30 results about "Tensor completion" patented technology

The invention discloses an image restoration method based on low-rank tensor completion and discrete total variation. The method comprises: introducing a discrete total variation regular term factor and integrating the discrete total variation regular term factor into a unified objective function; providing a method based on low-rank tensor completion and discrete total variation for natural imagerestoration, identifying a to-be-repaired pixel of the input damaged image; establishing the relation between known elements and unknown elements; performing diffusion in different directions on theboundary of the to-be-restored area, diffusing information of the undamaged area into the to-be-restored area to restore the image, and finally obtaining the restored image, so that the image restoration method is smoother in edge processing and more accurate in overall image restoration.

The invention belongs to the technical field of image or video data calculation methods, and particularly relates to a visual data completion method based on low-rank tensor estimation of a limited nuclear norm. The method comprises the following steps of A1, initializing a pre-prepared original tensor to be complemented to obtain a target tensor x, wherein the original tensor to be complemented comprises a color image and a video sequence; A2, constructing a limited kernel norm model for tensor completion by using the target tensor x to obtain a tensor completion target function; A3, performing optimization modeling on the objective function to obtain an alternative function capable of solving the optimal solution of the objective function; and A4, solving the substitution function to obtain a target tensor, converting the target tensor into a data source corresponding format, and obtaining a final completion result of the color image and the video sequence. According to the method, an improved restriction nuclear norm model is extended to tensor data completion to approach the rank of original tensor data, so that tensor data completion is completed.

The invention discloses a seismic signal completion method based on a synchronous sparse low-rank tensor completion model. The method comprises the following steps that: a four-dimensional earthquakeis preprocessed; a tensor completion model is established; the objective function of the tensor completion model is constructed through a truncated tensor nuclear norm regularization method; and the constructed objective function of the tensor completion model is solved through an imprecise augmentation Lagrange multiplier method. According to the method, a tensor model based on a transform domainis established; a tensor to be recovered is converted into a Curvelet transform domain, so that the sparsity and low rank of the tensor to be recovered is improved; and the accurate and efficient recovery of four-dimensional missing seismic data is realized through an ADMM algorithm.

The invention discloses a visual data tensor completion method based on smooth constraint and matrix factorization, which comprises the following steps of: firstly, acquiring missing overall data, determining a known data position set omega in the missing overall data, and constructing a corresponding visual data tensor model; then, taking the low-rank tensor completion model as a basic framework, introducing total variation and a tight wavelet framework to carry out smooth constraint, reducing the complexity by utilizing a matrix decomposition technology, and constructing a visual data tensor completion model based on smooth constraint and matrix decomposition; and finally, based on an alternating direction multiplier method, introducing a plurality of auxiliary variables to obtain an augmented Lagrangian function form of the visual data tensor completion model, converting an original optimization problem into a plurality of sub-problems, respectively solving the sub-problems, and outputting a convergence result after multiple iterations, namely a complete visual tensor of completed unknown data. According to the method, more efficient and accurate visual data recovery can be achieved under the condition that large-scale random missing exists in the acquired data.

The invention discloses a three-dimensional tensor completion method based on a Capped nuclear norm. The method comprises the following steps of S1, setting input original incomplete data as a form Xof a three-dimensional tensor; S2, defining a capped nuclear norm | | of the tensor X on a three-dimensional level; wherein the formula (1) is shown in the specification, S3, minimizing the capped nuclear norm until X is converged; outputting reduced tensor . According to the method, the data lost in the tensor is complemented through the rank of the three-dimensional tensor, specifically, the capped kernel norm of the tensor on the three-dimensional level is defined, and the capped kernel norm is iteratively solved so as to be minimized, so that the low-rank tensor is obtained, and the three-dimensional tensor is complemented quickly and accurately.

The invention discloses a tensor completion-based multi-energy CT (computed tomography) imaging method, a tensor completion-based multi-energy CT imaging device, storage equipment and a tensor completion-based multi-energy CT imaging system. The tensor completion-based multi-energy CT imaging method comprises the following steps of: firstly, respectively processing an obtained projection value ofeach section of narrow beam energy spectrum by using an FDK (frequency division duplex) algorithm to obtain a reconstructed image of each energy section; then, modeling the obtained reconstructed image of each energy segment into a third-order tensor, establishing a tensor nuclear norm and total variation regularization minimization model, and improving the precision of the reconstructed image ofeach energy segment; and finally, performing optimization weighting on each slice in the tensor obtained by modeling according to a weighting fusion algorithm to obtain a final imaging image. The beneficial effects of the invention are that the method achieves the data collection through the multifunctional CT simulation system based on GATE, combines the inherent multi-dimensional properties of aCT problem with the tensor, and achieves the more precise reconstruction of a CT scanning image.

