96 results about "Iterative filtering" patented technology

The invention concerns sound spatialization with multichannel encoding for binaural reproduction on two loudspeakers, the spatial encoding being defined by encoding functions associated with multiple encoding channels and the decoding by applying filters for binaural reproduction. The invention provides for an optimization as follows: a) obtaining a original set of acoustic transfer functions particular to an individual's morphology (HRIR;HRTF), b) selecting spatial encoding functions (g(θ,φ,n)) and / or decoding filters (F(t,n)), and c) through successive iterations, optimizing the filters associated with the selected encoding functions or the encoding functions associated with the selected filters, or jointly the selected filters and encoding functions, by minimizing an error (c(HRIR,HRIR*)) calculated based on a comparison between: the original set of transfer functions (HRIR), and a set of reconstructed transfer functions (HRIR*) from encoding functions and decoding filters, whether optimized and / or selected.

The invention provides a bearing-only underwater target tracking algorithm based on multilevel information fusion. Three levels are divided. In the first level, a least square method is used for initially tracking and positioning a target, and the estimated state value is calculated as an input value of a second-level AFIMM (Adaptive Fuzzy Interacting Multiple Model). In the second level, the AFIMM is designed, the motion model set of the AFIMM is designed optimally, and a method of a fuzzy adaptive model transition probability is brought forward. In the third level, resampling particle filter is designed as an iterative filtering structure at each step for the AFIMM, and information fusion is realized. During the bearing-only target tracking process for a multi-UUV (underwater unmanned vehicle) cooperative system, through the design of the multilevel information fusion framework, the filter accuracy is higher, and the target of multi-UUV cooperative underwater target tracking can be met.

A persistent filter refining system places a search result comprising objects in a dynamic non-persistent bucket and displays the search result in a dynamic non-persistent bucket display for selection of the objects by a user. The system transfers a selected object from the dynamic non-persistent bucket to a persistent bucket and displays the transferred selected object in a persistent bucket display. The system iteratively provides additional search results to the user for selection and transfer of additional selected objects to the persistent bucket to provide access by a user to take action on the transferred selected object and additional selected objects. In one embodiment, the present system displays one or more additional search results for comparison and selection of objects by the user.

The invention belongs to the technical field of transfer alignment in inertial navigation, and particularly relates to a method for measuring hull deformation angle based on inertia instruments, such as a gyroscope, an accelerometer and the like, and an iterative filtering algorithm. The method comprises the following steps: 1, evaluating an estimated value (FORMULA) of the deformation angle by utilizing a Kalman filtering method; 2, analyzing the power spectrum of an angular velocity observation residual; 3, using the angular velocity observation residual and a specific force observation residual as Kalman filtering observation variables for iterative filtering; 4, through calculation repetition of the second step and the third step, constantly correcting an estimated result in the first step till a corrected value (FORMULA) of the deformation angle is close to but not equal to a true value (FORMULA). By the method, in allusion to the environmental characteristics that a large ship is low in speed and swings weakly, the measurement accuracy of the deformation angle in a vertical direction can be significantly improved; meanwhile, the method is relatively small in calculation amount, and is suitable for online real-time measurement.

A guided trilateral filtering ultrasonic image speckle noise removal method comprises the steps of calculating the space domain distance weight of a guided image through a Gaussian function, and setting the standard deviation of the guided image to be increased along with increase of noise intensity; carrying out Histogram fitting on a local area of the guide image, and selecting a Fisher-Tippettprobability density function selected as a fitting function; Estimating a distribution parameter of the Tippett probability density function by adopting a maximum likelihood method, and calculating adistribution similarity weight according to the estimated parameter; calculating Pixel value difference weight of a guide image by using an exponential function, and setting a scale parameter of the guide image as an estimated Fisher-; Wherein the Tippett distribution parameters are in direct proportion change; And carrying out local iterative filtering on the ultrasonic image by using the three calculated weights, and carrying out iterative convergence to obtain the ultrasonic image with speckle noise removed. According to the method, the filtering weight value is calculated through three aspects of information of the spatial domain distance, the pixel value difference and the distribution similarity, speckle noise can be effectively reduced, meanwhile, detail and edge information of theimage can be better reserved, and therefore visual interpretation of the ultrasonic image is enhanced.

