316 results about "Phase spectrum" patented technology

A background noise estimate based upon a modified Doblinger noise estimate is used for modulating the output of a pseudo-random phase spectrum generator to produce the comfort noise. The circuit for estimating noise includes a smoothing filter having a slower time constant for updating the noise estimate during noise than during speech. Comfort noise is smoothly inserted by basing the amount of comfort noise on the amount of noise suppression. A discrete inverse Fourier transform converts the comfort noise back to the time domain and overlapping windows eliminate artifacts that may have been produced during processing.

The invention discloses a multi-speaker speech separation method and system. The method comprises the following steps : acquiring mixed voice signals, obtaining multichannel multi-speaker mixed voicesignals, and scanning the multi-channel multi-speaker mixed voice signals to obtain MUSIC energy spectra; obtaining S peak values from the MUSIC energy spectra, wherein each peak value corresponds toa beam direction; respectively enhancing S beams to obtain mixed voice in S directions; performing short-time Fourier transform on the mixed voice corresponding to each direction to obtain short-timeFourier amplitude spectra of the S target speaker voice, and respectively inputting the short-time Fourier amplitude spectra into a depth neural network to estimate a phase sensing mask correspondingto each target speaker; and performing element-by-element multiplication between the phase sensing mask of each target speaker and amplitude spectra of the corresponding mixed speech to obtain the amplitude spectrum of the target speaker, and the time domain signal of the target speaker is recovered by inverse short-time Fourier transform using the phase spectrum of the corresponding mixed speech.

A sound determination apparatus receives acoustic signals by a plurality of sound receiving units, and generates frames having a predetermined time length. The sound determination apparatus performs FFT on the acoustic signals in frame units, and converts the acoustic signals to a phase spectrum and amplitude spectrum, which are signals on a frequency axis, then calculates the difference at each frequency between the respective acoustic signals as a phase difference, and selects frequencies to be the target of processing. The sound determination apparatus calculates the percentage of frequencies at which the absolute values of the phase differences of the selected frequencies are equal to or greater than a first threshold value, and determines that the acoustic signal coming from the nearest sound source is included in the frame when the calculated percentage is equal to or less than a second threshold value.

The invention is a seismic wave impedance inversion method, determining amplitude and phase spectrums of pos-stack seismic data, and extracting seismic subwaves on a near-well seismic channel and determining amplitude and phase spectrums, spectrum widths and master frequencies of the seismic subwaves; picking up seismic stratigraphic horizons corresponding to geological stratigraphic horizon and checking and correcting as well as smoothing and interpolating the picked horizons; demarcating logging acoustic time difference and density curves, and demarcating depth-domain logging acoustic time difference and density curves into time-domain ones, consistent with horizons of stacked seismic data, generating in-well wave impedances; making wavelet odd analysis and wavelet characteristic extraction on the stacked seismic data; providing high resolution wave impedance parameters for reservoir description in the exploration, development and recovery of oil fields.

A frequency spectrum is detected by analyzing a frequency of a voice waveform corresponding to a voice synthesis unit formed of a phoneme or a phonemic chain. Local peaks are detected on the frequency spectrum, and spectrum distribution regions including the local peaks are designated. For each spectrum distribution region, amplitude spectrum data representing an amplitude spectrum distribution depending on a frequency axis and phase spectrum data representing a phase spectrum distribution depending on the frequency axis are generated. The amplitude spectrum data is adjusted to move the amplitude spectrum distribution represented by the amplitude spectrum data along the frequency axis based on an input note pitch, and the phase spectrum data is adjusted corresponding to the adjustment. Spectrum intensities are adjusted to be along with a spectrum envelope corresponding to a desired tone color. The adjusted amplitude and phase spectrum data are converted into a synthesized voice signal.

The invention discloses a time-varying mixed-phase seismic wavelet extraction method based on time-frequency spectrum simulation. The method is characterized in that on the basis that improved generalized S transformation is carried out on a non-stationary seismic record, a time-frequency filter is introduced for the first time to filter a time-frequency spectrum of the seismic record, then a wavelet amplitude spectrum at each moment is estimated according to the spectrum simulation method, a mixed-phase spectrum of the wavelets is reconstructed in the seismic record with the time-varying wavelet amplitude spectrums removed according to the bispectrum method of high-order cumulants under the assumed condition that the phases of the wavelets are time-invariant, and the amplitude spectrums and the phase spectrum are combined to achieve extraction of the time-varying mixed-phase wavelets. Compared with a conventional stage extraction method, the time-varying wavelet extraction method has the advantages that the assumption that stages of the wavelets are stable is not needed, the time-varying wavelets can be accurately estimated even though the attributes of the wavelets at adjacent stages vary greatly, the mixed phase of the extracted wavelets is more consistent with the reality and seismic data can be analyzed and processed more finely and more accurately. The method has great application value in improving the resolution of the seismic data.

