75 results about "Maximum a posteriori estimator" patented technology

The invention discloses a non-partial regularization prior reconstruction method for a low-dosage X-ray captive test (CT) image. The method comprises the following steps of: (1) acquiring a previously scanned standard dosage image of a patient by X-ray CT imaging equipment; (2) acquiring CT projection data of the patient by the X-ray CT imaging equipment under a Low-mAs scanning protocol, and simultaneously acquiring a corresponding correction parameter and a system matrix; (3) constructing a math model for image reconstruction according to statistical distribution met by the projection data acquired in the step (2); (4) constructing a non-partial regularization prior guided by the previously scanned standard dosage image in the step (1), performing model transformation by adopting a maximum posterior estimation method, and constructing a target function for image reconstruction according to the math model obtained in the step (3); and (5) calculating the target function for CT image reconstruction, which is constructed in the step (4), by adopting an iteration algorithm to finish image reconstruction. By the method, the low-dosage CT image can be reconstructed under the Low-mAs scanning protocol.

A timing error recovery system includes a phase locked loop that receives a continuous time input signal, samples the input signal at a sampling rate and generates a voltage control signal. A statistical estimator, such as a maximum a posteriori estimator, compares the voltage control signal with an expected error based upon a statistical model and produces an adjusted voltage control signal that drives a voltage controlled oscillator to adjust the sampling rate.

The invention discloses a maximal posteriori optimized image reconstruction method for leading in a general Gibbs experiment in the PET imaging. The method comprises the following procedures: (1) PET imaging equipment is utilized to collect detector data before imaging, and the corrective parameter value and the system matrix of various data in the imaging equipment are obtained simultaneously; (2) a mathematical statistic statistical model used for reconstructing an image is reconstructed according to a statistical feature that is met by the corrective data acquainted by the procedure (1) before imaging; (3) the general Gibbs experiment is led in aiming at the compute of a mathematical model in the procedure (2) , a maximal posteriori estimate method is adopted to perform the conversion of a reconstruction model, to obtain an optimized equation with a constrained objective function used for obtaining a PET reconstruction image; (4) a parabola is adopted to replace a coordinate descent algorithm, to perform the iterative computation treatment and to reconstruct the image based on the selection of a global parameter in the optimized equation through a result obtained by the procedure (3). The invention can greatly improve the quality of the PET reconstruction image.

The invention provides an online multi-target tracking method for coupling target detection and data association, which is used for real-time tracking of multiple interested targets in video, and belongs to the technical fields of computer vision and video monitoring. Starting from the intermediate results provided by the target detector, the present invention implements target tracking and optimizes target detection results by introducing sequential trajectory priors, which can effectively alleviate the problem of inaccurate detection results of the target detector; the present invention adopts the method of maximum a posteriori estimation The probabilistic framework calculates the association cost between the target and the detection result, improves the accuracy of data association, and can effectively deal with the mutual interference between targets; the present invention establishes a close connection between target detection and data association, making both Become two mutually promoting processes, which improves the performance of multi-object tracking.

The invention discloses a method for reducing the speckles of a synthetic aperture radar (SAR) image by combining dual-tree complex wavelet transform with a bivariate model, which mainly solves the problems that speckle noise cannot be well inhibited and part of edge information and detailed information are lost in the conventional method for reducing the speckles of the SAR image. The method comprises the following steps of: performing dual-tree complex wavelet decomposition on the original SAR image to obtain a real part and an imaginary part of a decomposition coefficient on each scale; solving the variance of a noise coefficient by using a non-logarithmic additive noise model; solving the edge variances of the real parts and the imaginary parts of the complex wavelet coefficient by using a local neighborhood window; solving a threshold contraction function by maximum posterior estimation and performing threshold contraction on the dual-tree complex wavelet decomposition coefficient; and performing dual-tree complex wavelet reconfiguration on the contracted coefficient to obtain an image of which the speckles are reduced. The method has the advantages of capability of effectively removing the speckle noise from the SAR image and high edge preserving performance, and can be used for reducing the speckles of the SAR images with abundant edge information and detailed information, particularly the airport, runway and road-containing SAR images.