The invention discloses a multi-channel collaborative spectrum sensing method based on iterative filtering. The method includes: time is divided into time slices with fixed length through a reasonably designed frame structure. In each time slice, a secondary user with poor sensing performance is removed preferentially and does not participate in next round of cooperative spectrum sensing, and therefore the number of sensing users is reduced remarkably, and spectrum sensing performance is improved. Through design of the frame structure and a collaborative sensing scheme, the multi-channel collaborative spectrum sensing method based on the iterative filtering can solve problems of too much collaborative spectrum sensing users and malicious user interference existing in a distributed cognitive radio network. Theory analysis and simulation results show that: compared with an existing collaborative spectrum sensing method, the multi-channel collaborative spectrum sensing method based on the iterative filtering can effectively reduce the number of the secondary users participating in collaborative sensing in a loose condition, malicious interference users are removed, and simultaneously spectrum sensing performance of the cognitive radio network is improved.

The invention discloses an ore prospecting method utilizing ground high-precision gravity measurement. The ore prospecting method mainly comprises six methods, namely a wave number domain iteration method of a potential field high-order derivative, a potential field separation method based on an iterative filtering method, a TASD method for identifying a boundary of a geologic body, an Euler deconvolution method for potential field data processing, a potential field abnormal cross-correlation imaging technical method and a gravity man-machine interaction profile forward modeling method. The ore prospecting method utilizing ground high-precision gravity measurement has the advantages of processing and explaining data of a plurality of gravity profiles, adopting the explaining methods covering the wave number domain derivative iteration method, the iterative filtering method of field separation and the cross-correlation imaging technology based on residual anomalies to be applied to gravity data, and achieves a good processing result.

The invention discloses a comprehensive energy system load prediction method considering multivariate load coupling characteristics. The method specifically comprises the following steps: 1) preprocessing historical data of a comprehensive energy system; 2) carrying out multivariate load modal decomposition; 3) establishing a multivariate load multi-task learning prediction model; 4) performing multivariate load prediction error compensation; The historical cold, heat and electric load time sequences are decomposed by adopting a self-adaptive local iterative filtering decomposition method, andthe periodic sequence, fluctuation sequence and trend sequence of each load are obtained through reconstruction on the basis, so that the complexity and non-stationarity of the multivariate load timesequence can be reduced, and the model prediction precision is improved on the data level.

The invention discloses a distributed asynchronous recursive filtering fusion based on track-before-detect technology, and belongs to the technical field of radar data processing. According to the invention, the method herein combines Dynamic Planing-track-before-detect (DP-TBD) algorithm and iterative filtering, and distributed asynchronous fusion algorithm, and increases the capability of DP-TBDalgorithm in detecting weak targets by conducting measurement data fusion on an iterative filter and a plurality of radars. In addition, the plurality of radars often operate at different times and have different scanning periods, which is to say that the plurality of radars have asynchronous data. According to the invention, the method herein achieves the fusion of asynchronous data of differentradars. According to the embodiments of the invention, the method herein overcomes the failure to use DP-TBD algorithm so as to conduct synergetic detection and information fusion in actual application when a plurality of asynchronous radars conduct synergetic detection, and can effectively increase the effects of tracking weak targets.

The invention is suitable for the technical field of intelligent automobiles, and provides an in-vehicle life detection method and device, equipment and a storage medium. The method comprises the steps of: acquiring an xth radar frame signal collected by a preset millimeter wave radar in a vehicle; performing demodulation sampling and time-frequency conversion on the xth radar frame signal to obtain a first range dimension signal array; filtering the first range dimension signal array to obtain a second range dimension signal array; extracting the phase and confidence of a preset detection point in the second range dimension signal array; performing iterative filtering on the phase and confidence of the preset detection point of the xth radar frame signal according to the pre-extracted phase and confidence of the preset detection point of an (x-1)th radar frame signal to obtain a filtering result of the preset detection point of the xth radar frame signal; and when the filtering result is greater than a preset threshold, determining that there is a life at the preset detection point. The in-vehicle life detection method can be suitable for in-vehicle life detection in various scenes, and the detection effect is good.