The invention relates to a 6G mobile communication system based on tensor calculation and a data processing method thereof. The 6G mobile communication system comprises a high-dimensional informationsource, a tensor compression unit, sending equipment, a channel, receiving equipment, a tensor completion unit and an information sink which are connected in sequence. The tensor compression unit is used for compressing the high-dimensional information source to be transmitted to reduce the information redundancy of the high-dimensional information source. The tensor completion unit is used for supplementing information loss items caused by interference of the complete high-dimensional signals in the transmission process. According to the method, the problem of pain points in a future 6G mobile communication system is solved, compression and completion of high-dimensional data can be achieved, and a good effect is achieved. The method improves the capability of processing high-dimensionaldata in 6G, and lays a foundation for the implementation of 6G in the future.

The invention discloses a wireless sensor network high-dimensional data completion method based on Hankel transformation. The method comprises the steps of S1, modeling received wireless sensor network high-dimensional data into a tensor form chi; S2, counting effective element positions of the tensor data chi to form a tensor data support set omega; S3, performing TUCKER decomposition on the tensor data chi to obtain a kernel tensor of the tensor data chi, wherein chi = C * 1U (1) *... * iU (i)... * NU (N); S4, establishing a simple kernel tensor nuclear norm tensor completion model; and S5,based on the simple kernel tensor nuclear norm tensor completion model, carrying out Hankel transformation and variable separation, carrying out calculation filling on missing data through optimization of an alternate iteration method, and outputting a completion result. According to the method, data complementation can be rapidly and accurately carried out.

The invention provides a third-order low-rank tensor completion method based on a GPU. The method comprises the following operation steps that (1) a CPU transmits input data DATA1 to a GPU, and the number of cycles is initialized; (2) the GPU obtains a third-order tensor of a current cycle based on a least square solution; (3) the GPU obtains a third-order tensor of a current cycle based on a least square solution; (4) the CPU checks whether an end condition is met, and if yes, the step (5) is executed, otherwise, the number of cycles is increased by one, and the step (2) is excuted to continue to circulate; and (5) the GPU transmits the output data DATA2 to the CPU. According to the method, the GPU is used for accelerating the calculation task of three-order low-rank tensor completion medium-high concurrency processing, so that the calculation efficiency is improved. Compared with traditional CPU-based third-order low-rank tensor completion, the method has the advantages that the calculation efficiency is obviously improved, and the same calculation can be completed within a short time.

The invention discloses a missing vibration signal recovery method based on variational Bayesian parallel factorization, which comprises the following steps of: constructing a three-dimensional tensor by a collected time domain vibration signal through segmenting a sampling point, introducing a likelihood model by combining the decomposed three-dimensional tensor with a Bayesian method, effectively utilizing the prior information of a factor matrix, introducing posteriori distribution with effective precision, processing a model by adopting a Bayesian method, deducing posteriori distribution of parameters of all unknown numbers including a factor matrix and a hyper-parameter, deducing posteriori distribution of the factor matrix and the hyper-parameter by adopting a variational Bayesian algorithm, and further deducing distribution prediction of missing signals. The performance of the method is evaluated by using a root-mean-square error, compared with a traditional low-rank tensor completion algorithm, the variational Bayesian parallel factorization algorithm has the advantages that the error is smaller, missing signals can be more effectively recovered, and the problem of signal missing caused by sensor failure in vibration signal analysis is effectively solved.

The invention discloses a greenhouse model construction and recovery method based on low rank tensor completion (LRTC). The method mainly aims at a large-scale multi-attribute wireless sensor network (WSNs) with slow environmental parameter change, and since greenhouse data tensors monitored by the network have inherent correlation: low rank and smoothness in time dimension, a tensor nuclear norm regularization (TNN) minimization model based on time smoothness is constructed, and the model is optimized and solved by adopting an alternating direction multiplier method (ADMM). By well balancing the low-rank performance and the time smoothness of the data tensors, the data recovery method provided by the invention is greatly improved in the aspects of stability and recovery precision.

The invention belongs to the technical field of air pollution monitoring, and particularly relates to an air pollutant concentration precise monitoring method and system based on tensor representation, and the method comprises the steps: obtaining the air quality data of a to-be-detected point in real time, inputting the obtained air quality data into a monitoring model, and obtaining a monitoring result; marking the monitoring point according to the monitoring result; According to the method, the multi-source data and the space-time air pollutant concentration data are fused to carry out tensor completion, so that high-precision tensor data are obtained.