The invention provides an adaptive variation remotely sensed image fusion method. In the method, multi-spectrum image and full-color image observation models are respectively established by analyzing the image degrading process, an inverse problem corresponding to the image fusion is described by utilizing a maximum posteriori estimation theoretical frame, and a variation image fusion model consisting of a multi-spectrum image data consistency constraint, a full-color image data consistency constraint and an image transcendental constraint is established; and in the resolving process, iterative solving is carried out by a gradient degressive algorithm, regularization parameters are adaptively solved by establishing a proper function, and an adjustable parameter is reserved simultaneously so as to meet the requirements of different users. The method can improve the original multi-spectrum image space resolution, effectively maintain the conventional spectrum information, and automatically select the proper regularization parameters adaptively according to different image data, so the method has the characteristics of high fidelity, high adaptation degree and the like.

Disclosed are a method and an apparatus for transmitting a sparse signal, and a method and an apparatus for recovering the sparse signal. The method for recovering a sparse signal by using a sparse signal recovering device that recovers a target signal from a received signal includes receiving a measurement signal with a noise signal from a sparse signal transmitting device which scans a target signal based on a measurement matrix, performing a mutual update procedure in which a likelihood probability is calculated by using a posterior probability of the target signal based on a relation between the target signal and the measurement signal, and the posterior probability is updated by using the likelihood probability, and recovering the target signal by performing maximum a posterior estimation for a final posterior probability output through the mutual update procedure.

The invention discloses a recommendation system and a recommendation method based on relationship type cooperative topic regression. The system at least comprises an RCTR (Relationship type Cooperative Topic Regression) model establishing module, a parameter studying module and a predicted value calculating module, wherein the RCTR model establishing module is used for integrating user-item rating information, item content information and a relation structure between items into a hierarchy bayesian model to establish an RCTR model; the parameter studying module is used for utilizing maximum posteriori estimation to study parameters in the RCTR model, and finally obtaining a parameter user implicit vector, an item implicit vector, an item relation vector and a full posteriori possibility of an item topic ratio; and the predicted value calculating module is used for utilizing the user implicit vector, the item topic ratio and point estimation of item implicit deviation to calculate a predicated value of evaluation by using a predicted value calculating formula. According to the recommendation system and the method disclosed by the invention, the user-item rating information, the item content information and the relation structure between the items is integrated to one hierarchy bayesian model seamlessly to integrate a social network between the items into a recommendation process, so that the recommendation accuracy is improved.

The present invention provides a low-dose X-ray CT image reconstruction method. The method comprises: obtaining imaging system parameters of a CT device and projection data in the low-dose CT scanning protocol; constructing a statistics generation model of the chordal graph data through the statistical law of the imaging process; constructing the priori statistic model of the chordal graph data; constructing a complete statistic model, and converting a model to a chordal graph data recovery model according to the maximum posterior estimation method; applying the chordal graph data recovery model, and obtaining estimated chordal graph data and other statistical variables; and performing CT image reconstruction according to the obtained chordal graph data to obtain and output a CT graph according to the obtained chordal graph data. The objective is to establish a high-quality chordal graph data recovery model and method based on the chordal graph data generation principle and the statistical prior therein so as to realize the CT image high-quality reconstruction capable of greatly reducing image noise / artifact and recovering image details.

The invention provides an online video foreground and background separation method based on regular error modeling. The method comprises the following steps: 1) obtaining in real time the video data of a monitoring system; 2) based on the real-time change of the video background environment, embedding into the model the transform operator optimization variables based on the video background; 3) based on the characteristics of the video foreground target featuring constant change, constructing a regular error modeling model; 4) combining the step 2 and step 3 for a complete statistical model; according to the maximum posteriori estimation method, obtaining a complete monitoring video foreground and background separation model; 5) performing down sampling to the video data; accelerating the calculation on the video foreground and background separation model of step 4 to realize the real-time solution to the model; and 6) according to the monitoring video data foreground and background separation result obtained from step 5, outputting the foreground and the background in real time. The invention provides a high-speed and high-precision online video foreground and background separation method, which is of great significance in detecting, tracking, identifying and analyzing the target in a monitoring video in practical use.

