38 results about "Shot noise" patented technology

The invention discloses an orchard unmanned aerial vehicle multispectral image noise restoration method based on a generative adversarial network, and relates to the technical field of multispectral image noise restoration, and the method comprises the steps: synthesizing a noisy sample containing stripes, photon particles, thermal noise, Rayleigh, shielding, compression artifacts, overexposure and waveband crosstalk on a noise-free reference image through a conditional generative adversarial network; a crosstalk matrix T obtained through optical calibration is introduced; based on the canopy / soil / sky semantic partition, main-cooperation-subdivision level superposition is implemented; generating hierarchical labels by taking energy conservation consistency and thermal noise correlation as thresholds; a branch and region self-adaptive repair module corresponding to classification is arranged in the image-to-image generative adversarial network, and joint loss training containing T regularization is adopted; and outputting the repaired image subjected to physical consistency checking. The method gives consideration to visual quality and spectral fidelity, inhibits cross-band pollution, improves the stability of indexes such as NDVI and NDRE, and is suitable for precision agricultural monitoring.

The invention relates to the technical field of image processing and extremely weak light imaging, and particularly provides a photon counting image reconstruction method and system, and the method comprises the steps: determining a target threshold value based on the adjustment of a dynamic adjustment coefficient under different illumination or gain conditions based on an initial threshold value; counting a first photon number of a single photon signal in the background image and a second photon number of the single photon signal in the signal image based on a target threshold, and generating noise photon number statistical distribution and signal photon number statistical distribution to determine a dynamic photon number processing interval; and performing time-space joint accumulation on the pixels conforming to the dynamic photon number processing interval, generating a two-dimensional photon accumulation matrix, and converting the two-dimensional photon accumulation matrix into a gray level image to form a single group of reconstructed gray level images. According to the method, real signals can be effectively extracted, the influence of dark noise, shot noise and readout noise is suppressed, zero noise detection is realized, and finally, the dynamic range, the signal-to-noise ratio and the stability of the image are remarkably improved through single-group image reconstruction and multi-group integration processing.

Collected light may be split into fields using grazing incidence mirrors to avoid illuminating gaps between the active areas for readout electronics. The grazing incidence mirrors may reduce lost field space in the integrating direction by splitting the imaging field in that direction to allow the readout electronics to be outside of the field. The fields may be split so there is space for readout circuits but no associated light on the readout circuits. The size of the illumination may then be decreased and / or the number and / or size of sensors increased to improve the photon collection efficiency. The better photon collection efficiency may reduce shot noise and / or improves tool throughput.

The present application relates to the technical field of quantum sensing based on atomic cell, in particular to a light glue inner mirror surface reflection multi-optical path atomic cell and application thereof. The light glue inner mirror surface reflection multi-optical path atomic cell of the present application increases the optical path of the probe laser in the atomic cell cavity and amplifies the signal without changing the size of the spectral analysis cell chamber; ultimately, the fundamental sensitivity of the measurement is limited by the quantum shot noise of the atoms, and the limit of such sensitivity is called the standard quantum limit, which is inversely proportional to the square root of the number of atoms simultaneously detected in the chamber. The multi-optical path design of the light glue inner mirror surface reflection multi-optical path atomic cell of the present application can make the probe laser cover the entire cross section of the atomic cell cavity and simultaneously detect almost all the atoms in the atomic cell cavity, thereby greatly improving the signal-to-noise ratio of the measurement and the sensitivity under the standard quantum limit caused by the shot noise.

The invention belongs to the technical field of commercial cipher system testing, and particularly relates to a quantum random number generator-based commercial cipher system entropy source quality detection method, which comprises the following steps of: acquiring shot noise of a light quantum detector at a high sampling rate, removing direct current bias by using a moving average subtraction method, and acquiring a quantum fluctuation sequence of a zero mean value; based on a variance ratio of the sequence and short-range autocorrelation, nonlinear mapping of an exponential function is utilized to obtain correlation degree so as to amplify a weak non-quantum memory effect; obtaining an evaluation model containing hyperbolic cosine and a logarithmic function by combining the drift potential energy of the bias voltage relative to the target value and the aging time of the device, and obtaining a dynamic entropy loss index; and comparing the entropy loss index with a preset threshold value, and executing self-adaptive bias current feedback regulation or triggering hardware fusing according to a comparison result. According to the invention, the problem of recessive degradation of the entropy source caused by drifting of the working point of the detector is solved, and accurate closed-loop control of the quality of the entropy source is realized.

