72 results about "Spectral processing" patented technology

The invention relates to the technical field of image processing, and provides an optimization method and device for a spectrum processing chip, electronic equipment and a storage medium, and the method comprises the steps: constructing a target function for designing modulation units, and enabling the target function to represent a mapping relation between the structure parameters of the micro-nano structure of each modulation unit and the demands of a target application scene; performing optimization solution on the target function based on the natural spectral image sample set in the target application scene, and determining structure parameters of the micro-nano structure of the modulation unit; and constructing the modulation unit according to the determined structural parameters of the micro-nano structure of the modulation unit, and generating the spectrum processing chip according to the constructed modulation unit and the image sensor. According to the invention, the spectrum processing chip is constructed, so that the plurality of modulation units and the image sensor in the chip jointly carry out optical simulation calculation of the inner product of the feature vectors on the incident natural spectrum image, visual light information of a spatial domain and a frequency domain can be directly detected, acquired and perceived in real time, and subsequent analysis of the feature information is efficient and rapid.

The invention provides a detection system of a micro optical fiber coupler based on a bidirectional optical frequency comb vernier effect, which can be applied to the technical field of photoelectric detection. The system comprises: an optical frequency comb light source configured to emit detection light; the beam splitter is arranged to split the detection light into first sub-detection light and second sub-detection light; the reference sensing unit is arranged to guide the first sub detection light to interact with a reference solution through an evanescent field of the first micro optical fiber coupler so as to output first transmission light; the detection sensing unit is arranged to guide the second sub detection light to interact with the solution to be detected through an evanescent field of the second micro optical fiber coupler so as to output second transmission light; the beam combiner is set to combine the first transmission light and the second transmission light to obtain a combined light spectrum; the spectrum analyzer is set to collect the combined light and generate a vernier spectrum; and the processor is set to determine the concentration of the liquid to be detected according to the vernier spectrum.

The invention discloses a saturated spectrum processing method for wide-range detection of an absorption spectrum of a tunable diode, and relates to the technical field of spectrum gas detection. According to the saturated spectrum processing method for wide-range detection of the absorption spectrum of the tunable diode, an effective spectrum signal is obtained by performing denoising and baseline correction on an original absorption spectrum signal, a plurality of spectrum characteristic parameters are extracted to calculate a saturated quantized value, and a complete absorption spectrum line contour is reconstructed by adopting a spectrum reconstruction algorithm according to the quantized value, so that the wide-range detection of the absorption spectrum of the tunable diode is realized. The method comprises the following steps of: constructing a saturation quantitative evaluation system, a layered self-adaptive spectrum reconstruction strategy, a dual-mode detection fusion mechanism and a working condition compensation model, calculating weights of an absorption spectrum and a wavelength modulation spectrum through an S-type function, correcting a preliminary concentration value by combining environment temperature, pressure and a compensation model, and outputting an accurate final gas concentration. Without additional hardware, wide-range high-precision gas detection is realized, and environmental interference and saturation influence are eliminated.

The invention mainly relates to the technical field of spectral data processing, and provides a spectral processing method suitable for a convolutional neural network in order to solve the problem that the classification or objective prediction accuracy of the convolutional neural network is not high due to the fact that spectral dispersion characteristic wavelengths are difficult to capture when the convolutional neural network processes one-dimensional spectral data. The core idea of the method is as follows: setting a target sample variable, screening out a wavelength points with high correlation with the sample target variable from M wavelength points of an original spectrum through a competitive adaptive reweighting algorithm based on the target sample variable as characteristic wavelength points, and setting the characteristic wavelength points as a fixed wavelength point combination; pre-processing the original spectrum according to a plurality of pre-processing modes; sorting the pre-processed fixed wavelength point combination of each spectrum and the other M-a wavelength points according to the generated K random sequences to generate spectrum matrix data; and training the convolutional neural network based on the generated spectral matrix data.