The invention discloses the hole repairing method and device of a three-dimensional crown mesh model and relates to the oral cavity technology field. In the prior art, the accuracy of the automatic repairing of a three-dimensional crown mesh model hole is low. In the invention, the above problem can be solved. The method comprises the following steps of acquiring the specific dimension of each three-dimensional tooth mesh model from multiple sets of three-dimensional tooth mesh models with a consistent topological structure of the same three-dimensional tooth mesh model; using a maximum posterior estimation method to construct a target three-dimensional tooth mesh model matching with the shape of the three-dimensional scanning crown model to be repaired based on the specific dimension; carrying out deformation from a target three-dimensional crown mesh model intercepted by the target three-dimensional tooth mesh model to a three-dimensional scanning crown model; repairing the three-dimensional scanning crown model according to the point correspondence relation of the deformed target three-dimensional crown mesh model and the three-dimensional scanning crown model; carrying out meshsmoothing on the repaired three-dimensional scanning crown model and acquiring a final result. The method and the device are mainly suitable for the scene of automatically repairing tooth holes.

The invention discloses a foundation and satellite-borne radar rainfall data fusion method based on wavelet domain regularization. The method comprises: based on radar rainfall data wavelet domain statistical characteristics, selecting an appropriate prior model of rainfall data; and determining a regularization function of scale coefficient fusion and wavelet coefficient fusion after wavelet decomposition of the rainfall data of the foundation and the satellite-borne radar, then solving the scale coefficient of the rainfall data and the maximum posteriori estimation of the wavelet coefficientin a wavelet domain by using a gradient projection method, and finally performing wavelet inverse transformation to obtain the optimal rainfall estimation. According to the method, uncertainty of rainfall estimation of different sensors and wavelet domain statistical characteristics of rainfall data are considered in the fusion process. The fusion result reduces the uncertainty of a single sensor, and can better maintain and reconstruct detail features such as a heavy rainfall extreme value and small-scale change, thereby being more beneficial to monitoring and forecasting of heavy disaster weather such as flood monitoring and the like.

The invention provides a pulse wave signal denoising method based on DTCWT-Spline. The method includes the steps that firstly, original noise-containing pulse wave signals are subjected to double-tree complex wavelet decomposition, wavelet coefficients on all layers are denoised through a Bayes maximum posteriori estimation threshold value, then double-tree complex wavelet inverse transformation is carried out, and pulse wave signals obtained after high-frequency noise is filtered out are obtained; the obtained pulse wave signals obtained after the high-frequency noise is filtered out are detected through a sliding window method, wave trough points in the pulse wave signals obtained after the high-frequency noise is removed are recognized, a wave trough curve is fit through a cubic spline interpolation method to serve as the estimated baseline drifting distance, finally subtracting the estimated baseline drifting distance from the pulse wave signals obtained after the high-frequency noise is removed is carried out, and the pulse wave signals are denoised. The high-frequency noise and baseline drifting can be effectively removed, overall characteristic information of the original pulse wave signals is well kept, the method is simple, low in calculated quantity and low in occupied memory, and a technical foundation is provided for research and development of miniaturized and mobile noninvasive continuous blood pressure detection equipment based on pulse waves.

The invention discloses a self-adaptive UKF (Unscented Kalman Filter) algorithm. In the algorithm, a Kalman filtering algorithm based on a maximum likelihood criterion and the self-adaptive UKF algorithm based on maximum posteriori estimation are combined. Real-time estimation is performed on covariance through the two algorithms, and averaging is performed to obtain an estimated value with a better prior covariance true value tracking effect, so that the filtering accuracy is increased, and the filtering stability is enhanced.

The invention relates to a robust online channel state information estimation method adaptive to dynamic changes of a noise environment. The robust online channel state information estimation method comprises the steps of: constructing an online machine learning model for dynamic noise estimation based on real-time varying characteristics of a communication noise environment; embedding a channel variation regular in the online machine learning model based on continuously varying characteristics of a channel, constructing a complete statistical model, and acquiring a complete online channel estimation machine learning model according to a maximum posteriori estimation method; saving environmental noise distribution parameters and channel state information of a previous period of time by utilizing base station storage equipment, and combining with the online channel estimation machine learning model to obtain high-precision channel state information estimation. The fast and high-precision online channel state information estimation method capable of being adaptive to the noise environment is implemented based on a machine learning principle, and is of great significance to the reduction of communication delay, reduction of pilot signal usage and increase of information transmission rate in actual application.