The present application belongs to the technical field of commercial cipher system test, and particularly relates to a method for detecting the quality of an entropy source of a commercial cipher system based on a quantum random number generator, which comprises: obtaining shot noise of an optical quantum detector at a high sampling rate, removing direct current bias by using sliding average subtraction, and obtaining a quantum fluctuation sequence with zero mean; obtaining a correlation degree by using nonlinear mapping of an exponential function based on the variance ratio and short-range autocorrelation of the sequence, so as to amplify weak non-quantum memory effects; obtaining an evaluation model containing a hyperbolic cosine and a logarithmic function by combining the drift potential of a bias voltage relative to a target value and the aging time of a device, so as to obtain a dynamic entropy loss index; and comparing the entropy loss index with a preset threshold value, and performing adaptive bias current feedback adjustment or triggering hardware fusing according to the comparison result. The present application solves the problem of implicit degradation of an entropy source caused by drift of a detector operating point, and realizes precise closed-loop control of the quality of the entropy source.

The invention discloses an all-normal dispersion super-continuum spectrum driven resonant optical gyroscope and a working method thereof, and belongs to the field of optical gyroscopes. The input end of the mode-locked laser ML is connected to a left port of the full-normal dispersion photonic crystal fiber ANDi-PCF, input ports of the full-normal dispersion photonic crystal fiber ANDi-PCF and the photoelectric detector PD are connected to two left ports of the circulator CIR respectively, and a right port of the circulator CIR is connected with a left port of the multifunctional integrated optical chip MIOC. And two ports on the right side of the multifunctional integrated optical chip MIOC are respectively connected with two ports of the waveguide ring resonator WRR. Based on the resonance mode of the full-normal dispersion super-continuum spectrum excitation waveguide ring resonator, the influence of relative intensity noise and shot noise on the gyroscope performance is further reduced by using the broadband and smooth light source spectrum pattern, and finally the angle random walk of the gyroscope is promoted.

The embodiment of the invention provides a sample image generation method and device, a model training method and device, equipment and a medium. The method comprises the steps of obtaining a noiseless image and a target image shot by a camera; predicting noise parameters of the camera based on the features of the target image; adding shot noise, readout noise and row noise in the noiseless image according to the noise parameters to obtain a noise image; and determining the noiseless image and the noise image as sample images corresponding to the camera. The electronic equipment can realize automatic generation of the sample image by predicting the noise parameter and adding the noise based on the predicted noise parameter, and the efficiency of obtaining the sample image is improved. Moreover, a large number of noise images can be generated on the basis of a public noiseless image only through one image shot by a camera, and the cost of obtaining a sample image is reduced. In addition, a noise image generated by adding the most important noise in three types of real images including shot noise, readout noise and row noise can be well fit with a real noise image.

The invention relates to the technical field of network security communication, in particular to a wireless network camera mainboard security communication protocol optimization method and system, and the method comprises the following steps: reading a signal-to-noise ratio and a signal intensity numerical value, calculating a link quality score, capturing photoelectric shot noise, converting the photoelectric shot noise into a true random entropy source sequence, and carrying out the real random entropy source sequence; generating a dynamic session key in combination with clock jitter; analyzing the monitoring video stream to separate an I frame and a P frame, when the score is lower than a threshold value, only encrypting the I frame to generate a core ciphertext, and generating an integrity check code for the P frame; and constructing a security protocol data frame according to the error correction coding rate and executing a data transmission action. According to the method, the encryption density of the video data is dynamically adjusted according to the signal-to-noise ratio of a real-time signal by introducing a dynamic fragmentation mechanism of channel state perception, so that the calculation overhead of a mainboard in a weak network environment is effectively reduced, and the traditional complete handshake process is replaced by lightweight re-authentication logic based on hardware noise characteristics; and the problem of picture interruption caused by frequent reconnection is fundamentally eliminated.

