14 results about "Spectral pattern" patented technology

To achieve high key performance. [Solution] The lock system according to the embodiment comprises a light source, a spectral pattern generation unit, a sensor, and an unlocking determination unit. The light source generates light. The spectral pattern generation unit generates incident light of a spectral pattern defined for a spectral pattern based on one spectral pattern selected from a plurality of spectral pattern generation patterns, based on the light generated from the light source. The sensor has one structural pattern selected from a plurality of structural patterns, and when incident light of the spectral pattern is incident on it, it outputs a different current value according to the selected one structural pattern. The unlocking determination unit determines, based on the output of the sensor when the incident light is incident, whether the combination of the one spectral pattern generation pattern and the one structural pattern of the sensor is an expected combination.

A spectroscopic device includes: an analysis optical system; a length measurement optical system; and a calculation device. The analysis optical system includes a moving mirror and a first light receiving element. The length measurement optical system includes a second light source configured to emit laser light, a gas cell with a gas that absorbs light of a predetermined wavelength sealed therein and configured to cause the laser light to be incident thereon, an emitted light amount detection unit configured to detect an amount of light emitted from the gas cell and output an emitted light amount detection signal, a light source control unit configured to control a wavelength of the laser light based on the emitted light amount detection signal, and a length measurement unit configured to use the laser light to obtain a displacement signal corresponding to a position of the moving mirror, and the calculation device includes a moving mirror position calculation unit, a light intensity calculation unit, and a Fourier transform unit configured to generate a spectral pattern.

This application discloses an image sensor, a camera module, an electronic device, and a method for generating depth images, belonging to the field of camera technology. The image sensor includes a first signal readout circuit, a first switching element, N third switching elements, and multiple pixel groups arranged in an array, where N is an integer greater than 1. Each pixel group includes N photodiodes and N photosensitive pixels, arranged in an array, with each of the N photosensitive pixels corresponding to one of N spectral patterns. The photosensitive pixels are used to sense light of a first spectral pattern, which is the spectral pattern corresponding to the photosensitive pixel. The N third switching elements are connected to the N photosensitive pixels in a one-to-one correspondence. The first connection terminal of each photosensitive pixel is electrically connected to the input terminal of the first switching element through the third switching element connected to the photosensitive pixel, and the output terminal of the first switching element is electrically connected to the input terminal of the first signal readout circuit.

The application discloses an echo cancellation method and system based on a heterogeneous multi-scale convolutional recurrent network, which first constructs a correlation encoder based on the extraction of relevant feature patterns in multiple frames of context, then constructs a spectral encoder according to the extraction of spectral patterns in a single frame, and then adopts the correlation encoder and the spectral encoder to form a heterogeneous multi-scale encoder and obtain multiple paths of encoded features; the application realizes the function of extracting different features from the input mixed signal spectrum and the reference signal spectrum by using the heterogeneous dual-path encoder, adopts correlation coding to adapt to the time delay between the reference signal and the mixed signal, and extracts correlation feature expressions from multiple frames of spectrum, simultaneously adopts spectral coding to extract spectral pattern expressions from the current frame in order to strengthen the spectral resolution of the current frame, and realizes the function of further adopting a channel attention feature soft selection mechanism to fuse the correlation feature coding and the spectral coding for the dual-path features, and is suitable for being widely promoted and used.

This invention provides an intelligent analysis system for photoelectric detection spectral data based on deep learning, belonging to the fields of photoelectric detection, spectral data analysis, and deep learning application technology. The system includes core units such as a spectral data acquisition unit, a preprocessing unit, a feature extraction unit, and a deep learning analysis unit, as well as auxiliary units for anomaly detection and data storage, forming a closed-loop architecture of acquisition-processing-analysis-output-optimization. It improves data quality through step-by-step preprocessing algorithms, reduces computational load by adaptively selecting key features, deeply mines spectral patterns using a model that integrates convolutional neural networks and attention mechanisms, achieves rapid scenario adaptation through transfer learning, and dynamically optimizes model parameters based on a feedback mechanism. The system solves the problems of low accuracy, poor adaptability, and long training cycles in traditional spectral analysis, achieving accurate, efficient, and intelligent spectral data analysis, and is suitable for various photoelectric detection scenarios such as qualitative and quantitative analysis of substances and environmental monitoring.