The invention provides an iterative filtering method for tracking before multiframe detection. A later target state can be estimated more precisely according to a short track obtained through tracking processing before detection, filtering processing is carried out on the target state, the precision of estimating the target state is improved, a target can be tracked in real time for a long time, and the complete track of the target is obtained. The target state is estimated iteratively, and overmuch memory space and the too large calculated amount are not introduced. According to the method, the precision of estimating the target state is improved, the target can be tracked in real time for a long time, the problem that missing inspection possibly happens to tracking before weak target detection with the low signal to noise ratio is effectively solved through track predication, and the continuity of the track is guaranteed.

The invention discloses a diagnostic method for a pump station unit based on a composite characteristic index and a depth limit learning machine and relates to the technical field of fault diagnosis of hydraulic machinery. The diagnostic method can directly decompose nonstationary time series from original vibrating signals by adopting self-adaptive iterative filtering, so that the nonstationary characteristic of the unit can be extracted effectively. A time domain statistics signal, a frequency domain statistics signal, an energy signal, a sample entropy signal and an arrangement entropy signal are extracted based on each nonstationary signal component. Characteristic learning is performed quickly and effectively by means of the depth limit learning machine, implicit fault information ofeach characteristic is extracted and boundedness of a manual design and characteristic extraction and a complex parameter adjusting process based on an artificial neural network are avoided, so that intelligent diagnosis of faults of a water pump unit is achieved, and therefore the accuracy and the effectiveness of fault diagnosis of the water pump unit are improved.

The invention discloses an adaptive-filtering target tracking and positioning method based on an embedded platform, comprising the following steps: (1) obtaining observation data of geographical position of a target aircraft at preset time by the data bus of the Advanced RISC Machine (ARM) embedded platform; (2) performing modeling according to the observation data of geographical position in three-order Lagrange's series in the three-dimensional space motion by a motion modeling module; (3) performing residual data processing for the error value output by the motion modeling module by an information extracting module; (4) performing parameter optimizing for the gain matrix according to the evaluated key parameter of the noise space obtained by the information extracting module; (5) obtaining the filtering data at preset time by an iterative filtering module; and (6) performing coordinate system of the filtering data and outputting the data by an output interface module. The adaptive-filtering target tracking and positioning method of the invention achieves high-stability and performance optimization when keeping low computation complex degree and less resource occupation of a core processor.

The invention discloses an image multistep residual feedback iterative filtering method based on fractional order difference weighting. In the filtering method, a fractional order singularity index calculating unit, a fractional order weight matrix calculating unit and a multistep residual feedback filtering unit are adopted. The method comprises the following steps of: firstly, estimating a fractional order singularity index which corresponds to each pixel point; secondly, generating a fractional order weighting coefficient matrix according to the coefficient calculating way of a fractional order difference format; and lastly, performing multistep residual feedback filtering iteration, updating and generating a middle image to be denoised by using a fractional order weighting combination of a plurality of denoised residual images, performing total variation filtering on the middle image to be denoised to generate an iteration denoised image sequence, and iteratively converging the image into a finial denoised image. Due to the adoption of the method, an iteration sequence can be rapidly converged into a denoised image with a high peak signal noise ratio, the sensitivity and dependency degree of an iteration result on an iteration terminating condition are low, and detailed information such as the textures of images and the like can be well kept while image noise is effectively restrained.