This invention relates to the field of ultrasensitive magnetic measurement technology based on quantum sensing, and in particular to a single-beam zero-magnetic-field dual-axis vector magnetometer and its usage and applications. This invention employs a single probe laser beam to simultaneously perform pumping and detection functions, and innovatively combines the tensor-vector polar moment conversion (AOC) effect, magnetic field modulation, and large-amplitude modulation of the optical frequency. This significantly reduces the resonant linewidth of the magnetometer while suppressing photon shot noise and detector electronic noise, and achieves dual-axis magnetic field measurement in a zero-magnetic environment. In laboratory desktop experiments, a sensitivity of level 10 can be achieved.

A detector can include a group of orthogonally oriented quantum detectors, wherein each quantum detector among the group of orthogonally oriented quantum detectors can detect variations in a quantum field. The detector can further include a signal processor that can determine a direction and magnitude of the detected quantum field variations based on outputs from the group of orthogonally oriented quantum detectors. Each quantum detector among the group of orthogonally oriented quantum detectors can include an array of Zener diodes biased to produce shot noise modulated by quantum field variations. The detector can constitute a three-dimensional quantum field detector. In some embodiments, a tunneling sensor comprising a tunneling capacitor may be implemented in place of Zener diode.

The present application relates to the technical field of quantum sensing based on atomic cell, in particular to a light cement total reflection multi-optical path atomic cell and its application. The light cement total reflection multi-optical path atomic cell comprises an atomic cell cavity, a window sheet and a right-angle prism; two side opening positions of the atomic cell cavity are provided with the window sheet, the window sheet is connected with the atomic cell cavity by light cement, and the side of the window sheet away from the atomic cell cavity is provided with the right-angle prism allowing multiple reflections of the probe laser; the window sheet is provided with an entrance and an exit allowing the probe laser to enter and exit. In use, the probe laser beam is totally reflected back and forth between the two right-angle prisms and passes through the atomic cell cavity multiple times; the atomic cell cavity allows the light intensity of the probe laser to be uniformly distributed in the atomic cell cavity, so that almost all the atoms in the atomic cell cavity can be detected at the same time, thereby improving the detection signal intensity and the sensitivity under the standard quantum limit determined by the atomic shot noise.

The invention discloses a wireless light transmission method based on characteristic value gradient distribution and a large-view-field MPPC array, and relates to the technical field of visible light communication. The method comprises the following steps: constructing an MPPC array receiving system, and adjusting a receiving signal-to-noise ratio by adjusting an array position; an array receiving self-identification merging algorithm is adopted to extract edge power gradient features and fuse signals; establishing a pixel channel model containing received signal power, environmental noise and shot noise; analyzing channel gradient change and calculating edge parameters; an EIGD algorithm is applied to allocate channel weights to balance noise suppression and signal enhancement; constructing a receiving model and calculating an output signal-to-noise ratio; and error code analysis and algorithm performance comparison are carried out. According to the method, the signal detection efficiency under low illumination is improved through the EIGD algorithm, the robustness is prominent in a low signal-to-noise ratio or dynamic environment, and the requirements of unmanned aerial vehicle communication for high-speed and reliable transmission are met.

A quantum random number generator (QRNG) system including an optical coupler, the optical coupler including at least one light emitter and at least two light receivers, the at least two light receivers being electrically coupled to each other in a configuration that suppresses common mode noise. The system also includes a controller configured to control the at least one light emitter to randomly emit photons to the at least two light receivers, measure shot noise signals generated by the at least two light receivers in response to receiving photons, and extract random numbers from the shot noise signals.