The invention discloses a plasma near-infrared spectrum mode recognition method for detecting latent hepatitis B. The method comprises the following steps: S1, collecting spectral data of a plasma sample to be tested in a near-infrared band; s2, acquiring multi-repetition near infrared spectrum data of each sample, and constructing a spectrum data set; s3, grouping samples of a latent hepatitis B infection group and a normal control group in the spectral data set; s4, constructing an OBI-normal control spectrum discrimination model of the plasma sample based on a classifier primitive algorithm; s5, preprocessing parameters are optimized based on the discriminant performance of the classifier; s6, constructing an optimal wavelength combination; and S7, checking the optimal model. Compared with the prior art, the plasma near infrared spectrum mode recognition method for detecting the latent hepatitis B has the advantage that a solution can be provided for developing a small special blood analyzer for detecting latent hepatitis B infection.

The purpose of the present invention is to make it possible to estimate the thickness of an object to be processed, which is subjected to plasma etching processing, with high accuracy even if a spectral pattern prepared in advance and a spectral pattern measured during actual processing deviate from each other. Therefore, this plasma processing / control device compares a measurement spectrum obtained by receiving plasma reflected light from a processing chamber at any time during etching with an estimation spectrum obtained in advance in association with the thickness of the same object to be processed. Spectral errors are calculated between the measured spectrum and the plurality of spectrums for estimation, and the thickness of the object to be processed at the arbitrary point in time is estimated using the spectrum for estimation that takes at least a first minimum value among the plurality of spectral errors and the spectrum for estimation that takes a second minimum value among the plurality of spectral errors.

A line-scan Brillouin microscopy apparatus is configured to operate in a coaxial configuration. The line-scan Brillouin microscopy apparatus includes an illumination source that provides a P polarized illumination light beam for illuminating a sample. A first optical assembly provides the P polarized illumination light beam to the sample and collects initial Brillouin scattered light from the sample. The first optical assembly includes an optical component that converts the initial Brillouin scattered light to S polarized Brillouin scattered light. A second optical assembly is configured to receive the S polarized Brillouin scattered light from the first optical assembly. Characteristically, the second optical assembly is configured to induce a spectral dispersion. A detection unit is configured to detect a spatio-spectral pattern of the initial Brillouin scattered light. Advantageously, multiple points of the sample along P polarized illumination light beam are measured simultaneously.

This invention discloses an online detection method for impurities in lithium carbonate smelting based on deep learning, belonging to the field of deep learning spectral detection technology. The method includes constructing a deep learning model and sequentially inputting a high-quality spectral frame set into a convolutional feature extraction layer, a feature sieving layer, a feature aggregation layer, and a fully connected output layer. Local spectral pattern extraction, feature sieving, and hierarchical fusion are performed to generate spectral feature vectors. The melt temperature is deduced from the spectral feature vectors, and the activity coefficient of lithium carbonate smelting impurities is estimated. Numerical calculation is performed using an activity-coupled spectral thermometry equation to obtain the residual of the activity-coupled spectral thermometry equation set, and simultaneously, the concentration of lithium carbonate smelting impurities is generated. This invention achieves deep representation learning of high-quality spectral frame sets for extracting multi-time-time, multi-band correlated features, enhancing feature stability and recognition accuracy; it also achieves collaborative calculation of temperature and impurity concentration for online quantitative detection, improving the accuracy and continuity of concentration calculation.

