160 results about "Zero-crossing rate" patented technology

The invention discloses a method for identifying local discharge signals of a switchboard based on a support vector machine model. The method comprises a model training process and an audio identifying process, and particularly comprises the following steps of: preprocessing audio signals; extracting effective audios according to short-time energy and a zero-crossing rate; segmenting the effective audios and extracting characteristic parameters such as Mel cepstrum coefficients, first order difference Mel cepstrum coefficients, high zero-crossing rate and the like of each segment of the audios; training a sample set by using a support vector machine tool, and establishing a corresponding support vector machine model; after preprocessing audio signals to be identified and extracting and segmenting the effective audios, classifying and identifying segment-characteristic-based samples to be tested according to the support vector machine model; and post-processing classification results, and judging whether partial discharge signals exist. By using the method, the existence of the partial discharge signals of the switchboard is accurately identified, the happening of major accidents involving electricity is prevented and avoided, economic losses caused by insulation accidents are reduced, and the power distribution reliability is improved.

The invention discloses a method and system for detecting abnormal sounds. The method includes the steps that short-time energy of each frame of a collected audio signal is compared with a first short-time energy threshold value, if the short-time energy is larger than the first short-time energy threshold value, the corresponding frames are recorded to be first-grade frames, if the short-time energy is smaller than the first short-time energy threshold value, the short-time energy of each frame of the collected audio signal is compared with a second threshold value or the zero-crossing rate of each frame of the collected audio signal is compared with a zero-passing rate threshold value, the frames with the short-time energy larger than the second short-time energy threshold value or the zero-passing rates larger than the zero-passing rate threshold value are recorded to be second-grade frames, and when the number of the continuous frames which are the first-stage frames or the second-stage frames is larger than N and the current frame is the first-stage frame, it is judged that the sounds are abnormal. By means of the method, the abnormal sounds are judged by calculating the short-time energy and the zero-passing rates; as the short-time energy and the zero-passing rates belong to the time domain characteristics and are not related to frequency domain conversion and characteristic parameter calculating, the calculating complexity can be lowered. Meanwhile, as audio information collected in real time is processed, real-time processing and real-time analysis can be carried out, and abnormality can be judged in time.

The invention relates to the technical field of speech, in particular to a real-time speech endpoint detection method and device. The method comprises the following steps that signal framing and emphasis are carried out; pulse removal processing is carried out; direct current components are removed; the short-time energy and zero-crossing rate of each frame of signal are calculated; windowing processing is carried out; spectrum reduction processing is carried out; spectral entropy is calculated; transformation smooth spectral entropy is calculated; a speech frame and a noise frame are preliminarily judged; the transformation smooth spectral entropy and a threshold are processed; a start frame and end frame in a speech segment are judged. The real-time speech endpoint detection method and device have the advantages that according to the conditions under which a signal is judged and a judged result, thresholds of parameters, such as a spectrum reduction threshold, the transformation smooth spectral entropy, the corresponding short-time energy, corresponding short-time average energy and a spectrum reduction power spectrum are weighted and updated, so that the thresholds are more andmore accurate, and finally the judged speech start frame and end frame are also more and more accurate; the method can efficiently and accurately detect speech in real time.

The invention discloses a heterogenous audio frequency splicing tampering blind detection method based on a mute segment; the method comprises the following steps: enframing and windowing a to-be-tested audio frequency; carrying out threshold determination on audio frequency frame frequency spectrum energy and zero-crossing rate, and detecting the mute segment; calculating mute segment audio frequency characteristics; using a slide window to solve an adjacent mute frame audio frequency characteristic correlation coefficient vector on the mute segment; detecting a tempering point; determining a tampering position. The method needs not to use the digit watermark as embedded information, and can effectively determine whether tampering exists or not and locate the specific tamping position while aiming at the heterogenous audio frequency splicing and insert operations on the time domain.

The invention discloses an isolated word speech recognition method based on an HRSF (Half Raised Sine Function) and an improved DTW (Dynamic Time Warping) algorithm. The isolated word speech recognition method comprises the following steps that (1), a received analog voice signal is preprocessed; preprocessing comprises pre-filtering, sampling, quantification, pre-emphasis, windowing, short-time energy analysis, short-time average zero crossing rate analysis and end-point detection; (2), a power spectrum X(n) of a frame signal is obtained by FFT (Fast Fourier Transform) and is converted into a power spectrum under a Mel frequency; an MFCC (Mel Frequency Cepstrum Coefficient) parameter is calculated; the calculated MFCC parameter is subjected to HRSF cepstrum raising after a first order difference and a second order difference are calculated; and (3), the improved DTW algorithm is adopted to match test templates with reference templates; and the reference template with the maximum matching score serves as an identification result. According to the isolated word speech recognition method, the identification of a single Chinese character is achieved through the improved DTW algorithm, and the identification rate and the identification speed of the single Chinese character are increased.