A 2.3 kb/s speech coding technique with harmonic excitation and linear prediction is disclosed. The speech signals are acquired by speech acquisition system and then inputted to the processor of computer. The fundational tone detecting method and voiced sound analysis method are used to split the LPC exciting signal into lower band associated with voised sound and higher band relative to voiceless sound. A united harmonic model with a fixed phase spectrum of voiced sound of man speech is used to express the harmonic component and pseudo-noise component. An interpolation method is used to reduce noise. A short-lag filter is used for high quality of synthetic speech.

Method and apparatus for embedding binary data in audio signals. The magnitude of the power spectrum at the perceptual holes of each frame of a host speech utterance, may be altered so as to embed digital data. The phase spectrum at perceptually masked spectral points of each frame of a host speech utterance, may also be altered so as to embed digital data.

The invention belongs to the technical field of remote sensing image target detection, and in particular relates to a multi-spectral image ship target rapidly detection method based on a selective visual attention mechanism. The method comprises the following steps: constructing a multi-dimensional database into a biquaternion form; and extracting a salient region for detecting a salient object by use of the phase spectrum of biquaternion Fourier transform, thereby achieving integrated treatment of the multi-dimensional data and fully utilizing the information of the spectrum domain and the space domain. By utilizing the method, the defects of the conventional multi-spectral image ship detection method, such as high calculation complexity and complex parameter setting can be effectively overcome. The test results of the analog data and real multi-spectral remote sensing data show that the method has the advantages of good detection result and stronger robustness to noises. The method has great application value in fields of the marine fishery, marine transportation regulation, marine military monitoring and the like.

The invention discloses a frequency-domain-analysis-based method for detecting a region-of-interest of a visible light remote sensing image, belonging to the technical field of remote sensing image processing and image recognition. The method comprises the following implementation processes of: (1) preprocessing the image; (2) carrying out quaternion Fourier transformation on the preprocessed result to obtain the frequency domain information of the image; (3) retaining a phase spectrum and obtaining the high-frequency information of a magnitude spectrum by using a Butterworth high-pass filter; (4) carrying out quaternion Fourier inversion on the phase spectrum and the high-frequency information of the magnitude spectrum to obtain a characteristic pattern; and (5) filtering the characteristic pattern by using a gaussian pyramid and reducing dimensions to obtain a saliency map; (6) carrying out threshold segmentation and binaryzation on the saliency map to obtain a region-of-interest template; and (7) raising the dimensions of the template and carrying out masking operation on an original image so as to obtain the final region-of-interest. The method realizes the rapid and accurate location of the region-of-interest, has the advantages of high accuracy of region description, low computation complexity and the like and can be applied in fields such as environmental monitoring, land utilization, agricultural investigation and the like.

The embodiment of the invention discloses a voice noise reduction method and device, a storage medium, and a mobile terminal. The method includes the steps: receiving a voice signal collected by a single microphone of an earphone, extracting an initial amplitude spectrum and a phase spectrum of the voice signal, and transmitting the initial amplitude spectrum to a pre-trained noise reduction modelto obtain a first noise reduction amplitude spectrum, masking the first noise reduction amplitude spectrum, generating a second noise reduction amplitude spectrum, and recombining the second noise reduction amplitude spectrum and the phase spectrum to generate a first noise reduction speech. Through the above scheme, the method maintains the phase spectrum of the voice signal, and performs the noise reduction of the amplitude spectrum of the voice signal based on the noise reduction model, performs the masking of the processed first noise reduction amplitude spectrum, eliminates the distortion in the noise reduction process, solves a problem that the noise of the voice signal collected in a noisy environment is big, and achieves the quick and precise noise reduction of the voice signal collected by the earphone.

The invention discloses a deconvolution method of multi-path double spectral domain mixed phase wavelets, comprising the following steps: selecting n earthquake records from the collected seismic records; determining the amplitude spectrum of the wavelet according to the selected n earthquake records; determining the phase spectrum of the wavelet according to the selected n earthquake records; implementing the inverse Fourier transform according to the amplitude spectrum and the phase spectrum of the determined wavelet to obtain the mixed phase wavelet of the time domain; determining the zero phase wavelet which is expected to output according to the amplitude spectrum of the expected output wavelet input by users based on the amplitude spectrum of the determined wavelet; adopting the mixed phase wavelet as input and adopting the zero phase wavelet as output to determine the deconvolution filter through the filter wave matching; and obtaining the earthquake records of which the resolution ratio is improved based on the deconvolution filter and the earthquake records.