This invention belongs to the field of artificial intelligence time series prediction technology. Addressing the problems of semantic distortion due to fixed time-domain segmentation, poor adaptation to heterogeneous distributions, and low prediction accuracy in existing multivariate time series prediction methods, this invention discloses a multivariate time series prediction method based on frequency-domain adaptive variable-length segmentation. This invention performs reversible instance normalization on multivariate time series data, obtains segment tensors, variable identifiers, and gating weights through frequency-domain adaptive variable-length segmentation, and completes joint modeling of spectral relationships using a two-stage attention mechanism consisting of pattern-level spatial attention and variable-level temporal attention. After frequency-domain spectral expert learning to fuse seasonal and trend components, refined segmentation is obtained, and the prediction results are output after reconstruction. This invention achieves precise alignment between segmentation and spectral patterns, adapts to heterogeneous time series distributions, improves feature representation accuracy and prediction robustness, and reduces model computational overhead. It can be widely applied to multivariate time series prediction scenarios such as traffic flow and power load.

The present disclosure relates to an information authentication method and system, the method is applied to a receiving spectrometer, the method comprises: detecting a first spectral pattern to obtain a first detection signal; if a first similarity between the first detection signal and a preset signal is within a preset range, or a difference between a similarity threshold and the first similarity is less than a first preset value, it is determined that the first spectral pattern is identical to a target spectral pattern, and it is judged that the first detection signal is a true value; the preset signal is obtained by detecting the target spectral pattern by a sending spectrometer, and the similarity between the signals obtained by the sending spectrometer and the receiving spectrometer for identifying any identical spectral pattern is within a preset range. The embodiment of the present disclosure can quickly realize the authenticity identification of the first spectral pattern. Due to the physical characteristics of the spectrometer itself, the spectrometer is difficult to be counterfeited, so the information authentication has the characteristics of high security, and through the spectral pattern storage information, the information storage capacity can be improved.

The invention discloses a kidney biological age calculation method based on glycosylation markers and integrated machine learning, and belongs to the technical field of bioinformatics and health monitoring. The method comprises the following steps: firstly, acquiring N-glycosylation spectral pattern data of plasma immunoglobulin G of an individual and preprocessing; secondly, carrying out feature dimension reduction on the glycosyl data by adopting a self-adaptive group elastic network, and screening candidate glycosyl structures; then, further screening out a target glycosyl structure associated with kidney functions by using an integrated machine learning algorithm, and performing causal inference through Mendel randomization to determine a final glycosyl structure; and finally, based on the final glycosyl structure and the individual age, constructing a multi-factor Logistic regression model, and calculating to obtain the kidney glycosyl biological age of the individual. According to the method, quantitative evaluation of the kidney specific biological age is realized by integrating glycomics data and a machine learning algorithm.

This invention discloses a rapid identification method for unknown radio signals based on a large electromagnetic spectrum model, relating to the field of electromagnetic spectrum monitoring technology. The method involves acquiring clean spectrum samples for self-supervised mask modeling pre-training to construct an electromagnetic spectrum spatiotemporal cognitive model capable of predicting expected spectral patterns based on context. Real-time frequency sweep data is input into the model to obtain the expected spectrum map, and its similarity difference with the actual spectrum map is calculated. Cosine distance is calculated as the feature difference by extracting features from the intermediate layers of the model, and the two are fused to obtain a cognitive difference matrix, which is then mapped to a two-dimensional surprise map. Spatiotemporal correlation tracking is performed on multiple consecutive frames of surprise maps to form surprise trajectories. The type of credible unknown signal is determined based on the frequency of occurrence, the mean surprise level, and frequency stability conditions. Original spectrum segments from the signal occurrence region are collected as positive samples, and the model parameters are incrementally fine-tuned using a low-rank adaptation matrix. This invention enables the discovery of unknown radio signals without relying on prior knowledge.

Provided is an acoustic signal analysis system and the like that can accurately separate a sound into its individual components even if the components of the constituent units overlap in time. [Solution] An approximation calculation unit 31 performs nonnegative matrix factorization on a spectrogram matrix obtained by short-time Fourier transform of an acoustic signal of a speech sound to estimate a basis matrix indicating a spectral pattern and an activation matrix indicating time changes in signal strength. A normalization unit 32 normalizes the basis matrix using the Euclidean norm of the basis matrix and scales the activation matrix using the Euclidean norm. A separation unit 33 separates the acoustic signal into basis components based on the normalized basis matrix and the scaled activation matrix.