A method for recovering a sparse signal of a finite field may include: updating discrete probability information of a target signal element of the finite field and discrete probability information of a measurement signal element of the finite field by exchanging the discrete probability information of the target signal element with the discrete probability information of the measurement signal element a predetermined number of times, wherein the target signal element and the measurement signal element are related to each other; calculating a final posteriori probability based on a priori probability of the target signal element and the discrete probability information of the measurement signal element, acquired as the exchange result; and recovering the target signal by performing maximum posteriori estimation to maximize the final posteriori probability.

The invention provides a high-speed rail track response prediction method based on sparse Bayesian width learning. The method comprises the steps: carrying out the linear and nonlinear feature extraction of an input temperature field variable, carrying out the maximum posteriori estimation of a hidden layer neuron node output layer weight, predicting a structure response output result, and evaluating the structure state of the track preliminarily. According to the method, the sparse Bayesian width learning method is adopted to carry out correlation mining on the data of the high-speed rail monitoring system, and the over-fitting problem of regression prediction can be effectively avoided through sparse solution of the weight w reflecting the relation between data variables. The method has the advantages of high prediction precision, high calculation speed and loose equipment hardware requirements, so that the mining of the correlation between the temperature load and the structural strain implied in a large amount of monitoring data can be realized, and meanwhile, the evolution of a monitoring data model is found in time to serve as a basis for judging the abnormal service state of the track structure.

The invention provides an underwater vehicle positioning method based on baseline geometric structure constraint, comprising the steps of calculating to obtain a 3D measurement distance between everytwo beacons; calculating the distance between every two beacons, and constructing an equality constraint condition and an inequality constraint condition of a 2D calculated distance between each beacon and the current position of an underwater vehicle; and obtaining predicted positioning information, establishing a relative distance measurement equation between the water surface beacon and the underwater vehicle, and obtaining the position, the course and the pitch estimation value of the underwater vehicle by solving a constraint optimization problem. According to the method, an underwater inertia / dead reckoning positioning method and an underwater acoustic positioning method are combined, and long-time effective and accurate positioning of the underwater vehicle is ensured; the distanceinformation between the acoustic beacons is effectively utilized in the positioning process; and by means of the maximum posteriori estimation criterion and extended Kalman filtering, constraint conditions are effectively fused into the positioning solving process, and the positioning precision of the underwater vehicle is improved.

The invention discloses an abdominal component analyzer and an analysis method thereof; an abdominal component analysis model of a maximum posteriori estimation method is firstly used, then the model is trained by detecting sample human body abdominal bioelectrical impedance, and finally the human body abdominal bioelectrical impedance of a detected person is sent into the trained model so as to obtain the content of abdomen components of the detected person; meanwhile, in the process of measuring the bioelectrical impedance of the abdomen of the human body, the accuracy of impedance data measurement is improved in a multi-frequency detection and self-correction mode, and experimental results show that the error of a measured value of the abdomen of the human body can be controlled within 10%. In addition, compared with doctors and common large-scale human body composition analyzers and abdominal fat analyzers, the device has the advantages of higher pertinence, smaller size, noninvasive property, safety, simplicity, convenience and low cost, so that the product is a feasible solution for current household detection.

The invention discloses a tag recommending system and method based on synergistic topic regression with social regularization. The tag recommending system comprises a CTR (Common Technical Regulation) model establishing module, a CTR model establishing module with social regularization, a parameter studying module and a tag recommending module, wherein the CTR model establishing module is used for establishing CTR models to all tags; the CTR model establishing module with social regularization is used for integrating an article-tag matrix, content information of the articles and a social network of the articles into a level Bayesian model to establish a CTR-SR model; the parameter studying module is used for studying parameters in the model established by the CTR model establishing module with social regularization by utilizing maximum posterior estimation, and finally obtaining the whole posterior probability of all the parameters; the tag recommending module is used for carrying out tag recommendation according to the studied parameters. The tag recommending system and method disclosed by the invention has the advantages that the CTR model is applied in tag recommendation, and the level Bayesian model is provided by expanding the CTR, so that the article-tag matrix and the content information of the articles are effectively integrated, the network relationship among the articles are utilized, and further the accuracy of recommendation is improved.