The invention discloses an image noise filtering method via median and mean value iterative filtering of a minimal cross window, and aims at solving the problems that a present image de-noising method is hard to protect edge details and set parameters during de-noising. The image noise filtering method is realized by the steps that (1) noise points are found via an extremum method, and a binary map whose size is the same with that of an image is constructed; (2) the noise density p is calculated; (3) if p < / = 0.525, the iterative median of the minimal cross window is used for filtering, namely, the median of non-noise points in the minimal cross window replaces the value of each noise point in the noise image, iteration is carried out for another two times, whether all the noise points are processed is determined after iteration in each time, and if yes, a de-noising result is output; and (4) if p>0.525, the iterative mean value of the minimal cross window is used for filtering in a way similar to that via the iterative median. The image noise filtering method has the advantages that it is not required to consider the size of windows, operation is easy, the de-noising efficiency is high, details of the image are kept effectively, and the method is more suitable for real-time application.

A persistent filter refining system places a search result comprising objects in a dynamic non-persistent bucket and displays the search result in a dynamic non-persistent bucket display for selection of the objects by a user. The system transfers a selected object from the dynamic non-persistent bucket to a persistent bucket and displays the transferred selected object in a persistent bucket display. The system iteratively provides additional search results to the user for selection and transfer of additional selected objects to the persistent bucket to provide access by a user to take action on the transferred selected object and additional selected objects. In one embodiment, the present system displays one or more additional search results for comparison and selection of objects by the user.

The invention discloses a ground common-frequency multi-motion radiation source tracking method and system based on time-frequency difference and direction finding, and belongs to the technical fieldof positioning. The ground common-frequency multi-motion radiation source tracking method based on time-frequency difference and direction finding includes the steps: establishing a measurement modeland an iterative filtering model for multi-motion radiation source targets of a dual-satellite time-frequency difference positioning system; determining the iterative stop condition of iterative filtering of the filtering model; and realizing tracking of the common-frequency motion source target in an unknown motion state. The ground common-frequency multi-motion radiation source tracking method and system based on time-frequency difference and direction finding can filter all the fuzzy time difference and frequency difference measurement as the target measurement to avoid understanding the complicated problem of the fuzzy, and can determine the optimal motion target filtering estimation result by setting the weight threshold and realize tracking of the common-frequency multi-motion radiation source. Besides, for UHF and L / S frequency band targets that often receive multiple common-frequency signals at the same time, the ground common-frequency multi-motion radiation source tracking method and system based on time-frequency difference and direction finding has great reference significance for improving tracking of low-frequency band multi-motion radiation sources.

The invention provides image processing equipment and an image processing method. The image processing equipment comprises a filtering device for iteratively filtering an input image, and an iterative filtering stopping device for determining whether iterative filtering is stopped or not according to a change rate of a filtered image obtained by filtering with an iterative frequency in an iterative filtering process.

Methods, systems, and computer-readable storage mediums for policy-based testing of computer networks may involve identifying a set of network nodes within a network, and iteratively filtering the set of network nodes in accordance with a testing policy comprising at least one policy rule. Iteratively filtering may include, in an initial iteration, filtering out network nodes that are endpoints of the network from the set of network nodes. Iteratively filtering may further include, for each subsequent iteration, determining a filter criteria from one or more of the policy rules, and filtering out network nodes from the set of network nodes produced by the immediately previous iteration based on the filter criteria. The policy-based testing further comprises invoking a network test scheme that is compliant with the testing policy and is limited to execution against the set of network nodes produced by the iterative filtering.

The invention discloses a method for automatically picking up a seismic reflection event based on Kalman filtering. According to the scheme, the method comprises the following steps of: establishing a state equation and a measurement equation of a Kalman filtering system for picking up seismic reflection event information on the basis of a seismic reflection hyperbolic time-distance equation according to movement rules of seismic reflection waves; detecting the time of arrival of seismic wavelets in each seismic signal, thereby obtaining measurement data for Kalman filtering; and automatically initiating the event according to data in a free measurement point set, and performing gradual iterative filtering according to a sequence of prediction, observation and correction. A data association technology is used for solving the problem of simultaneous tracking of a plurality of events.

