45 results about "Frequency domain algorithm" patented technology

The invention discloses a brain electrical signal independent component extraction method based on convolution blind source separation. The brain electrical signal independent component extraction method based on the convolution blind source separation includes concrete steps: building a brain electrical signal independent component extraction system based on the convolution blind source separation, which comprises an AD (analog to digital) sampling module, a short time Fourier transformation module, a frequency domain instantaneous blind source separation module, a sequence adjustment module and a short time inverse Fourier transformation module; using the AD sampling module to sample brain electrical signals; using the short time Fourier transformation module to transform the brain electrical signals from a time domain to a frequency domain; using the frequency domain instant blind source separation module to separate instantaneous mixing signals in the frequency domain; using the sequence adjustment module to perform sequence adjustment on independent components in a vector on each frequency domain segment; using the short time inverse Fourier transformation module to transform a frequency domain separation result into an independent component on the time domain. The brain electrical signal independent component extraction method based on the convolution blind source separation extracts the independent components of brain electrical signals based on a true convolution mixing model, uses a convolution blind source separation frequency domain algorithm, and is simple to achieve, good in separation effect, and low in calculation complexity.

The invention belongs to the field of optical communication, and is applied to the polarization demultiplexing of a polarization diversity multiplexing coherent optical communication system. According to the invention, for the problem of the polarization demultiplexing in the polarization diversity multiplexing coherent optical communication system, a special training sequence and related algorithms are designed. Through the method, a coherent receiver can fast, automatically and accurately perform the polarization demultiplexing. As the track of the training sequence, viewed from a planisphere, rotates along a unit circle after sign power is normalized, the training sequence provided by the invention is called a constant rotation sign training sequence. The invention discloses two processing algorithms for the constant rotation sign training sequence, which are a frequency-domain algorithm and a time-domain algorithm respectively, wherein the frequency-domain algorithm can be combined with frequency-domain dispersion compensation without adding a frequency spectrum calculation module. Through the method, the receiver does not need synchronizing with an initial position of the training sequence.

The invention relates to a large frequency deviation estimation and correction method for quadrature frequency division multichannel signal carriers, comprising the following steps: time domain frequency deviation estimation of frequency deviation, and time domain frequency deviation correction according to estimation results. The invention also comprises the following steps: smooth processing of the time domain frequency deviation estimation is performed, namely a low pass filter is adopted to smooth down peak values and valley values of time domain frequency deviation estimation values; after the time domain frequency deviation correction, Fourier transformation is performed; frequency domain frequency deviation estimation is performed on transformation results; frequency domain frequency deviation estimation results or time domain frequency deviation estimation results after smooth processing participate in time domain frequency deviation correction once again. The method of the invention overcomes limitation of the prior frequency domain algorithm on frequency deviation estimation and has the advantages of capability of processing large frequency deviation, capability of guaranteeing smooth signal processing, capability of quickly tracking frequency deviation caused by movement, and capacity of being conveniently realized in an integrated circuit design.

The invention discloses a vehicle-induced bridge transverse vibration reaction calculation method based on a frequency domain method, provides a novel railway bridge transverse vibration response determination method. Compared with a traditional time domain analysis method, the novel railway bridge transverse vibration response determination method has the advantages that a train load spectrum issubdivided into a load sequence spectrum and a bridge harmonic modal spectrum, the influence relation of transverse simple harmonic frequency, train speed, a load sequence and different load amplitudes on the spectrums is researched, and calculation is simplified theoretically; transverse vibration of an axle is directly decoupled, bogie force is simplified for wheel set transverse force, and random vibration analysis is facilitated by adopting a frequency domain algorithm according to the power spectral density of the wheel set transverse force; therefore, the method provided by the inventionis small in calculation amount and high in calculation efficiency; and based on the determined modal number, the calculation precision is relatively good.