The present application relates to the technical field of image processing, and provides a spectral processing chip optimization method and apparatus, an electronic device, and a storage medium. The method comprises: constructing and designing a target function of modulation units, the target function representing a mapping relationship between a structural parameter of a micro-nano structure of each modulation unit and a requirement of a target application scenario; performing optimized solving on the target function on the basis of a natural spectral image sample set in the target application scenario, and determining structural parameters of micro-nano structures of the modulation units; and constructing the modulation units on the basis of the determined structural parameters of the micro-nano structures of the modulation units, and generating a spectral processing chip on the basis of the constructed modulation units and an image sensor. According to the present application, the spectral processing chip is constructed, the plurality of modulation units and the image sensor in the chip jointly perform optical analog computing on a feature vector inner product of an incident natural spectral image, and can directly detect, acquire and sense visual light information of a spatial domain and a frequency domain in real time, making subsequent feature information analysis efficient and rapid.

The application belongs to the technical field of optical communication and relates to a frequency-shift-free microwave photonic channelized receiving device, which comprises a carrier light frequency comb generating unit, a local light frequency comb generating unit, an electro-optical modulation unit, an optical phase shift unit, a channel selection and spectrum processing unit and a plurality of photoelectric conversion units. The received broadband radio frequency signal is modulated on each comb tooth of the carrier light frequency comb through the electro-optical modulation unit, and the frequency alignment between the optical local oscillator and the required receiving sub-channel can be realized based on the "vernier" effect caused by the different free spectral ranges of the carrier light frequency comb and the local light frequency comb and the inherent frequency alignment of the same pump laser. The mirror image sub-channel suppression is realized based on the opposite optical phase shifts on both sides of the pump light. The application has a simple structure, omits the frequency shift and special modulation structure required for frequency alignment, effectively suppresses the inter-channel crosstalk problem in the channelized receiving process, and can flexibly increase the channelization quantity according to actual needs.

The invention discloses a multi-component gas analysis method based on a photoacoustic spectrometry technology, which comprises the following steps: S100: exciting to-be-detected gas molecules; s200, decomposing the light into two beams, wherein each beam is subsequently guided to a different spectral processing path; s300, receiving and screening light in a specific wavelength range, and generating a targeted first spectrum and a targeted second spectrum which are matched with absorption characteristics of target gas; s400, modulating the first spectrum and the second spectrum to generate an optical signal with a specific frequency for subsequent signal processing; and S800, based on the amplification of the signal, analyzing the respective concentrations of various gases in the signal.

The audio and video reproduction method, device, integrated chip, and storage medium provided by embodiments of the present invention include receiving an audio signal, performing frame processing on the audio signal according to a preset frame length to obtain multiple audio frame signals, performing spectral processing on each audio frame signal to obtain spectral information, and then determining a sound and image perception width range, randomly generating an azimuth angle for each audio sampling point in the audio frame signal within the sound and image perception width range, determining an impulse response for each audio sampling point based on the azimuth angle of each audio sampling point in the audio frame signal, and determining an output audio signal of a sound-generating device based on each audio sampling point and the impulse response corresponding to each audio sampling point. The present invention can quantitatively determine the sound and image perception width range, and by feeding each audio sampling point to any azimuth angle within the sound and image perception width range, simulate a sufficiently fast-jumping virtual sound source, thereby allowing the user to perceive an expanded sound and image width and achieve a better sound and image reproduction effect.

The invention relates to the field of spectrum processing, and provides a micro spectrometer and a manufacturing method thereof, the micro spectrometer comprises a metasurface, the metasurface comprises a thin film with a plurality of grating structures, the plurality of grating structures are configured to generate qBIC modes for light with different wavelengths and output corresponding light intensities, the energy band of the grating structure presents flat-band dispersion in an incidence angle range of-10 degrees to + 10 degrees; the photoelectric detector is used for detecting the light intensity output by the plurality of grating structures of the metasurface and outputting an electric signal; and the calculation module is used for decoding the electric signal to obtain spectral information. According to the invention, the photon quasi-bound state (qBIC) and photon flat band characteristics in the continuous medium are utilized to realize high-sensitivity, high-performance and angle-insensitive spectrum detection.