The embodiment of the invention provides a combined model training method. The method comprises the following steps: extracting the phase spectrum and the logarithm magnitude spectrum of a noisy voicetraining set in an implicit manner; by utilizing the magnitude spectrum fragments of the logarithm magnitude spectrum after expansion as the input features of a time frequency masking network, and byutilizing the noisy voice training set and a clear voice training set, determining a target masking label used for training the time frequency masking network, based on the input features and the target masking label, training the time frequency masking network, and estimating a soft threshold mask; and enhancing the phase spectrum of the noisy voice training set by utilizing the soft threshold mask, wherein the enhanced phase spectrum is adopted as the input features of a DOA (direction of arrival) estimation network, and training the DOA estimation network. The embodiment of the invention further provides a combined model training system. According to the embodiment of the invention, by setting the target masking label, the input features are extracted in an implicit manner, and the time frequency masking network and DOA estimation network combined training is more suitable for the DOA estimation task.

A coating density ultrasonic measurement method based on sound pressure reflection coefficient phase spectrum, belongs to the field of material ultrasonic non-destructive inspection and assessment technique, which uses a spike pulse ultrasonic waterlogging focusing echo system composed of an ultrasonic flaw analyzer, an ultrasonic spike pulse waterlogging focusing probe, a coating sample, a flume, a digital oscilloscope and a computer. First, the system is used to collect one aliasing signal comprising a coating surface sampling signal reflection signal and a matrix boundary reflection signals, then the system is used to collect a surface echo signal of a standardized block, FFT transformation is respectively performed on the collected sample signal and the standardized signal, data of a real part and an imaginary part are obtained, finally coating density is computed. Ultrasound of 1-20 MHz is used to measure the coating density, also is extendible to measure density of homogeneous coating or heterogeneous coating. The invention is not restricted by the material and uniformity, the device is simple, operation is strong, cost is low, practicality is easy, measurement accuracy is high, repeatability is good.

The present invention provides a demodulation method for the optical distance variation of an optical fibre interferometer sensor, including steps of (1) detecting the spectrum of the optical fiber interferometer sensor; (2) transmitting the detected spectrum to a computer for a Fourier transformation or Z spectrum conversion to obtain an amplitude spectrum and a phase spectrum of an optical fiber interferometer sensor reflected signal; (3) obtaining an approx center frequency of the optical fiber interferometer sensor from the amplitude spectrum and then fixing the frequency point to monitor the change of the point phase; obtaining the optical distance change of the optical fiber interferometer sensor by the phase spectrum. By monitoring the phase change around the center frequency point in the phase spectrum after the optical spectrum and the frequency spectrum conversion of the optical fiber interferometer sensor, it is capable of obtaining the optical distance change of the interferometer sensor, comparing with the traditional detection of the change of the intensity frequency spectrum center frequency, the sensor center frequency point of the phase spectrum is more sensitive to the optical distance difference, thus the invention has a high sensitivity or resolution.

The invention discloses a high-speed Fourier stacking imaging device based on annular illumination and a reconstruction method. The device adopts an annular LED board as an illumination light source.The illumination numerical aperture of each LED on the ring is equal to the numerical aperture of an objective lens. Iterative reconstruction is realized through LED unit brightness calibration, LED location calibration and correction, original image acquisition, original image preprocessing and high resolution image initialization in order. According to the invention, the illumination angle of each LED on the ring is inscribed with the numerical aperture of the objective lens, so that a phase transfer function completely covers all low frequency parts of the phase spectrum; simply by taking asmall amount of low-resolution pictures of the bright field, the large-visual-field high-resolution phase distribution of an object can be stably and accurately reconstructed; and the device is suitable for label-free high-speed quantitative phase microscopy imaging of living cells.

An echo processor includes an echo suppression amount decider which compares a state of an outgoing input signal to be sent including an echo with a state of an incoming input signal and calculates an echo suppression amount based on the comparison, a time-frequency transformer which transforms the outgoing input signal in the time domain into the frequency domain, thereby generating an amplitude spectrum and a phase spectrum, a noise-spectrum estimator which estimates a noise spectrum from the amplitude spectrum, a mixer which mixes an amount of the noise spectrum based on the echo suppression amount with the amplitude spectrum, and a frequency-time transformer which generates an outgoing output signal including a noise in the time domain on the basis of the amplitude spectrum, with which an amount of the noise spectrum is mixed by the mixer, and the phase spectrum, and outputs the outgoing output signal.