The invention discloses a non-local low-rank conversion domain and full-connection tensor decomposition image reconstruction method and device. The method comprises the following steps: S1, inputting a to-be-restored image; s2, constructing a tensor decomposition model: S2.1, segmenting the input image to obtain non-local tensor blocks; s2.2, introducing the non-local tensor block into a B spline conversion domain to obtain a conversion domain form of the non-local tensor block; s2.3, through the non-local tensor blocks, constructing a non-local similar tensor block group; s2.4, combining full-connection tensor decomposition, and constructing a full-connection tensor decomposition factor; s2.5, constructing a low-rank tensor completion model, and performing optimization according to the steps S2.1 to 2.4 to obtain a non-local low-rank conversion domain and full-connection tensor decomposition-based model; and S3, constructing an image restoration model, obtaining a to-be-restored image, and obtaining a restored image through a restored image tensor block group obtained through the tensor decomposition model. In the spectral image restoration process, the whole image reconstruction is more accurate.

The invention discloses a shielded face detection method based on tensor completion.The data features of a face image sequence are extracted by employing a 3D CNN, so that the time sequence features in the face image sequence aresupplemented while the features of a single-frame face image are obtained, the richness of the features of the face image is improved, and the method plays an important role in improving the accuracy of a face detection result. Tensor completion is utilized to supplement the preliminarily extracted face image sequence data features, so that the adverse effects of face shielding and the like on the face detection result can be reduced, and the shielded face detection result is improved.

The invention discloses an acoustic vector sensor DOA estimation method based on matrix decomposition under data loss, and the method specifically comprises the following steps: firstly, arranging any acoustic vector sensor array in a measurement scene, obtaining a received signal, and constructing a received signal tensor model; then, by utilizing the low-rank characteristic of the tensor model, performing missing data completion on the constructed received signal tensor model by using a tensor completion algorithm based on matrix decomposition; and finally, expanding the complemented tensor data according to a folding direction to obtain a complete received signal matrix after data complementation, and performing information source direction-of-arrival estimation on the complemented complete received signal matrix by using an ESPRIT algorithm for any acoustic vector sensor array. According to the method, accurate DOA estimation can be realized when part of array elements in any acoustic vector sensor array are damaged.

The embodiment of the invention provides a tensor completion method, a tensor completion device, tensor completion equipment and a storage medium, and belongs to the technical field of image processing. The method comprises the following steps: acquiring original tensor data and tensor data to be observed, and establishing a first objective function according to the original tensor data and the tensor data to be observed; and performing wavelet tight frame transformation on the original tensor data, updating the first objective function, and combining the updated first objective function with the non-convex penalty function to obtain a second objective function. And solving the second objective function by adopting an alternating direction multiplier method to obtain a complete tensor image corresponding to the tensor data to be observed. According to the method, a tensor completion method of non-convex low-rank tensor approximation based on a wavelet tight frame is adopted, and a low-rank tensor approximation model which is better in recovery effect and can retain more structural information can be provided, so that a complete tensor image is obtained.

A low-rank tensor completion method based on three-dimensional total variation and Tucker decomposition comprises the following steps: reading a damaged video into MATLAB software, converting the damaged video into a three-dimensional tensor with the tensor size of X * Y * Z, optimizing a solved objective functional by using an augmented Lagrange formula, decomposing a mixed objective functional into a plurality of optimization sub-problems, introducing three auxiliary variables, dividing the three auxiliary variables into three independent parts, introducing a three-dimensional weighted difference operator into three-dimensional total variation constraint, retaining a multi-factor structure of the three-dimensional tensor, and describing a segmented smooth structure of a three-dimensional space domain of the tensor data; continuously iteratively updating the introduced three auxiliary variables and the tensor y needing to be repaired, and determined that the tensor completion is completed when the maximum number of iterations is reached or the relative error of the tensor y complemented for two consecutive times is smaller than a given parameter value epsilon. According to the method, multi-channel data can be effectively processed, the low-rank performance of the tensor is described, the proposed convex functional is efficiently solved, and the restoration of the high-loss-rate damaged video is completed.

The invention discloses a seismic signal completion method based on a synchronous sparse low-rank tensor completion model. The method includes preprocessing four-dimensional earthquakes, establishing a tensor completion model, and adopting a truncated tensor kernel norm regularization method The objective function of the tensor completion model is constructed, and the objective function of the constructed tensor completion model is solved by using the non-exact augmented Lagrangian multiplier method. The invention transforms the tensor to be restored into the Curvelet transform domain by establishing a tensor model based on the transform domain to improve its sparsity and low rank, and combines the ADMM algorithm to realize accurate and efficient recovery of four-dimensional missing seismic data.