An inspection system includes a controller including one or more processors configured to execute program instructions. The program instructions cause the one or more processors to receive at least a first portion of a first set of repeat swaths of a first scan path of a sample. The program instructions cause the one or more processors to generate an image by averaging the first portion of the first set of repeat swaths. Averaging the first portion of the first set of repeat swaths reduces a noise in the image. The program instructions cause the one or more processors to detect one or more defects in an inspection region of the sample using the image.

The present application relates to a kind of in-situ optical microscopic imaging high temperature growth system, the system includes: high temperature growth furnace, high temperature microscopic objective lens built in the high temperature growth furnace, very narrow single-wavelength incoherent light source and observation device.Therein, high temperature microscopic objective lens built in the high temperature growth furnace uses high-temperature-resistant material, need not cooling to work normally under high temperature environment;The single-wavelength characteristic of very narrow single-wavelength incoherent light source avoids chromatic aberration, can be close to diffraction limit to imaging resolution, and the incoherence of very narrow single-wavelength incoherent light source can avoid shot noise, to obtain clear imaging picture, so that the system can realize 20-1200 ℃ high temperature under the in-situ optical microscopic imaging of material growth process, to obtain the nucleation rate of material growth process, growth rate, material morphology and structure etc.Experimental data, it has extremely important significance to material high temperature growth research.

This invention discloses a tunneling photodiode suitable for high-energy photon detection and its fabrication method. It employs a very thick intrinsic perovskite crystal as the high-energy photon absorber, achieving high photon absorption and conversion efficiency. A perovskite P-type layer and an N-type layer are grown at both ends of the intrinsic perovskite crystal using solution-doped epitaxy, forming a perovskite PIN junction. Dark current and noise are suppressed using the depletion layer barrier in the junction region. Narrow-bandgap quantum dots are epitaxially grown in the junction region, forming band traps. When high-energy photons are incident, photogenerated carriers generated by the intrinsic perovskite absorber are injected into the quantum dot energy level traps, resulting in tunneling field emission and a high-gain photocurrent. Compared to conventional high-energy photon PIN detection structures, this invention achieves higher photocurrent gain due to the introduction of the tunneling field emission structure. Compared to high-energy photon avalanche diode detection structures, this invention exhibits lower shot noise because it does not generate random avalanche effects.

A method in an electronic device manages thermal energy events affecting the performance of an image capture device. The method includes detecting an elevated thermal energy event at the image capture device, determining if a second wireless communication subsystem is situated farther from the image capture device than a first wireless communication subsystem, and checking if the second wireless communication subsystem is inactive and has a lower temperature than the first wireless communication subsystem. If these conditions are met, the method transfers electronic communication operations from the first wireless communication subsystem to the second wireless communication subsystem. Additionally, the method can perform thermal mitigation operations as well.

In an instrument configured to spectroscopically divide fluorescences emitted from a plurality of capillaries and collectively measure the fluorescences using an image sensor, when the number of pixels of a binning region on the image sensor on which a predetermined wavelength-band component of each fluorescence is projected is denoted by Bm, the number of pixels of hardware binning is denoted by Bh, the number of pixels of software binning is denoted by Bs, Bm=Bh×Bs, the total noise measured in a case where Bm=Bh=Bs=1 is denoted by N, the readout noise is denoted by Nr, the dark-current noise is denoted by Nd, and the shot noise is denoted by Ns, Bm, Bh, Bs, N, Nr, Nd, and Ns satisfy a predetermined relationship, thereby realizing high sensitivity and high dynamic range in fluorescence measurement.

The application discloses a kind of single light source single-axis fiber-optic gyroscope electric signal demodulation method, it is related to quantum optics precision measurement technical field, including, the joint probability density function model of quantum fluctuation noise is constructed, through the dynamic correction measured noise parameter of bayesian inference, generate the noise model containing shot noise variance and thermal noise mean;Utilize Y waveguide beam splitting characteristics to generate multiple groups of reference-measurement optical path pairs, calculate the cross-correlation coefficient of interference signal of each group of optical path pairs, and select the two groups of high correlation signal pairs with the highest correlation degree;According to high-precision phase trajectory, based on the phase-velocity conversion relationship of Sagnac effect, the Sagnac phase shift formula is applied to calculate the rotation angular velocity value and output.The application realizes the efficient suppression of shot noise and thermal noise by dynamic noise modeling and parameter adaptive adjustment mechanism, significantly improves the integrity and accuracy of the phase signal after noise reduction.