The invention discloses a double-talk detection and echo cancellation method based on zero-crossing rate, comprising the following steps: S1, zero-crossing rate calculation and double-talk monitoring: different echo cancellation strategies are adopted for double-talk and single-talk scenarios; S2, echo filter estimation and echo cancellation; there is difference between an echo signal collected by a microphone and a far-end signal because of room impulse response, and the echo component is eliminated from the signal collected by the microphone; and S3, recovery of a target speech harmonic structure. A harmonic structure analysis method is used to compensate for the missing harmonic components of near-end speech to further inhibit speech distortion. Compared with the traditional echo cancellation technology, double-end monitoring is realized based on the zero-crossing rate, and elimination of target speech is avoided. Based on the criterion of frequency-domain minimum mean square error, an echo cancellation filter converges rapidly. By adopting a frequency-domain parallel processing framework, the complexity is low. A speech distortion inhibition module is added, and the distortion degree of target speech is reduced.

The invention discloses a self-adaptive endpoint detection method and a self-adaptive endpoint detection system for isolate word speech recognition. The self-adaptive endpoint detection method for isolate word speech recognition comprises the following steps: a, a voice input step, wherein a voice signal containing an isolate word to be recognized is input; b, a voice preprocessing step, wherein the voice signal is subjected to amplitude translation and normalization and framing processing operation, and short time average energy and a short time average zero-crossing rate of each frame of voice are calculated; c, an isolate word endpoint rough detection step, wherein isolate word endpoints are roughly estimated through utilization of the short time average energy and the short time average zero-crossing rate of each frame of the voice signal and constraint on the shortest length of continuous voice frames before and after the end points, d, a detection threshold self-adaptive adjustment and accurate endpoint detection step, wherein through utilization of constraint on the smallest time duration and the largest time duration of the isolate word, the detection threshold is subjected to dynamic adjustment operation, the voice endpoints are subjected to front and back fine adjustment, and accurate isolate word endpoints are obtained; e, an isolate word endpoint output and isolate word voice recognition step, wherein the accurate isolate word endpoints are output and isolate word recognition is realized by using voice recognizing technologies.

The invention provides an encoding method and a decoding method of voice frequency data. The encoding method comprises acquiring original voice frequency, carrying out endpoint detection through short-time energy and a short-time zero-crossing rate, eliminating non-voice-frequency data in the original voice frequency, and then acquiring voice section data; collecting characteristic parameters from the voice section data, recognizing statuses of each frame of the voice section data through trained hidden Markov model and Viterbi algorithm, and then confirming status switch and status duration; and encoding the status switch and the status duration, and then a voice frequency data package is generated. The encoding method and the decoding method of the voice frequency data can maintain high voice quality under the condition of low encoding rates.

The invention relates to a method for analyzing similarity based on voice features, which belongs to the technical field of audio signal processing. To compare the similarity of two audios to be tested, the method uses the amplitude and the zero-crossing rate in physical characteristics as basic parameters to compare the audio similarity degree and compares three physical feature parameter algorithms including waveform comparison, envelope comparison and zero-crossing rate comparison. The similarity value is calculated through a correlation function; a similarity threshold is set; the similarity value and a similarity threshold are compared to determine the similarity. The method can be used for comparing the similarity of audio signals and can be applied to the monitoring of broadcast television signals. Compared with the prior art, the algorithm of the invention is simple, the theory is clear, and the technology is easy to implement.

The invention discloses a fundamental frequency detection method based on improved experience wavelet transformation. The method comprises steps of step 1, preprocessing a speech signal: calculating short-term energy and zero-crossing rate of the speech signal, using a dual-threshold method to carry out rhythm segmentation and passing the segmented signal through a 50-1500Hz band-pass filter for filtering so as to obtain pre-processed speech signal; step 2, using the improved experience wavelet transformation method to decompose the preprocessed speech signal so as to obtain each mode function of the speech signal; step 3, according to the mode function, selecting the main mode of the speech signal; step 4, using Hilbert transformation to solve a transient fundamental frequency value of the main mode; and step 5, using a rectangular window function to carry smoothing processing on the transient fundamental frequency value obtained in the step 4 to finish the fundamental frequency detection. The method is characterized by high accuracy, quite good robustness and high time resolution.