The present invention provides a microorganism association network prediction method and apparatus. The method comprises obtaining abundances of various microorganisms in metagenome sequencing samples and an environmental factor measurement value corresponding to each metagenome sequencing sample; establishing a hierarchical Bayesian model according to a data generation process of each metagenome sequencing sample and the abundances of the various microorganisms and the corresponding environmental factor measurement values in each metagenome sequencing sample; learning the hierarchical Bayesian model by using a maximum posteriori estimation algorithm, and determining an objective function of the hierarchical Bayesian model; optimizing the objective function; and predicting association between the microorganisms and association between the microorganisms and the environmental factor measurement values by using the optimized hierarchical Bayesian model. Accuracy and practicability of prediction tasks of microorganism association and microorganism and environmental factor association are improved.

The invention provides an X-ray pulsar signal time delay estimation method based on maximum posteriori estimation. The method comprises the following implementation steps: (1) initializing parameters; (2) correcting the time sequence of the X-ray pulsar photons reaching the spacecraft; (3) obtaining a mean value and a variance of prediction phase delay corresponding to a spacecraft position prediction error delta r; (4) solving a maximum value of a maximum posteriori estimation MAP cost function; and (5) obtaining the time delay estimation value of the X-ray pulsar signal. The maximum value of the maximum posteriori estimation MAP cost function is solved through the corrected time sequence and the mean value and variance of the prediction phase delay corresponding to the spacecraft position prediction error, and the time delay estimation value of the X-ray pulsar signal is obtained through the maximum value of the maximum posteriori estimation MAP cost function and the rotation frequency of the X-ray pulsar, so that the influence on the estimation precision due to the fact that only the photon arrival time sequence is considered and the spacecraft position prediction error is not considered in the prior art is avoided, and the time delay estimation precision is improved under the same observation time.

The invention discloses an image motion deblurring method, which includes a first step of establishing infinite student-t distribution mixed models based on the Dirichlet process as an image gradient distribution model and a point spread function model, and obtaining the number of infinite student-t distribution models automatically based on an observation image; and a second step of taking the image gradient distribution model and the point spread function model respectively as an image a priori model and a point spread function priori model, conducting motion deblurring processing on the image using a maximum posteriori estimation method, and estimating a model parameter using variational Bayesian inference. The adaptive selection of the number of models is realized, the degree of fitting of the image gradient distribution is improved, and the technical effect of accurate image motion deblurring is achieved.

The invention provides an iterative massage passing (IMP) PN code capturing method based on a sum-product algorithm on the condition of a low signal to noise ratio. According to the method, an m sequence is modelized into a special linear block code; a sum-product decoding algorithm is executed on a Tanner graph to obtain maximum posteriori estimation of the m sequence; and an initial state is selected by introducing a posteriori log-likelihood ratio to generate a local m sequence; and at last, serial correlation verification is carried out with reception data and capturing is completed. The greatest advantage of the method provided in the invention is as follows: maximum posteriori estimation of an m sequence is directly decoded by utilization of a sum-product decoding algorithm and a posteriori log-likelihood ratio is introduced to select an initial state. Besides, the method has advantages of rapid capturing speed, low complexity and excellent performance under a low signal to noise ratio. According to the invention, a fundamental principle on IMP PN code capturing is mainly introduced and then an IMP PN code capturing method based on a sum-product algorithm is provided; and an analysis on a selection process of an initial state of an m sequence is emphasized and a detailed flow chart is provided; at last, a capturing performance and a complexity of the method are analyzed though examples.