The invention provides a Bayesian tensor complementation algorithm based on complex noise, which aims at target data with missing values and complex noise, expresses the target data as a tensor which is the sum of a tensor estimated value and noise, and extracts low-rank information of the tensor by adopting CP decomposition, so that the target data with missing values and complex noise can be complemented. Gibbs sampling is carried out by combining CP decomposition and a Bayesian method framework, a tensor estimation value is obtained through iteration, and target data are complemented and denoised simultaneously based on the tensor estimation value. The low-rank information of the tensor is fully mined by adopting CP decomposition, the observed tensor information is fully utilized, and iterative sampling is carried out, so that the completion algorithm can realize good completion and denoising on abnormal values and complex noise, is a robust and effective tensor completion algorithm, and compared with a completion method in the prior art, the tensor completion algorithm has the advantages that the complexity is low, and the efficiency is high. According to the complementation algorithm provided by the invention, a more accurate tensor estimation value can be obtained, so that more accurate target data complementation and denoising are realized.

The invention belongs to the field of network public opinion analysis, and particularly relates to a rumor refuting game propagation control method based on sparse representation and tensor complementation, which comprises the following steps: acquiring user data information, and preprocessing the user data information; extracting related attributes of the preprocessed user data information; inputting related attributes of the user data information into a rumor-refuting game propagation model based on sparse representation and tensor completion, and predicting a rumor propagation trend of theuser; and according to the predicted rumor propagation trend of the user, controlling the rumor propagation user to prevent rumor propagation. According to the method, the evolutionary game theory andthe neural network are used for predicting whether the user participates in the rumor-refuting topic or not, the time when the user participates in topic discussion can be dynamically predicted, andsituation awareness is carried out on the rumor topic development trend.

The invention provides a Bayesian tensor complementation algorithm based on multiple measurement values, which comprises the following steps of: representing multiple measurement value data by using multiple tensors, and setting each measurement value of each tensor element of the tensors to obey Gaussian distribution; performing CP decomposition on the tensor to obtain a corresponding factor matrix, and setting parameters of the factor matrix to obey conjugate prior distribution; and then a Gibbs sampling method is adopted to sample the posterior condition distribution of each parameter, an estimated value of the tensor is output, interpolation is performed on missing values in the multi-measurement-value data based on the estimated value of the tensor, and thus data completion is realized. In conclusion, according to the complementation method, aiming at measurement data which is low in measurement precision and high in cost and is repeatedly measured for multiple times in some areas, the Gibbs sampling method is combined with CP decomposition to realize data complementation, and compared with a complementation method in the prior art, due to the fact that information of all the measurement data can be utilized by the method, the data complementation efficiency is improved. Therefore, a more accurate estimation value can be provided, and more accurate data completion is realized.

The invention discloses a non-local low-rank transformation domain and a fully connected tensor decomposition image reconstruction method and device, including: S1, inputting an image to be repaired; S2, constructing a tensor decomposition model, including: S2.1, segmenting the input image, Obtain the non-local tensor block; S2.2, introduce the non-local tensor block into the B-spline transformation domain, and obtain the transformation domain form of the non-local tensor block; S2.3, construct the non-local similarity through the non-local tensor block Tensor block group; S2.4, combine fully-connected tensor decomposition to construct fully-connected tensor decomposition factors; S2.5, construct a low-rank tensor completion model, and optimize it according to S2.1‑2.4 to obtain a Local low-rank transformation domain and fully connected tensor decomposition model; S3, construct an image inpainting model, obtain the image to be inpainted, and obtain the inpainted image through the inpainted image tensor block group obtained by the tensor decomposition model. This makes the image reconstruction more accurate in spectral image restoration.

The invention discloses a large-scale target three-dimensional reconstruction method and system based on binocular vision. The three-dimensional reconstruction method comprises the following steps: obtaining binocular images of a target at different visual angles; performing dual-threshold judgment on pixel information in the binocular image, positioning an information missing position, and performing information enhancement on missing pixels by using a tensor completion algorithm; extracting a target in the binocular image after information enhancement; and carrying out stereo matching on the binocular image of the target, calculating local point cloud information according to the obtained depth map, and restoring the overall three-dimensional structure of the target through point cloud splicing. The binocular vision-based three-dimensional reconstruction technology can effectively overcome the influence of environment change on the reconstruction effect, the tensor completion technology is utilized to enhance the missing pixel information of the target caused by overexposure and shadow, and then the target is segmented and extracted, so that the interference of the background on the three-dimensional reconstruction is solved. The point cloud reconstruction effect is improved.