The invention discloses a low-illumination color image enhancement method based on a local extreme value, and relates to the field of color image enhancement methods. The method comprises the steps ofstep 1, converting an original color image from an RGB space to a YUV space, and extracting an intensity channel Y of the YUV space as a grayscale image I; step 2, performing iterative filtering on the grayscale image I by using a local extreme value filter with gradually increased cores, and taking a filtering result as an illumination component L of the image; step 3, separating a reflection component R from the grayscale image I according to the illumination component L; step 4, carrying out gamma transformation on the contrast component L, and reconstructing the contrast component L and the reflection component R to obtain an enhanced image. The grayscale image is iteratively filtered by using the local extreme value filter with gradually increased cores, so that the problems of high-brightness region detail loss and low-brightness region contrast enhancement of the enhanced image caused by the existing method are solved, and the color fidelity of the enhanced image, the definition of dark region details and the detail information richness are improved.

The invention discloses a multi-channel collaborative spectrum sensing method based on iterative filtering. The method includes: time is divided into time slices with fixed length through a reasonably designed frame structure. In each time slice, a secondary user with poor sensing performance is removed preferentially and does not participate in next round of cooperative spectrum sensing, and therefore the number of sensing users is reduced remarkably, and spectrum sensing performance is improved. Through design of the frame structure and a collaborative sensing scheme, the multi-channel collaborative spectrum sensing method based on the iterative filtering can solve problems of too much collaborative spectrum sensing users and malicious user interference existing in a distributed cognitive radio network. Theory analysis and simulation results show that: compared with an existing collaborative spectrum sensing method, the multi-channel collaborative spectrum sensing method based on the iterative filtering can effectively reduce the number of the secondary users participating in collaborative sensing in a loose condition, malicious interference users are removed, and simultaneously spectrum sensing performance of the cognitive radio network is improved.

The invention relates to a disoperative target trajectory tracking method under multiple observation nodes. The method comprises the following steps: the locating of a disoperative target is realized through a plurality of signal transmitting points and a plurality of observation receiving points in the sea to generate target locating bright points; the target locating bright points near the base line between each pair of signal transmitting point and observation receiving point are rejected; time alignment is performed on the target locating bright points to obtain the target locating bright point in each pulse period; the clustering process is performed on the bright point obtained in each pulse period through clustering analysis, and the bright points which are away from the class center are rejected through Kalman filtering; the bright point obtained in each pulse period is processed, the remaining bright points are averaged, and segmentation fitting is performed on the average value to obtain a target trajectory; segmentation prediction is performed on the target trajectory again, and the bright point which are away from the predicted trajectory are rejected through the Kalman filtering to realize iterative filtering; and the remaining bright points are averaged, and then segmentation fitting is performed through the least square method to obtain the final target trajectory.

The invention discloses a maximum correlation kurtosis deconvolution method without a period. The method comprises steps of firstly, carrying out truncation and mean removal processing of an acquiredvibration signal; performing Hilbert transform of the vibration signal to obtain an analytic signal of the vibration signal; selecting a maximum value point of the part behind a first zero crossing point in the autocorrelation spectrum as an iteration period to carry out deconvolution operation; then, continuously calculating the maximum value point of the part behind the first zero crossing pointin the autocorrelation spectrum of the signal after each iterative filtering as an iterative period to update a filter so as to obtain an optimal filter coefficient; and lastly, performing envelope analysis of the filtered signal, and extracting the fault characteristic frequency from the envelope spectrum. The method is advantaged in that human participation is not needed in the extraction process of the characteristic frequency, automation of fault characteristic extraction and diagnosis monitoring can be achieved, the time is saved, and efficiency is higher.

In a method and an apparatus for filtering a digital image acquired with a medical device by means of a spatial-frequency operator that automatically allow an optimum filtering of the source image, the integral signal-to-noise ratio of the digital image is calculated, a limit value for the integral signal-to-noise ratio of the filtered image is calculated from the integral signal-to-noise ratio of the digital image, the digital image is iteratively filtered with the spatial-frequency operator R with variation of the dosage g and with calculation of the integral signal-to-noise ratio in the filtered image, afterward iteration of the integral signal-to-noise ratio of the filtered image is compared to the limit value for the integral signal-to-noise ratio, the iterative filtering of the digital image with the spatial-frequency operator R is aborted when the integral signal-to-noise ratio of the filtered image reaches the limit value for the integral signal-to-noise ratio or when a range of control (if present) for the dosage g has been exhausted, and the dosage g most recently employed in the iterative filtering is employed as the optimum dosage gopt for a filtering of the entire digital image under consideration with the spatial-frequency operator R employed.

