they can also be pressure, velocity, and acceleration, to name just a few. These different aspects can then be analyzed using, for example, time series, and spectral similarities. Further, the spectral similarities can be analyzed using wavelet-transforms. In another embodiment, these analysis systems and methods can be applied to written signatures as well as dynamic written signatures.

The invention provides an object-oriented timing sequence remote sensing image cloud coverage area automatic detection method, comprising six steps of data preparation, image registering, multiscale image segmentation, radiation difference elimination, spectrum similarity calculation, threshold segmentation and detection result generation. The method employs an object-oriented interpretation method, and can effectively perform pattern classification on a cloud coverage area, fully utilize timing sequence information in timing sequence images, and effectively utilize the multi-feature advantage among the timing sequence images; in a timing sequence remote sensing image cloud coverage area detection process, the method employs a multi-feature fusion strategy, fully utilizes timing sequence and spectrum information, and makes detection results more reliable and stable.

The present invention relates to a method for the detection of noise and speech segments in a digital audio input signal, said input signal being divided into a plurality of frames comprising: - a first stage (10) in which a first classification of a frame as noise is performed if the mean energy value for this frame and the previous N frames is not greater than a first energy threshold, N>1; - a second stage (20) in which for each frame that has not been classified as noise in the first stage it is decided if said frame is classified as noise or as speech based on combining at least a first criterion of spectral similarity of the frame with acoustic noise and speech models, a second criterion of analysis of the energy of the frame and a third criterion of duration; and of using a state machine for detecting the beginning of a segment as an accumulation of a determined number of consecutive frames with acoustic similarity greater than a first threshold and for detecting the end of said segment; - a third stage (30) in which the classification as speech or as noise of the signal frames carried out in the second stage is reviewed using criteria of duration.

The invention relates to a high-spatial resolution remote sensing image-based road region building change extraction method and device. According to the method and device, an object-oriented image processing strategy is adopted, the spectral information and spatial information (including structural indexes and spatial relationships) of images are comprehensively utilized, and a single-class classification method is adopted. In order to avoid interference (spectral similarity) on extracted results caused by ground feature classes except road regions, it is required that existing road information such as an existing road vector diagram, is provided in advance; the existing road vector diagram is adopted to extract a road region range; and newly increased buildings are extracted within the road region range. With the high-spatial resolution remote sensing image-based road area building change extraction method and device provided by the invention, time for a traditional method to obtain newly increased buildings such as illegal buildings by using image visual interpretation can be greatly decreased, efficiency can be improved, and human resources can be saved. The method and device can be used for road maintenance and monitoring business operation systems.

The invention relates to a method for maintaining a cigarette blending formulation, and belongs to the field of tobacco leaf quality control. The method comprises the following steps of sample collection, chemical component detection, sensory evaluation, spectral preprocessing, feature spectrum extraction, and spectral similarity matrix construction for substitute tobacco leaf search. According tothe method for maintaining the cigarette blending formulation, the tobacco leaves with the similar quality can be quickly searched for in the formulation maintenance process; the method is short in period, high in efficiency and low in cost; the problems of stable cigarette quality, tobacco leaf interior quality, availability and the like, and a cigarette formulation technology are comprehensively considered; and the requirements of rapid evaluation of raw materials and formulation maintenance can be met.

The invention discloses a method for applying the near infrared spectral analysis technology for undamaged identification of the truth of tobacco. The method is as follows: firstly, spectrum data of true tobacco is acquired and stored into a computer; secondly, the method for qualitative analysis of the similarity matching in TQ analyst7.1 software is applied to establish a similarity matching model of the tobacco of the brand; thirdly, a similarity matching critical value of the tobacco of the brand is obtained after calculation. When the tobacco is identified, the same method is applied for acquiring spectrum data of the tobacco to be identified and calculating a spectrum similarity matching value of the tobacco to be identified; when the similarity matching value of the tobacco to be identified is more than or equal to the similarity matching critical value of the tobacco of the brand, the tobacco to be identified is identified to be true tobacco with the same brand, or else, the tobacco is false tobacco. The amount of samples required by the identification method is small; the identification process does not need pretreatment and sampling procedure, does not damage cigarettes and is simple, thereby the method is extraordinarily suitable for being promoted in the tobacco industry and the tobacco law enforcement and supervision department and has good market application prospect.