The disclosure relates to photolithography systems, methods and masks for semiconductor fabrication, focusing on overcoming diffraction-limited critical dimensions in optical lithography. Current extreme ultraviolet (EUV) systems typically achieve critical dimensions of -12-20 nm. However, finer critical dimension yields are constrained by stochastic effects including photon shot noise, unacceptable line-edge roughness, and unacceptable random bridging or break defects. These random failures, rather than systematic aberrations, dominate yield loss for the 12-20 nm features' technologies. Attempts to reduce defects by increasing source power or resist sensitivity worsen the resolution / roughness / sensitivity trade-off, further narrowing the process window. Therefore, new lithographic methods are provided, producing sub- 10 nm features with high fidelity, low probability of stochastic defectivity, and compatibility with both 300 mm and the underserved 200 mm fabrication infrastructures.

The invention relates to a video generation method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a multi-lens prompt text, a multi-lens noise video and lens cutting time indication information; respectively coding the multi-shot prompt text and the multi-shot noise video to obtain a multi-shot text feature and an initial multi-shot noise feature; performing rotation position coding on the resectoscope time indication information to obtain a first rotation position coding feature (used for indicating a resectoscope time position); performing rotation transformation on the initial multi-lens noise feature according to the first rotation position coding feature to obtain a target multi-lens noise feature added with a tangent time position identifier; and performing multi-shot video generation processing on the multi-shot text features and the target multi-shot noise features based on a video generation diffusion model to obtain a multi-shot generated video. According to the technical scheme provided by the invention, flexible lens cutting can be realized in multi-lens video generation, and the video clip corresponding to each lens can be in any duration.

The invention discloses a dark light diagram optical flow synthesis algorithm based on data enhancement and a dark light diagram optical flow intelligent test data set (ReLOLFlow). The improved ISP simulation algorithm is provided for solving the problem that the performance of a model is reduced in a dark environment due to the fact that an existing optical flow data set lacks low-illumination scene data. A normal illumination RGB image is converted into RAW data through a reverse ISP process, shot noise and reading noise are simulated and introduced in a RAW domain, the exposure degree is adjusted, and a high-trueness synthetic dark light image is generated through a forward ISP process. The invention further provides a construction method of the ReLOLFlow dark light graph optical flow intelligent test data set. Through manual screening and multi-model prediction and correction, a low-illumination video sequence with a real optical flow label is obtained, and the low-illumination video sequence is used for accurately evaluating the performance of the model in a dark environment. Large-scale dark light flow training data can be generated through a synthesis algorithm, evaluation is performed in combination with a ReLOLFlow test set, and the generalization ability and precision of a light flow estimation model in a complex low-light environment can be remarkably improved.

The application relates to the technical field of quantum secure communication, and discloses a device and a method for measuring shot noise and electric noise in real time, which comprises an optical coupler, a balanced detection module and a data acquisition and analysis module connected in sequence, and further comprises a first optical switch module and a second optical switch module connected with the optical coupler respectively; the first optical switch module receives quantum signal light, and the second optical switch module receives local oscillation light. The application solves the problems of the prior art, such as the incapability of monitoring shot noise and electric noise in real time, the inaccuracy of parameter estimation, the low safety code rate and the low system safety.

In an instrument configured to spectroscopically divide fluorescences emitted from a plurality of capillaries and collectively measure the fluorescences using an image sensor, when the number of pixels of a binning region on the image sensor on which a predetermined wavelength-band component of each fluorescence is projected is denoted by Bm, the number of pixels of hardware binning is denoted by Bh, the number of pixels of software binning is denoted by Bs, Bm=Bh×Bs, the total noise measured in a case where Bm=Bh=Bs=1 is denoted by N, the readout noise is denoted by Nr, the dark-current noise is denoted by Nd, and the shot noise is denoted by Ns, Bm, Bh, Bs, N, Nr, Nd, and Ns satisfy a predetermined relationship, thereby realizing high sensitivity and high dynamic range in fluorescence measurement.