The invention discloses a method and device for measuring the frequencies of high-frequency cosine signals under the undersampling rate, and relates to the field of digital signal processing. By combining the type of the zero crossing point with the distribution characteristic of apFFT phase spectrums, L spectrum position index values are selected from 2L peak value magnitude spectrum positions; according to the remainder selecting result, each apFFT phase spectrum is corrected through the all-phase ratio interpolation method so that the frequency of the apFFT phase spectrum can be estimated; the frequency estimation value obtained after each apFFT phase spectrum is corrected is used as a remainder, the remainders are processed according to the Chinese remainder theorem of the closed analysis form, and the frequencies of the original high-frequency signals are reconstituted through the frequency estimation values. According to the method and the device, the frequencies of the multiple paths of high-frequency cosine signals under the undersampling rate are measured, and the measuring range of the frequencies of the high-frequency signals is greatly widened; the novel spectrum correction method is adopted, and therefore the measuring accuracy of the frequencies of the high-frequency cosine signals is improved, and novel anti-noise performance of the high-frequency cosine signals under the low undersampling rate is ensured.

The invention discloses a voice enhancement method and system based on phase compensation. The method comprises steps that a to-be-processed noisy voice signal is obtained; short-time Fourier transform of the noisy voice signal is carried out to obtain the amplitude spectrum and the phase spectrum of the noisy voice signal; a phase spectrum compensation function is obtained, and a Sigmoid type function whose factor is correspondingly changed with the change of the noisy voice signal-to-noise ratio is compensated; the phase spectrum of the noisy voice signal is compensated according to the phase spectrum compensation function to obtain the compensated phase spectrum; the amplitude of a pure voice signal is obtained according to the amplitude spectrum of the noisy voice signal; the compensated phase spectrum and the amplitude of the pure voice signal are reconstructed to obtain an enhanced voice signal. The method is advantaged in that compared with a traditional phase compensation voiceenhancement method, estimation of the noise is closer to the real noise power spectrum, the noise in an audio signal can be effectively inhibited, and intelligibility of the voice signal is improvedwhile the quality of the voice signal is enhanced.

The invention discloses a training method of a voice bandwidth expansion model and a voice bandwidth expansion method, wherein the voice bandwidth expansion method comprises the following steps: acquiring to-be-expanded narrowband voice; calculating an amplitude spectrum and a phase spectrum of the to-be-expanded narrowband voice, and extracting auxiliary characteristics of the to-be-expanded narrowband voice ; processing the amplitude spectrum and the auxiliary characteristics of the to-be-expanded narrowband voice by virtue of the voice bandwidth expansion model which is obtained from training, so that the amplitude spectrum of a reconstructed bandwidth voice high-frequency band can be obtained; conducting mirror image reversing on the phase spectrum of the to-be-expanded narrowband voice in a frequency domain, and determining the phase spectrum of the bandwidth voice high-frequency band; and on the basis of the amplitude spectrum and the phase spectrum of the narrowband voice, and in combination with the amplitude spectrum and the phase spectrum of the reconstructed bandwidth voice high-frequency band, determining a bandwidth voice signal. With the application of the voice bandwidth expansion method provided by the invention, an effect of improving the tone quality and naturalness of the narrowband voice can be achieved.

The invention discloses a direction finding device for a phase interferometer and a phase spectrum interferometer based on an orthogonal/non-orthogonal multiple baselines. According to direction-finding algorithms of the phase interferometer and the phase spectrum interferometer, the direction finding device is designed by integrally considering factors such as an actual direction-finding environment, the frequency range of an incident signal, the direction-finding requirement, hardware sources and the like. According to the direction finding device, the processing of a digital signal is cooperatively finished by two FPGAs (Field Programmable Gate Array) and a DSP (Digital Signal Processor); and two direction-finding algorithms of the phase interferometer and the phase spectrum interferometer based on the orthogonal/non-orthogonal multiple baselines are realized respectively for a narrow frequency band single-source signal and a wide frequency band multi-source signal. The direction finding device has the advantages of simple algorithm, high speed of calculation, higher direction finding precision and lower requirement on hardware resources; and the direction finding device can beused in application occasions such as quick acquisition of two-dimensional arrival direction of a single-source or multi-source signal and the like.