A large-batch automatic hyperspectral remote sensing mineral mapping method comprises seven major steps of: step 1, reading in spectroscopic data; step 2, intercepting a spectral band; step 3, performing continuum removal of the spectrum; step 4, calculating the absorption depth of the image spectrum; step 5, standardizing the image spectrum; step 6, calculating the similarity of the image spectrum to a reference spectrum; and step 7, extracting mineral abnormity according to the calculation results of the step 5 and the step 6. The method does not need to extract the reference spectrum from the hyperspectral remote sensing and perform artificial interpretation; the same reference spectrum is used for the processing of all data so that large-batch and automatic processing of data can be realized conveniently; and the spectrum similarity and the characteristic parameters are comprehensively utilized for mineral abnormity extraction, and the reliability and the accuracy of mineral recognition are improved. The large-batch automatic hyperspectral remote sensing mineral mapping method has practicable value and wide application prospect in the field of hyperspectral remote sensing geological exploration engineering application.

The invention discloses a remote sensing image missing data restoration method based on multi-image local interpolation. According to the method, based on the principle that satellite remote sensing image pixels and identical ground object pixels within a local range around the satellite remote sensing image pixels have similar spectral features, a missing data pixel is searched for in a to-be-restored image, and windows with a preset maximum search size are constructed respectively with the location of the pixel and the location of a corresponding pixel in a non-missing-data image being center points; the spectral features of pixels, corresponding to pixels with values in the window of the to-be-restored image, in the non-missing-data image are classified in the non-missing-data image to determine the location of the pixel in a minimum spectral difference with the center pixel; the spectral value of the pixel at the same location in the window of the to-be-restored image is used to replace a missing spectral value of the center point pixel in the window of the to-be-restored image; and the non-missing-data image is adopted to determine whether the missing data pixel in the to-be-restored image is the same as a certain non-missing-data pixel in a local window or has a maximum spectral similarity, and the spectral value of the non-missing-data pixel is used to fill up the spectral value of the missing data pixel. In this way, complicated calculation such as regression statistical analysis, image segmentation and geostatistical analysis on the to-be-restored image is avoided, and the method has the advantages that calculation efficiency is moderate, restoration precision is high, and the method is fast and easy to realize.

The invention provides a method and system for identifying consistency of articles by using Raman spectroscopy. The method comprises the following steps: 1) acquiring the spectral data of different samples by using a spectrometer; 2) preprocessing the acquired spectral data; 3) setting a spectral interval and separately calculating similarity of local spectral intervals at different wave number positions; 4) acquiring a curve representing changes of local spectral similarity with wave numbers of Raman spectra; and 5) determining consistency of the articles according to the curve.

The invention discloses a compact-dictionary sparse-representation-based hyperspectral-remote-sensing-image classification method, which effectively improves class interference solving time-consumingproblems caused by using fixed full-class dictionaries by existing residual-based sparse-representation classification method. In a classification process of the method, neighborhood label informationand a spectral similarity measurement method are utilized to construct a compact dictionary of self-adaptive classes for each test sample, limited local label information is transmitted to a wider area in a hyperspectral remote-sensing image through a space position expansion strategy, thus space information of the hyperspectral remote-sensing image is more fully explored, a scale of the dictionary and a classification decision range are reduced at the same time, solving time of sparse coefficients is enabled to be greatly decreased, and a speed and an accuracy rate of classification are significantly increased. The method can be used in the fields of agriculture, environment monitoring, military defense and the like.

The invention discloses an unsupervised hyperspectral image implicit low-rank projection learning feature extraction method, and aims to provide an unsupervised hyperspectral feature extraction methodcapable of realizing rapidness and high robustness. The method is realized through the following technical scheme: firstly, dividing input hyperspectral image data into a training set and a test setin proportion; designing a robustness weight function, calculating the spectral similarity between every two training set samples, and constructing a spectral constraint matrix and a graph regularization constraint according to the training set; approximately decomposing row representation coefficients of the hidden low-rank representation model; constructing an implicit low-rank projection learning model by combining the spectral constraint matrix and the image regularization constraint; and optimizing and solving the hidden low-rank projection learning model by adopting an alternating iterative multiplier method, obtaining a low-dimensional projection matrix, outputting the categories of all test set samples, taking the low-dimensional features of the training set as the training samplesof the support vector machine, classifying the low-dimensional features of the test set, and evaluating the feature extraction performance according to the quality of a classification result.

The invention discloses a MeanShift based distance measurement optimization method in high-resolution remote sensing image segmentation during area merging. According to the MeanShift based high-resolution remote sensing image segmentation distance measurement optimization method, remote sending image data characteristics are fully considered, traditional Euclidean distance is replaced by spectrum matching distance measurement, specifically, spectrum angle matching measurement, spectral similarity measurement and Kernel mapping spectrum matching measurement, and the segmentation result is accurate.