The invention discloses a voice signal endpoint detection method based on a characteristic value code. Characteristic parameters including a short-time energy and a short-time zero-crossing rate are extracted in frames, an average value of the short-time energy, an average value of the short-time zero-crossing rate and a maximal value of the short-time zero-crossing rate are calculated, four thresholds are set for the short-time energy via a statistical result and empirical parameters, one threshold is set for the short-time zero-crossing rate, a voice characteristic is coded according to the thresholds, and endpoint detection is carried out on a voice signal according to five-grade determination rules on the basis of the characteristic value code of each frame. The lowest threshold is set for the short-time energy of an audio segment, suspected voices are accepted/rejected according to the rules by combining characteristics of the adjacent frames, the five-grade determination rules can handle with different complex conditions effectively, false detection and neglected detection are avoided, and the correct rate of endpoint detection of the voice signal can be improved substantially.

The syllable division of Tibetan speech is an important part and essential link in Tibetan speech information processing, and lays the foundation for Tibetan speech recognition and speech database making; however, a software for the syllable division of Tibetan speech is still absent. Through extracting the time domain parameter short-time energy and short-time zero-crossing rate of Anduo Tibetan speech, the invention realizes Tibetan syllable division through a specific algorithm. The invention is used to carry out syllable division of 30 continuous Anduo Tibetan speech samples, and the division accuracy reaches to 30.6 percent.

A method and apparatus to detect voice activity by using a zero-crossing rate includes removing noise included in an audio signal, adding a random signal having energy of a predetermined size to the audio signal from which noise is removed, extracting predetermined voice detection parameters from the audio signal to which the random signal is added, and comparing the extracted predetermined voice detection parameters with a threshold value and determining voice and non-voice activities.

The invention discloses an audio tampering detection device based on time-frequency domain joint features. The device includes an RFID authentication module mainly used for identity authentication; a signal transmission module mainly used for communication with the RFID authentication module and a signal processing module; a signal processing module used for recording device recognition and voice tempering detection on received voice information, wherein voice tempering detection is mainly realized through a tempering detection algorithm of zero insertion for zero crossing rate, EM (Expectation Maximization) algorithm-based resampling voice tempering detection and LPC model-based initiative forensicvoice tempering detection; and recording device recognition is realized through BP neural network-based multimedia mobile equipment audio tempering recognition and detection. The device provided by the invention can store frequency spectruminformation in real time and has distinctive advantages in aspects of interference rejection, multi-band frequency and precision. The device provided by the invention has high economical efficiency, reliability and practical applicability.

The invention discloses a baseline drift correction method for electrocardiosignal. The method includes the steps of conducting improved self-adaption noise set empirical mode decomposition on the original electrocardiosignal to obtain intrinsic mode functions and residual components, counting the zero-crossing rates of all the intrinsic mode functions and all the residual components, and removing the intrinsic mode functions with the zero-crossing rates smaller than the set threshold value and the residual components with the zero-crossing rates smaller than the set threshold value from the original electrocardiosignal to obtain the electrocardiosignal where baseline drift is removed. According to the method, by conducting improved self-adaption noise set empirical mode decomposition on the original electrocardiosignal, the mode aliasing phenomenon is eliminated, and residual noise is reduced; the step of conducting self-adaption baseline drift amount selection according to the zero-crossing rates is added, and therefore the problem of the lack of effective baseline drift component selection means in the prior art is solved. The method can be widely applied to the field of baseline drift correction of the electrocardiosignal.

The invention discloses a human voice highlighting processing method and device in an audio. The method comprises the following steps: carrying out framing processing on audio signals to obtain audio frame signals; grouping the obtained audio frame signals, wherein every N audio frame signals are grouped into one group, analyzing characteristics of frequency band range, frequency band energy, low-energy frame rate and zero-crossing rate of the audio frame signals in each group, and determining whether the audio frame signals in each group have human voice according to the analysis result of each group, N being a positive integer larger than 1; and if the audio frame signals in some group have human voice, carrying out band-pass filtering on the audio frame signals in the group to output a first audio signal obtained after filtering. According to the human voice highlighting processing method and device, the human voice can be accurately identified and the identified human voice is highlighted.

The invention relates to a single station infrasonic wave signal recognition and extraction method, which comprises the steps that a single infrasonic sound monitoring station receives infrasonic wave signals; abnormality recognition is performed on the received infrasonic wave signals based on four elements such as the energy flow peak, the zero-crossing rate, the wind speed and noise and wavelet time-frequency characteristics so as to find out a time period in which an abnormal moment is distributed, thereby acquiring four groups of time period sequences; comparative analysis is performed on the four groups of time period sequences to find out overlapped time periods in the four groups of time period sequences, the overlapped time period sequence is marked on the waveform of the infrasonic wave signals so as to be used for indicating the waveform and the location of valid event suspected signals, and then further recognition and classification are performed on the valid event suspected signals by using valid event historical signal characteristics, so that the type or the source of the infrasonic wave signals is judged.