The invention discloses a data calibration method and device. The method comprises the following steps: establishing a closed loop control system of a filter, performing iterative filtering processing for n times on the to-be-calibrated data by using a self-adaptive filter, wherein the filter parameter used in each iterative filtering processing is the filter parameter obtained after updating the difference between the output data obtained after former iterative filtering processing and the preset reference output data, thereby reducing the error of the output data and improving the data precision.

The invention discloses an optimization method of pipeline surveying and mapping internal inspection IMU data preprocessing. The method comprises the following steps: firstly, determining an estimatedvalue of an initial attitude angle of an IMU by measuring a physical dimension of a ball serving barrel; then performing initial alignment of a static base by using two iterative filtering data to becompared to respectively obtain a course angle and an elevation angle, and calculating alignment errors of the course angle and the elevation angle; and then using a difference value of absolute values of the alignment errors as a criterion to evaluate the effect of the twice iterative filtering. By adoption of the method, the filtering effects of different iterations can be compared to a certainextent to prompt the over-filtering situation and to optimize the preprocessing algorithm. Compared with the manner of directly comparing the matching degree of a solution trajectory and a checkpoint, the method is more efficient, the most of confounding factors are removed from the iteration for terminating the iteration, the method is more convincing, the over-filtering situation is basically avoided, and the optimization of the data preprocessing algorithm is achieved.

The invention discloses a correlation imaging optimization method based on an interpolation algorithm. The correlation imaging optimization method is based on a corresponding imaging idea, a bilinearinterpolation algorithm and a differential correlation imaging scheme. The scheme of the invention effectively combines the bilinear interpolation and differential correlation imaging methods, compresses image information collected by the system by using bilinear interpolation, uses the compressed pixel information to participate in differential correlation imaging, performs time-based corresponding grouping on differential signals by means of the corresponding imaging idea, reconstructs an image of a target object after performing certain iterative filtering, and finally restores the image size to the original size by using the bilinear interpolation algorithm. The interpolation algorithm greatly reduces image data information by processing the gray value, and also improves the image quality, so that the scheme of the invention has the advantages of the convenience of the interpolation algorithm and strong anti-interference ability of correlation imaging.

The invention provides a joint channel and data estimation method and a system for a fast time-varying OFDM channel. According to the method, the system presets a Doppler model for a wireless channel, an input output model formed by at least one tap sequentially arranged based on preset time delay, and a channel state equation for the wireless channel built by a correlation matrix formed based on the Doppler model and all channel impact responses, wherein each tap in the input output model is provided with the corresponding channel impact response; and the state parameter in the channel state equation is formed by each channel impact response; based on the multiple data from the wireless channel and the input output model, a channel measurement equation based on the state parameter for the wireless channel is determined; and the channel measurement equation and the channel state equation are subjected to iterative filtering estimation, and data estimation corresponding to each data without interchannel interference is obtained. The interchannel interference problem in a fast time varying condition can be effectively solved.

The invention discloses a wavelet denoising method based on an adaptive threshold. The wavelet denoising method comprises the following steps: (1) performing wavelet decomposition on a noise-contaminated signal to obtain wavelet coefficients of each layer and a highest-layer scale coefficient; (2) obtaining wavelet coefficient adaptive thresholds of all layers through iterative filtering; (3) extracting a pulse corresponding coefficient from each layer of wavelet coefficients by utilizing a chi-square energy window method; (4) setting a non-pulse corresponding coefficient in each layer of wavelet coefficient to zero to obtain a processed wavelet coefficient; and (5) performing reconstructing by using the processed wavelet coefficient of each layer and the scale coefficient of the highest layer to obtain a denoised signal.