The invention discloses an intelligent dysphagia complication early warning method and system based on artificial intelligence, and relates to the technical field of intelligent early warning monitoring. The method comprises the following steps: acquiring basic physiological parameters, chewing behavior data and related environment data of a target person, constructing a scene intention feature vector, and identifying a dysphagia behavior intention priority; performing short-period frequency spectrum processing on the basic physiological parameters and the chewing behavior data to generate a micro-rhythm offset spectrum, identifying micro-rhythm offset abnormity, and generating a rhythm precursor response value; constructing a respiration coupling curve, calculating a coupling deviation value, and forming a respiration coupling deviation sequence; and establishing a precursor fusion model, calculating a dysphagia risk index, generating a dysphagia risk level, automatically generating an intervention strategy, pushing the intervention strategy to a target person, and carrying out adaptive correction on the precursor fusion model. According to the invention, prospective and individual accurate early warning of the risk of dysphagia complications is realized, and the early warning effect is continuously optimized through an intervention feedback closed loop.

The present application relates to the technical fields of spectral analysis and substance detection, and discloses a deuterium lamp device and an atomic absorption spectral analysis method and system using the deuterium lamp. The system comprises a hollow cathode lamp unit, a built-in hollow cathode lamp emitting a first light beam after preheating under power supply; a first lens, a second lens and a third lens, a deuterium lamp device for emitting a second light beam at a first frequency under the driving of a pulse power supply; a sample cell for internally placing a sample to be measured in a combustion state; a monochromator for performing spectral processing on the incoming first light beam and / or the second light beam, and outputting the spectrally processed light signal to a photoelectric detector; the photoelectric detector for photoelectrically converting the light signal output by the monochromator to obtain an electric signal; and a data processing unit for analyzing and processing the electric signal of the photoelectric detector to obtain an analysis result. The present application can realize more accurate background deduction and improve the accuracy and reliability of atomic absorption spectral measurement.

The application relates to a spectrum processing method, a spectrum processing device and a spectrum processing system. The spectrum recovery method comprises the following steps: acquiring parameter information of a spectrum chip; configuring parameters corresponding to a shot spectrum image according to the parameter information of the spectrum chip, and collecting a spectrum image; and recovering spectrum information of a to-be-detected object based on the spectrum image and a corresponding spectrum recovery algorithm. In this way, a miniature spectrum processing device with small volume, convenient carrying, direct interaction with a terminal, convenience and high recovery precision can be realized.

The application provides a holographic image anti-counterfeiting pattern laser processing method, a storage medium and a product, a polarization angle sequence is determined through spectrum processing, a neural network is used to optimize microstructure arrangement, an encoding matrix is generated, hidden data is integrated to improve information dimension, then simulation calculation and iterative updating are used to enhance the anti-counterfeiting effect of the encoding matrix, resistance characteristics are extracted and robustness is verified, finally, material surface writing is guided to form a holographic image anti-counterfeiting pattern, and finally, a multi-dimensional verification protocol is combined to ensure the stability of the polarization response in different environments. The application significantly improves the cracking resistance and environmental adaptability of the anti-counterfeiting pattern through the deep combination of the polarization angle and the microstructure optimization.

The invention discloses a method for measuring a residual birefringence difference value of a transparent ferroelectric electro-optical material based on a Fabry-Perot resonant cavity. The method comprises the following steps: constructing a Fabry-Perot resonant cavity measuring light path consisting of a broadband light source, a polarization control assembly and a sample to be measured; the polarization state of incident light is accurately controlled through a magnetic control Faraday rotating mirror, and transmission spectrums in the o-light direction and the e-light direction are obtained respectively; a high-precision spectrum acquisition module is adopted to record transmission spectrum characteristics; based on an original spectrum processing algorithm, the residual birefringence difference value of the to-be-detected material is calculated by analyzing the resonant wavelength offset and the free spectrum range change of the o-light and e-light transmission spectrums. The method solves the problem that the weak birefringence detection capability of the existing measurement technology is insufficient, has the advantages of high measurement precision, strong anti-interference capability and the like, and is suitable for performance evaluation, process optimization and device performance improvement of the transparent ferroelectric electro-optical material.