The invention discloses a noise diagnosis method for an electrical equipment break-down arc. The method comprises the following steps that A, an arc sound signal of the electrical equipment is collected, and pre-emphasis processing, FIR digital filtering and framing are carried out on the arc sound signal; B, short-time energy and short-time zero-crossing rate of each frame of the arc sound signalare calculated, short-time average energy and short-time average zero-crossing rate are determined, and an arc sound signal abnormal interval is detected by adopting a dual threshold determinate method; C, filtering is carried out on the arc sound signal abnormal interval by adopting M FIR filters, an energy value of each frequency domain sub-band is calculated to form an M-dimensional feature vector; D, a primary linear kernel model is established by using sample set data and an optimal parameter of a kernel function, an optimal linear kernel model is obtained by modifying the parameter repeatedly, the M-dimensional feature vector is diagnosed by using the optimal linear kernel model to determine whether or not a running state of the electrical equipment is normal. The noise diagnosis method for the electrical equipment break-down arc has the advantages that on-line detection and real-time alarming of the running state of the electrical equipment are achieved, and the damage and probability of the break-down arc are reduced.

The invention discloses a pattern recognition method for a fiber-optic distributed disturbance sensor. The method includes: sampling an output signal I1 (t) of the fiber-optic distributed disturbance sensor for M times according to periods of time so as to obtain N sample points divided according to the periods of time, N >/=M; calculating zero-crossing rate F(i) of the sample point in the i-th period of time, and calculating equivalent frequency f(i) in the current period of time according to the zero-crossing rate F(i) to obtain frequency and time distribution characteristics of the sensor output signal in the current period of time; and recognizing patterns according to the frequency and time distribution characteristics of the sensor output signal. By the use of the method, recognition efficiency and accuracy are increased, and false alarm rate is reduced.

The invention discloses a distributed optical fiber sensing positioning method based on zero crossing point analysis. Frequency distribution of disturbance frame signals is obtained with an improved zero crossing point analysis method, and different from zero crossing rate analysis, the distributed optical fiber sensing positioning method is characterized in that average zero crossing amount is not directly calculated, the positions of zero crossing points in the signals are firstly found, the frequency of a certain section of the signals is estimated by comparing signal point numbers between adjacent zero crossing points, and the smaller the signal point numbers between the adjacent zero crossing points are, the higher the frequency of the section of the signals is. Points next to the maximum frequency point serve as effective signal sections to carry out cross-correlation time delay estimation, the effective signal sections contain high-frequency information of the signals and have the wide frequency bandwidth characteristics, and therefore the positioning error is small; meanwhile, due to the fact that the signal processing method based on zero crossing point analysis only contains simple calculation such as adding and subtracting, the calculation complexity is low, the signals with cross-correlation time delay estimation are only small parts of effective sections of a signal, and the signal processing speed is remarkably increased.

The invention discloses a method for reducing breathing noises in a breathing mask. The method comprises the following steps of: (a) sampling a voice signal in the breathing mask, and dividing the voice signal to voice frames of 10 to 30 ms; (b) calculating the short-time energy and zero-crossing rate of each voice frame; (c) determining the voice frame to be the breathing noise frame when both the short-time energy and the zero-crossing rate of the voice frame are higher than or equal to a predetermined threshold level, and resetting the voice frame; and (d) combining the voice frames which are partially superposed after treatment to obtain the voice signal, in which the breathing noises are eliminated. The method provided by the invention can determine the noises through combining the short-time energy and the zero-crossing rate of the voice frames to reduce or eliminate the breathing noises, so that the method not only can obviate the speech distortion of users wearing the mask but can obviate the extra hardware refit of the existing mask, thereby lowering the cost.

Methods and apparatus for monitoring audio quality and controlling noise in telephone conferences. In one example, voice activity detection processing is implemented based on the energy in the audio signal of each channel and the zero crossing rate of the audio signal to differentiate noise from speech. Certain examples further include the use of a voice quality index and activity timer to determine audio quality.

The present invention discloses an MIMO-OFDM (multiple input multiple output-orthogonal frequency division multiplexing) detection method for sub-carrier grouping and a broadband wireless communication system adopting an MIMO-OFDM technique, which relate to an MIMO-OFDM detection algorithm used in the system. In the present invention, the frequency domain zero-crossing rate range and sub-carrier group interval table of a typical frequency selective channel is set by off-line optimization with a computer emulation method in the design phase of an MIMO-OFDM detector; a frequency domain zero-crossing rate is calculated in real time according to the estimated coefficient of the channel during MIMO-OFDM on-line detection, relative sub-carrier group interval is obtained by referring to the table, and MIMO balance processing is conducted according to the group interval. The present invention largely reduces the calculation amount of detection algorithm and solves the problem that the existing MIMO-OFDM detection algorithm has large calculation amount.