The present invention provides a method and system for pattern recognition and processing. Information representative of physical characteristics or representations of physical characteristics is transformed into a Fourier series in Fourier space within an input context of the physical characteristics that is encoded in time as delays corresponding to modulation of the Fourier series at corresponding frequencies. Associations are formed between Fourier series by filtering the Fourier series and by using a spectral similarity between the filtered Fourier series to determine the association based on Poissonian probability. The associated Fourier series are added to form strings of Fourier series. Each string is ordered by filtering it with multiple selected filters to form multiple time order formatted subset Fourier series, and by establishing the order through associations with one or more initially ordered strings to form an ordered string. Associations are formed between the ordered strings to form complex ordered strings that relate similar items of interest. The components of the invention are active based on probability using weighting factors based on activation rates.

The invention relates to the field of analytical chemistry, particularly to an analysis identification method of ophiocordyceps sinensis through an infrared spectrum, and aims to solve the technical problem that the authenticity of the existing ophiocordyceps sinensis is difficult to distinguish. In order to solve the above technical problem, the invention provides a technical scheme to rapidly and accurately distinguish ophiocordyceps sinensis. The analysis identification method through the infrared spectrum comprises the following steps: a, drawing a standard spectrum of ophiocordyceps sinensis; b, collecting an essential characteristic peak and a common characteristic peak of an infrared spectrum of ophiocordyceps sinensis; c, distinguishing the essential characteristic peak and similarity of the infrared spectrum of measured ophiocordyceps sinensis. The invention provides a novel method for rapidly and accurately distinguishing ophiocordyceps sinensis.

The invention discloses a calibration set and validation set selection method based on spectral similarity and a modeling method. The calibration set and validation set selection method comprises thefollowing steps of performing near infrared spectrum determination on an original sample to obtain an original sample spectrum matrix; randomly extracting m samples as an independent testing set; foreach sample in the independent testing set, respectively computing the spectral similarity between the sample and each remaining sample in the original sample in order to obtain g samples with the highest similarity to be written into the validation set; and for each sample in the independent validation set, respectively computing the spectral similarity between the sample and each remaining sample in the original sample in order to obtain n samples with the highest similarity to be written into the calibration set. According to the validation set and the calibration set which are selected based on the method of the invention, an obtained model can more accurately predict an unknown model.

The invention discloses a method for monitoring the irrigation area of an irrigation area based on high-resolution satellite data. The method comprises the following steps: step 1, obtaining and preprocessing satellite remote sensing data; step 2, actually measuring sample point data; step 3, extracting an end member spectrum according to the real sample point data obtained in the step 2; step 4, calculating spectral similarity by adopting a statistical algorithm and a spectral waveform feature algorithm in a spectral matching method, and quantitatively analyzing the matching degree of the end member spectrum of the main crops in the research area and the target spectrum through three indexes; and step 5, calculating an SSV segmentation threshold value by adopting an OTSU adaptive threshold value algorithm to judge whether the irrigation area is an irrigation area, and if the SSV segmentation threshold value is smaller than the SSV segmentation threshold value, identifying the irrigation area spatial distribution condition of the research area, and finally obtaining the irrigation area range. The method is suitable for high-resolution satellite remote sensing data, meets the requirement for extracting small plot irrigation information, and can improve the accuracy of an irrigation area monitoring result.

The invention discloses an engine lubricating oil recognition algorithm based on an infrared spectrum. The engine lubricating oil recognition algorithm is mainly realized by two steps of obtaining a new spectral similarity evaluation indicator sim and applying the new spectral similarity evaluation indicator sim to distinguish lubricating oil. Compared with the prior art, the engine lubricating oil recognition algorithm has the most obvious characteristics that a variable importance degree is introduced into a fingerprint spectral similarity evaluation indicator so as to overcome the disadvantages of an existing algorithm that a small change spectrum can not be effectively distinguished, caused by the fact that the variable importance degree is not considered; furthermore, the recognition capability of the similarity evaluation indicator on the small change spectrum is improved.

The invention aims to provide an alternative method for a tobacco leaf and a cigarette leaf group formula based on a near infrared spectrum. The alternative method eliminates interference factors suchas base lines and scattering in the spectrum through adopting a locally correcting spectral preprocessing method and an integrated spectral similarity algorithm, realizes the conversion from spectralsimilarity to tobacco leaf similarity, realizes the tobacco leaf alternative method in the auxiliary cigarette leaf group formula maintenance process, reduces the smoke panel test amount of the formula maintenance personnel, and avoids the subjectivity and instability in the sensory smoke panel test process.