This invention discloses a method for controlling the addition of supernatant from kitchen waste as a carbon source for denitrification based on three-dimensional fluorescence spectroscopy. The method includes steps such as sample collection and detection, three-dimensional fluorescence spectral processing, establishment of a multiple linear regression model, and control of the carbon source dosage. By acquiring real-time online data on the changes in three-dimensional fluorescence spectra during the denitrification process of nitrate-containing wastewater and combining this with a multiple linear regression model to predict carbon source consumption, this invention achieves precise control of the carbon source dosage, ensuring a sufficient supply of carbon source during the denitrification stage while avoiding waste caused by excessive carbon source addition. This significantly improves the denitrification rate, reduces energy consumption and operating costs, and ensures that the effluent quality meets standards.

The invention relates to the technical field of laser-induced breakdown spectroscopy data processing, in particular to an LIBS spectrum processing method and device based on residual skewness estimation and a storage medium. The invention discloses an LIBS (Laser-induced Breakdown Spectroscopy) processing method based on residual skewness estimation. The method comprises the following steps: step S10, acquiring an observation signal of an LIBS; s20, on the basis of morphological analysis of the LIBS spectrum, constructing a spectrum model by using prior distribution of a Bayesian framework; step S30, based on the updated hidden variable observation signal, analyzing the updated hidden variable spectrum model by using variational inference to obtain iteration deviation and posterior analysis data; s40, judging whether the updating hidden variable iteration deviation meets an iteration termination requirement or not; if yes, iteration is stopped, and actual spectrum data is obtained according to the updated hidden variable posteriori analysis data; and if not, executing the step S20. According to the method, the spectrum model is analyzed by utilizing variational inference, the optimal convergence point is judged, effective separation of observation signals is realized, posterior analysis data is obtained, and the spectrum estimation precision and the calculation efficiency are greatly improved.

A method for analyzing a mixture includes obtaining a first spectrum of the mixture comprising a plurality of components; selecting at least one position on the spectrum; and estimating a mixing weight for each of the components using an algorithm based on an intensity of the at least one position. A system for analyzing a mixture includes a spectrometer configured to obtain a first spectrum of the mixture comprising a plurality of components; a processor configured to select at least one position on the spectrum and estimate a mixing weight for each of the components using an algorithm based on an intensity of the at least one position.

The application discloses a wavelength modulation surface plasmon resonance imaging sensing method, which generates multiple groups of illumination spectra by controlling a light-emitting diode, processes the spectra into p-polarized light, excites a sensing chip to generate surface plasmon resonance through prism coupling, collects corresponding reflected light intensity images by using a CMOS camera, and reconstructs surface plasmon resonance reflection spectra of each pixel on the sensing chip by using a smooth constraint reconstruction algorithm according to intensity observation values of the pixel in all images and corresponding illumination spectrum data, so as to further extract a phase signal and generate a spatially resolved phase image. The phase signal and the imaging scheme realize high-sensitivity and high-throughput sensing performance. The method realizes low-cost, high-signal-to-noise ratio and high-throughput sensing imaging.

The present invention discloses a method for screening grafting affinity rootstocks using a spectral method and a fluorescent probe. The method comprises the following steps: first, the rootstock and the scion cut surface are tightly fitted together and then a grafting clamp is clamped thereon to perform graft healing period management. Before measurement, the bottom of a hole plate of a grafted seedling is immersed in a fluorescein disodium salt solution, the grafted seedlings are fluorescently stained after grafting, and then a 365nm ultraviolet lamp is used to obliquely illuminate the surface of the junction between the grafted seedling rootstock and the scion. A micro-fiber spectrometer is used to receive the fluorescence reflection spectrum at the junction between the grafted seedling rootstock and the scion, the spectral data is processed, and the relative fluorescence intensity of the scion healing part in the scion area at a wavelength of 525nm is calculated to determine whether the scion and the rootstock are effectively connected, thereby determining the grafting affinity of the melon rootstock.