A method and system of the present inventions reduces both near-end crosstalk (NEXT) and far-end crosstalk (FEXT) interferences due to handshake tones into upstream and downstream neighboring services, both at the customer premises equipment (CPE) and at the central office (CO), when handshake is experienced. An embodiment of the present inventions is directed to significantly reduce the NEXT and/or FEXT interferences due to handshake tones. For example, NEXT and/or FEXT interferences due to G.994.1 handshake tones, identified as, bins 7 and 9 for the Upstream channel and bins 12, 14 and 64 for the Downstream channel may be reduced. An embodiment of the present inventions provides an algorithm that may be used for both NEXT and FEXT Handshake Interferences reduction at the CO and at the CPE. In addition, the algorithm may operate in the time domain and in the frequency domain. Frequency domain algorithm offers two options, both versions taking advantage of the high correlation time of the Handshake tone signals.

The invention discloses a method and device for detecting seeding quality of a duckbill drum type seeding machine, and the detecting method comprises the following steps that an off-line method is used to establish a seed vector signal characteristic vector library; during normal operation of the seeding machine, the signal conditioning circuit collects capacitance information of the detecting capacitance sensor and the reference capacitance sensor when the seed passes through the detection field of the detecting capacitance sensor and performs differential operation to obtain differential capacitance information, and the wireless transmission module transmits the differential capacitance information to the vehicle-mounted terminal; the vehicle-mounted terminal extracts the peak value, pulse width and peak area of the signal by using a frequency domain algorithm according to the differential capacitance information, matches with the seed response signal characteristic vector library, and combines the seed variety, bulk density and water content information to determine whether the seed is a single particle or double particles; and if double particles, the vehicle terminal prompts an alarm and records the positioning information corresponding to the double particles. The invention also provides a detecting device for implementing the above detecting method.

The invention belongs to the field of signal and information processing, particularly relates to a signal processing method, system and device based on a phased array incoherent scattering radar, andaims to solve the problems of low real-time performance and accuracy of signal processing of the existing incoherent scattering radar. The method comprises the following steps: acquiring an echo signal scattered by an ionized layer to aquire an IQ digital signal as a first signal through down-conversion, AD sampling and digital orthogonal down-conversion; performing complex weighting and summationoperation on the first signal to obtain a second signal; removing clutters of the second signal through a preset clutter removing method, and obtaining autocorrelation data through a frequency domainFFT algorithm; circularly acquiring autocorrelation data, accumulating the autocorrelation data, removing background noise after accumulation, and performing calibration correction through a preset data calibration method and a spectrum ambiguity function; and fitting the calibrated and corrected data with theoretical autocorrelation data to obtain ionospheric parameters. According to the invention, the real-time performance and the accuracy of signal processing of the incoherent scattering radar are improved.

The invention discloses a Staggered-SAR echo signal reconstruction method, and the method employs fonformal Fourier transform to process a radar signal after distance compression. The method gives full consideration to the characteristics of Staggered-SAR echoes, and can obtain a uniform azimuth frequency spectrum suitable for the high-quality imaging of a radar. The frequency spectrum obtained through the method can accurately reflect the frequency spectrum characteristics of azimuth echoes under the conditions of data loss and non-uniform sampling, and finally can be used by a Staggered-SAR system for high-quality imaging. Compared with a conventional method, the method can obtain the uniform azimuth frequency spectrum, is suitable for a classical frequency domain algorithm, and can obtain a high-quality imaging result.

The invention relates to an echo cancellation method and system based on an all-phase multi-delay partitioning frequency domain. The method comprises the following steps: receiving remote audio signals, performing sample processing to obtain continuous remote audio sampling signals, and separately storing the continuous remote audio sampling signals; receiving near-end audio signals, performing sample processing to obtain continuous near-end audio sampling signals, and separately storing the continuous near-end audio sampling signals; respectively integrating the remote audio sampling signals and the near-end audio sampling signals, and respectively performing all-phase processing to obtain remote signals and near-end signals; performing Fourier transform processing on the remote signals and the near-end signals, and performing adaptive echo cancellation processing on the remote signals and the near-end signals converted into the frequency domain by using a multi-delay partitioning frequency domain algorithm to obtain audio signals; and encoding the audio signals, and sending the audio signals to a remote receiving end. Therefore, the spectrum leakage problem resulting from that the multi-delay partitioning frequency domain algorithm needs to be converted from the time domain into the frequency domain is solved, the echo estimation is better, and a better echo cancellation effect is achieved.