The invention provides a Bayesian filtering-based multi-vehicle cooperative positioning method. According to the method, a probability density function for positions a target vehicle and an adjacent vehicle and a probability density function for relative distance between vehicles are used; in each iteration of extended Kalman filtering, a posterior probability estimate of a target vehicle positionis obtained via Bayesian filtering, a final position estimate of the target vehicle is determined according to a maximum posteriori estimate, cooperative positioning is realized via dual filtering operation, vehicle position errors can be reduced, and accuracy of vehicle position information can be improved.

The invention relates to a hybrid recommendation algorithm based on a deep neural network and probability matrix decomposition. The method comprises steps of S10, collecting and preprocessing information; S20, establishing a deep neural network model and a probability matrix decomposition model; S30, obtaining a potential feature vector fusing real information of the user and the project accordingto the three steps; and S40, performing personalized recommendation on the user by utilizing the feature vector. The invention discloses a hybrid recommendation algorithm based on a deep neural network and probability matrix decomposition. The beneficial effects of the invention are as follows: the method is based on previous research; the user scoring matrix data is fully utilized; and meanwhile, feature extraction is performed on the user description information and the project description information by using a deep neural network to generate a user and project real feature set includinguser preference features, and performing iterative optimization processing on the probability matrix model and the features generated by the deep neural network by using maximum posteriori estimation.

The invention provides an iterative massage passing (IMP) PN code capturing method based on a sum-product algorithm on the condition of a low signal to noise ratio. According to the method, an m sequence is modelized into a special linear block code; a sum-product decoding algorithm is executed on a Tanner graph to obtain maximum posteriori estimation of the m sequence; and an initial state is selected by introducing a posteriori log-likelihood ratio to generate a local m sequence; and at last, serial correlation verification is carried out with reception data and capturing is completed. The greatest advantage of the method provided in the invention is as follows: maximum posteriori estimation of an m sequence is directly decoded by utilization of a sum-product decoding algorithm and a posteriori log-likelihood ratio is introduced to select an initial state. Besides, the method has advantages of rapid capturing speed, low complexity and excellent performance under a low signal to noise ratio. According to the invention, a fundamental principle on IMP PN code capturing is mainly introduced and then an IMP PN code capturing method based on a sum-product algorithm is provided; and an analysis on a selection process of an initial state of an m sequence is emphasized and a detailed flow chart is provided; at last, a capturing performance and a complexity of the method are analyzed though examples.

The invention relates to a statistical inverse based electric / supersonic double-mode content boundary reconstruction method. According to the method, a radial shape model is used to implement parameterized representation of content boundary to be reconstructed, namely the target boundary, an electrical imaging model is used to construct a likelihood model for shape factor estimation, a supersonicreflection imaging model is used to construct a prior model of shape factor estimation, a maximal posteriori estimate method is used to solve an optimal shape factor, and the content boundary is reconstructed.

The invention relates to the technical field of bridge deck elevation measurement in bridge engineering, and discloses a bridge deck elevation fitting method based on a Bayesian-Kriging model. The method comprises the steps that S1 a Bayesian-Kriging fitting model is established; S2 test optimization of an elevation test point sample is carried out; S3 a Bayesian-Kriging prediction model is established; and S4 elevation fitting evaluation is carried out on the whole bridge deck. The model has the advantages that Bayes and Kriging are combined: based on non-information super-priority, multi-layer priori constraints are imposed on the basis function coefficients and related parameters of a Kriging model; an EM algorithm is used for the solving of the basis function coefficients and the maximum posteriori estimation of the related parameters; the Kriging model is improved, and the Bayesian-Kriging model is established; the adaptability and robustness of the model are enhanced; and the test optimization design of an elevation measurement sample based on a number theory method is carried out.

Methods for ascertaining a control strategy for a technical system using a Bayesian optimization method. The control strategy is created based on model parameters of a control model and is executable. The method includes providing a quality function whose shape corresponds to a regression function and that evaluates a quality of a controlling of the technical system based on model parameters; carrying out a Bayesian optimization method based on the quality function in order to iteratively ascertain a model parameter set having model parameters within a model parameter domain that indicates the permissible value ranges for the model parameters; and determining the model parameter domain for at least one of the model parameters as a function of an associated maximum a posteriori estimated value of the quality function.