This invention discloses a multi-channel spectral imaging system and method, belonging to the field of spectral imaging technology. The system includes an array optical imaging lens, an array multi-channel spectral sampling filter, an array color camera, and a spectral processing circuit module. This system enables solid-state snapshot imaging without moving parts, featuring a compact structure, high stability, and suitability for dynamic scenes, vehicle-mounted / airborne, and other complex operating conditions. The method of this invention completes multi-channel spectral imaging by constructing a measurement matrix, acquiring and converting target image signals, establishing the correlation between spectral signals and acquired signals, sparse representation and optimization solutions, super-resolution reconstruction, and data cube construction. It exhibits excellent reconstruction accuracy and noise resistance, and can flexibly expand the array size to adapt to the spectral coverage requirements of different application scenarios.

The present disclosure relates to an apparatus and a method for detecting and analyzing foreign matter in a cathode, the apparatus comprising: a first imaging device for dividing a cathode conveyed in one direction into a predetermined region and imaging the predetermined region; an image processing processor for processing an image of the predetermined region captured by the first imaging device to specify a position where the foreign matter is formed; a second image capturing device for capturing an image of the foreign matter specified by the image processing processor; and a spectral processing processor for analyzing a spectral signal of an image of the foreign matter captured by the second image capturing device to classify a component of the foreign matter.

An optical spectrum obtaining apparatus includes: a beam splitting unit, a first / second optical spectrum computation unit, and a processing unit. The beam splitting unit is configured to: receive a light beam from a light source, split the light beam into a first polarization component and a second polarization component with different polarization directions, transmit the first polarization component to the first optical spectrum computation unit, and transmit the second polarization component to the second optical spectrum computation unit. The first optical spectrum computation unit is configured to compute an optical spectrum of the first polarization component. The second optical spectrum computation unit is configured to compute an optical spectrum of the second polarization component. The processing unit is configured to obtain an optical spectrum of the light beam based on the optical spectrum of the first polarization component and the optical spectrum of the second polarization component.

The invention provides a multi-parameter optical fiber health monitoring device. The device comprises a light source system, a light source conduction system, a sensing system, a sensing light conduction system and a spectrum processing system, the light source system can emit light signals of a single wavelength or a plurality of different wavelengths; the light source conduction system can transmit light signals, provide multi-channel light signal channels and realize random switching of the light signal channels; the sensing system comprises at least one optical fiber sensor, and one end of the optical fiber sensor acts under the action of a marker to cause the change of an optical signal; the sensing light conduction system can pass back a changed light signal and gate of a path; and the spectrum processing system can collect and process the returned optical signals to obtain spectrum data. According to the invention, multi-channel synchronous monitoring can be realized, a plurality of physiological and biochemical markers can be detected at the same time, and the detection efficiency is greatly improved; the method has high sensitivity and stability; the invention has a wide clinical application prospect.

The invention discloses a tobacco online near infrared spectrum detection device and method based on a linear optical filter, and relates to the technical field of spectrum processing. Near-infrared light is emitted to a target area, a plurality of near-infrared bands are obtained through a linear optical filter, and reflected light of each near-infrared band is collected to obtain an original near-infrared band set; preprocessing the target original near-infrared band to obtain a target near-infrared band; performing feature extraction on the target near-infrared band to obtain target sub-features; and constructing a target feature vector according to all the target sub-features, and searching a preset database according to the target feature vector to obtain a detection result. Multiple near-infrared bands and frequency-combining and frequency-doubling absorption information of substance molecules corresponding to different bands are obtained through the linear optical filter, and then the different near-infrared bands are preprocessed and then subjected to feature extraction, so that prediction of different chemical substance components in tobacco is improved, and the online detection efficiency of the tobacco is improved.