The invention belongs to the field of rail traffic rail detection and provides a rail traffic rail detection system based on acoustic waves. The rail traffic rail detection system based on acoustic waves comprises an impact hammer, a sound pick-up, a signal amplifier and a microprocessor; the impact hammer is a metal hammer; by using the impact hammer to strike a rail, the rail produces a sound; the sound pickup receives the sound produced by the rail; after amplifying a sound signal received by the sound pickup, the signal amplifier outputs the sound signal to the microprocessor; the microprocessor calculates the received signal by using a time domain-to-frequency domain algorithm to obtain frequency spectrum data of the received signal; the calculated frequency spectrum data are comparedwith hub frequency spectrum saved in the system; and therefore, rail cracks and screw loosening are detected.

The invention discloses a modified geosynchronous orbit SAR (synthetic aperture radar) frequency-domain imaging method. A part with zero Doppler frequency rate can be effectively imaged by the method, with good focusing effect; meanwhile, for the part with non-zero Doppler frequency rate, the method allows better focusing performance than the traditional frequency-domain algorithm does. According to the method, the Doppler frequency rate caused by an acceleration component along a slant range direction is compensated, nonlinear azimuth time-frequency relation of raw echo is converted into linear time-frequency relation, and thus, in following operations, the SAR frequency-domain imaging algorithm can be used to correct a range migration in range Doppler. Meanwhile, an acceleration value subjected to compensational optimization is selected by numerical retrieval, variance of azimuth is greatly reduced, and focusing of a larger scene can be achieved during azimuth focusing. Verifications show that the method is suitable for imaging different parts of a GEO (geosynchronous orbit) SAR rail.

The main feature of the invention includes following points: after called party responses the calling party, the both parties firstly initialize the logical connection and negotiates parameters required by security communication containing voice name, frequency domain algorithm and compression algorithm of voice in security transmission; the calling party transmits the security-transmitted voice that is hidden in the voice in public transmission; when the public-transmitted voice reaches the called party, the security-transmitted voice is restored and transmitted to the called party.

The embodiment of the invention provides an image processing method and device, a server and a storage medium. The method comprises the steps of acquiring a watermark image and an original image; processing the original image based on a preset first frequency domain algorithm to obtain a first frequency spectrum matrix; processing the watermark image based on the first frequency domain algorithm to obtain a second frequency spectrum matrix; embedding the second frequency spectrum matrix into the first frequency spectrum matrix to generate a third frequency spectrum matrix; and processing the third frequency spectrum matrix based on a preset second frequency domain algorithm to obtain a target image. Therefore, based on the image copyright protection of the hidden watermark, the hidden watermark information is added in the frequency domain, operations such as shearing, smearing, compressing and filtering can be resisted, and the image robustness is effectively improved.

The invention discloses a frequency domain active noise control system based on coefficient smoothing, which comprises a rear-end control filter, a coefficient smoother and a front-end control filter which are connected in sequence, and is characterized in that the coefficient smoother is used for smoothing the time domain coefficient of the front-end control filter according to the obtained time domain coefficient of the rear-end control filter; and a front-end control filter is updated to limit system noise caused by overlarge coefficient change, so that the stability and the effectiveness of the active noise control system are improved. According to the invention, the problem of noise occurring in an active noise control system using a frequency domain algorithm is solved.

The invention discloses an information hiding method and device and a computer storage medium, relates to the technical field of information processing, and solves the technical problem of low confidentiality of hidden information. The information hiding method comprises the following steps: acquiring a first image and a second image; processing the first image through a frequency domain algorithm to obtain a grey-scale map; carrying out gray processing on the second image according to the gray-scale map, and carrying out sub-affine transformation scrambling on the image subjected to the gray processing to obtain a third image; and embedding the third image into the grey-scale map, and obtaining a fourth image hiding the second image through a frequency domain reduction algorithm.

The invention discloses a target fruit identification method in a night vision environment, and the method comprises the following steps: 1, obtaining an environment image with the target fruit in the night vision environment, carrying out the noise reduction of the environment image through employing a frequency domain algorithm based on edge direction features, and obtaining a denoised environment image; 2, performing target fruit recognition from the denoised environment image by adopting an a * CbR fusion space method to obtain a target fruit image; and step 3, optimizing the circle center and the corresponding radius of the target fruit image obtained in the step 2 by adopting a three-point circle setting method improved by a genetic algorithm to obtain a final target fruit recognition result.