The present invention relates to a light source quality evaluating method by using spectral resemblance with respect to the blackbody radiation, which mainly comprises 5 method steps. This method is used for evaluating the quality of light based on physiological perception of human. In evaluating operation, the method firstly transfers a power spectrum of a light source to a luminance spectrum of light source through a luminosity function. Next, the method compares the luminance spectrum of light source with a luminance spectrum of the blackbody radiation thereof. Therefore, an index of spectral resemblance with respect to the black body radiation (SRBR) would be calculated and then obtained, such that the SRBR can be used for evaluating the quality of the light source. Moreover, comparing to conventional color rendering index (CRI), SRBR is a better light source quality evaluating method because of having fairness and consistency.

The invention provides a spectral similarity calculation method which can comprehensively evaluate the similarity between a sampling spectrum and a standard spectrum so as not only to reflect the similarity of the common peaks between a to-be-evaluated sampling spectrum and a standard spectrum, but also to consider the influence of the common peak number on the magnitude of the similarity and reflect the similarity degree between the spectrums generally. The similarity degree can sensitively reflect the differences of the sampling spectrum and the standard spectrum in quantity and quality. Therefore, the to-be-evaluated sample could be evaluated reliably.

The invention relates to the field of analytical chemistry, specifically to an infrared spectroscopic analysis and identification method for the polypide of Cordyceps sinensis. The objective of the invention is to overcome the technical problem that whether the polypide of Cordyceps sinensis is true or false is difficult to determine at present. Thus, the invention provides the method for rapid and accurate discrimination of the polypide of Cordyceps sinensis. The method comprises the following steps: a, drafting the standard spectrum of the polypide of Cordyceps sinensis; b, acquiring necessary infrared spectroscopic characteristic peaks and common infrared spectroscopic characteristic peaks of the polypide of Cordyceps sinensis; and c, comparing infrared spectroscopic characteristic peaks and infrared spectroscopic similarity of a detected sample of the polypide of Cordyceps sinensis. The invention provides the novel method for rapid and accurate discrimination of the polypide of Cordyceps sinensis.

The present application relates to the field of substance detection, and in particular to a method, device and terminal for substance detection and a readable storage medium. The method for substancedetection comprises: obtaining the spectral data of a substance to be tested; determining a matching result of the substance to be tested and matching samples in a database according to the spectral similarity of the spectral data of the substance to be tested and the spectral data of each sample in the database, wherein the database includes the spectral data and the attribute information of a known sample, as well as the spectral data and the attribute information of a unknown sample; and determining the detection result of the substance to be tested according to the matching result. The method reduces the probability of misidentification of the substance to be tested.

The invention discloses a terminator time sea fog detection algorithm based on Himawari-8 satellite data, and relates to the technical field of new generation static satellite data application. The method comprises the following steps of preprocessing Himawari-8 data to obtain the sea area data at a terminator moment; removing the medium-high cloud and low cloud through image segmentation; settinga brightness temperature threshold of a far infrared band to remove missing middle and high clouds; extracting a sensitive wave band, constructing a judgment index to form a judgment condition, setting a corresponding threshold value to remove a sea surface, and finally obtaining a sea fog recognition result. According to the method, the change characteristics of the ground object at the terminator moment and the spectral similarity characteristics of sea fog and low cloud are comprehensively considered, sea fog detection at the terminator moment can be achieved through combination of image segmentation and the judgment threshold value of the set judgment condition, and a foundation is provided for sea fog detection all day long.

A magnetic resonance imaging system capable of conducting spectroscopic imaging with an improved SNR without degrading the spatial resolution includes edge-preserving filter processing means for spectroscopic imaging. The edge-preserving filter processing means executes processing including the steps of calculating spectral similarity in spatial neighborhoods (spatially neighboring voxels) at each voxel in spectroscopic imaging data, calculating a spectral weight according to the spectral similarity, and conducting weighted smoothing for compounding spectra of spatial neighborhoods (spatially neighboring voxels) according to the spectral weight.

The invention relates to the field of analytical chemistry, and in particular to an infrared spectroscopic analysis and authentication method for cordyceps sinensis stromata. The invention aims at solving the technical problem that cordyceps sinensis stromata are difficult to distinguish the true from the false at present. The technical scheme for solving the technical problem of the invention is to provide a method for rapidly and accurately authenticating cordyceps sinensis stromata. The method comprises the following steps: a, drawing a standard spectrum of the cordyceps sinensis stromata; b, acquiring the necessary characteristic peaks of the infrared spectrums of the cordyceps sinensis stromata and the common characteristic peaks of the infrared spectrums of the cordyceps sinensis stromata; c, authenticating the characteristic peaks of the infrared spectrums of the tested cordyceps sinensis stroma samples and the similarity of the infrared spectrums. The invention provides a new method for rapidly and accurately authenticating the cordyceps sinensis stromata.