The application discloses a kind of intelligent preparation systems of laser navigation milk teeth root canal, it is related to milk teeth root canal treatment field, including: image processing module, self-checking registration module, structure acquisition module, spectral processing module, anomaly detection module, fusion determination module, compensation tracking module, digital modeling module, simulation learning module, trajectory generation module, execution monitoring module, safety protection module, interactive command module and record audit module;The application can more accurately distinguish infected dentin and healthy dentin, reduce the risk of misjudgment or missing infected areas, effectively reduce the risk of root tip penetration and root tip tissue damage, so that the laser cutting process always revolves around the real infected area, both can fully remove infected tissue, and can maximize the retention of healthy dental structure, improve root canal preparation efficiency and safety, reduce the dependence of doctors on experience judgment, so that milk teeth root canal preparation operation is more standardized and repeatable.

A method and system for performing grayscale image generation. The method includes acquiring, by a color image sensor, color information of an environment in a field of view of an imaging assembly. The color image sensor includes a plurality of color pixels, each color pixel configured to acquire corresponding color information of the environment. Each pixel in the array of pixels has a different detection spectrum than an adjacent neighboring pixel. A processor determines a grayscale value for each pixel in the plurality of color pixels, the grayscale value determined from the color information of a two-by-two array of pixels from the color image sensor. The processor then generates a grayscale image from the grayscale value of each pixel.

The invention discloses a cloud collaborative atomized liquid spectral data intelligent analysis system, and relates to the technical field of atomized spectral data analysis. The cloud collaborative atomized liquid spectral data intelligent analysis system comprises a spectral data acquisition module used for acquiring spectral time sequence data of set atomized liquid and uploading the spectral time sequence data to a cloud platform for storage; the cloud spectrum processing module is used for inputting the spectrum time sequence data into a pre-trained atomization spectrum analysis model and analyzing a spectrum spraying mapping time sequence feature set; and the cloud spectrum analysis module is used for analyzing an atomization absorption cooperation characteristic value and an atomization dispersion quality characteristic value based on the set spectrum spraying mapping time sequence characteristic set of the atomization liquid. Combined spraying control is performed on the set atomization liquid based on the atomization absorption cooperation characteristic value and the atomization dispersion quality characteristic value in the cloud control feedback module; therefore, collaborative perception and dynamic adjustment of the characteristics of the atomized liquid are achieved, the quality of the spraying effect is guaranteed, and then the spraying precision is guaranteed.

The invention belongs to the technical field of optical communication, and relates to a frequency-shift-free microwave photon channelized receiving device, which comprises a carrier optical frequency comb generation unit, a local oscillation optical frequency comb generation unit, an electro-optical modulation unit, an optical phase shift unit, a channel selection and spectrum processing unit and a plurality of photoelectric conversion units. A received broadband radio frequency signal is modulated on each comb tooth of the carrier wave optical frequency comb through the electro-optical modulation unit, and based on a vernier effect caused by different free spectral ranges of the carrier wave optical frequency comb and the local oscillator optical frequency comb and inherent frequency alignment of the same pumping laser, frequency alignment of the optical local oscillator and a required receiving sub-channel can be realized. And mirror image sub-channel suppression is realized based on opposite light phase shifts on two sides of the pump light. According to the invention, the structure is simple, a frequency shift and a special modulation structure required by frequency alignment are omitted, the problem of inter-channel crosstalk in a channelization receiving process is effectively suppressed, and the channelization number can be flexibly increased according to actual demands.

This application relates to a method, apparatus, computer device, and computer-readable storage medium for generating fiber Raman spectroscopy preprocessing models. The method includes: determining the true noise spectrum and random noise of the target object; generating a clean spectrum corresponding to the true noise spectrum based on a spectral synthesis method; generating fluorescence background data corresponding to the true noise spectrum based on a fluorescence background synthesis strategy; training a model using the true noise spectrum, clean spectrum, and fluorescence background data to obtain a trained background removal model; and training a multi-channel model based on random noise and clean spectrum to obtain a trained noise removal model. By employing this method, a reliable model processing foundation can be provided for background separation processing of Raman spectral signals and for collaborative denoising in both the spectral and frequency domains, improving the generation efficiency and quality of training datasets and achieving efficient and high-quality spectral processing in complex environments.