The photoelectric detector modeling method comprises a photoelectric detector modeling method based on an equivalent circuit technology, the final output current of a photoelectric detector is mainly composed of a light current generated by photon absorption, a dark current, shot noise and thermal noise, and the model output current is shown in the specification.

The present application relates to the field of image processing, in particular to a confocal microscope image filtering enhancement method based on adaptive compensation, the method comprising: collecting multiple frames of dark signal images under no excitation light, obtaining pixel-by-pixel noise weighted mean and variance through exponential decay weighted statistics; determining the deduction ratio according to the ratio of local variance of the original image to noise variance, the ratio is constant when low and decreases when high, and the noise mean is deducted to obtain a compensation image; low-pass filtering the compensation image to extract the detail layer, calculating the local signal-to-noise ratio using the local variance of the compensation image and the noise variance, and determining the enhancement coefficient according to the local signal-to-noise ratio, and adding the weighted detail layer back to the compensation image to obtain an enhanced image. The method dynamically follows the time-varying noise to realize adaptive bias compensation and selective detail enhancement, effectively suppresses non-uniform background and shot noise, protects edge details, and improves the quantitative analysis accuracy of long-time imaging of living cells.

The application discloses a neuromorphic camera denoising method for general lighting conditions, and belongs to the field of video denoising. According to the relationship between the intensity of the pulse stream noise and the scene lighting, the continuous space-time pulse stream is converted into the picture feature representation under a specific window, and the input of the traditional neural network is adapted; the dark current noise and the response inconsistency noise are removed through the calibrated dark current matrix and the response inconsistency matrix; the image features are converted into the frequency domain features through the two-dimensional discrete wavelet transform, the low-cost noise removal is realized by using the frequency domain noise separation characteristics; the shot noise and the clipping noise are removed by fusing the frequency domain features of adjacent windows; the residual noise on the frequency domain is thoroughly removed through the deep self-attention network; finally, in order to inhibit the excessive smoothing caused by the denoising, the fusion features on the frequency domain are used to refine the denoising result. The application is suitable for the field of video denoising, is used for reconstructing a clean image, reducing the noise in the image and reducing the influence of the noise on the downstream task.

The invention belongs to the information security related fields of identity authentication, anti-counterfeiting, encryption, random number generation and the like, and particularly relates to an ultra-low authentication error rate Hash authentication method based on an optical physical unclonable function (PUF), which comprises the following steps of: obtaining a binary coded image taking zero as a threshold value by using resampling and a DoG filtering method based on two-dimensional speckles generated by the optical PUF; the internal correlation of a binary coded image is reduced through a method of combining image pixels and removing image edge pixels, so that the randomness of a hash key is improved; through adjustment of filtering kernel parameters, low spatial frequency components and high spatial frequency components of speckles collected under different environment noise and shot noise conditions are adaptively filtered. According to the invention, while the randomness of the excitation-response characteristics of the optical PUF is ensured, the authentication error rate can be reduced, and the authentication reliability is enhanced.

The invention relates to the technical field of induction heating temperature control, in particular to an induction furnace heating control method, device and system, and the method comprises the steps: obtaining an infrared thermal imaging image in the heating process of an induction furnace; environmental medium interference is obtained through shot noise features and edge gradient features in the infrared thermal imaging image of the induction furnace; obtaining induction heating sufficiency according to skin effect characteristics and temperature change trend characteristics in the infrared thermal imaging image of the induction furnace; obtaining a heating control factor according to the environmental medium interference and the induction heating sufficiency; and evaluating the overall heating condition of the material in the induction furnace through the heating control factor, and adjusting the temperature of an induction coil of the induction furnace according to the overall heating condition. The invention aims to accurately adjust the temperature of the induction coil of the induction furnace, improve the energy transfer efficiency of the induction furnace, improve the automatic control precision and reduce the energy consumption.