79 results about "Global structure" patented technology

A method for identifying voice emotion based on support vector computer includes characteristic picking up-analyzing to collate characteristic parameter selection of global structure and sex, as well as to collate characteristic parameter selection of time sequence structure and to collate sex and vowel number; support vector computer training to carry out identification on five emotions of happy, angry, sad, fear and surprise.

The invention discloses a fine-grained emotion polarity prediction method based on a hybrid attention network, and aims to overcome the problems of lack of flexibility, insufficient precision, difficulty in obtaining global structure information, low training speed, single attention information and the like in the prior art. The method comprises the following steps: 1, determining a text context sequence and a specific aspect target word sequence according to a comment text sentence; 2, mapping the sequence into two multi-dimensional continuous word vector matrixes through log word embedding;3, performing multiple different linear transformations on the two matrixes to obtain corresponding transformation matrixes; 4, calculating a text context self-attention matrix and a specific aspect target word vector attention matrix by using the transformation matrix, and splicing the two matrixes to obtain a double-attention matrix; 5, splicing the double attention matrixes subjected to different times of linear change, and then performing linear change again to obtain a final attention representation matrix; and 6, through an average pooling operation, inputting the emotion polarity into asoftmax classifier through full connection layer thickness to obtain an emotion polarity prediction result.

A coating system (1) for blown containers (9) made of plastic material, with high production rate and flexibility so as to allow an efficient coupling with the most advanced one-stage or blowing machines. Such coating system (1), despite its high production rate, envisages a compact global structure with low implementation costs and contained energy consumption. Along with the system, a corresponding coating process is described, which consists in the effective and rapid application of several paint layers on plastic containers (9).

A scale-Space feature extraction technique is based on recursive decomposition of polyhedral surfaces into surface patches. The experimental results show that this technique can be used to perform matching based on local model structure. Scale-space techniques can be parameterized to generate decompositions that correspond to manufacturing, assembly or surface features relevant to mechanical design. One application of these techniques is to support matching and content-based retrieval of solid models. Scale-space technique can extract features that are invariant with respect to the global structure of the model as well as small perturbations that 3D laser scanning may introduce. A new distance function defined on triangles instead of points is introduced. This technique offers a new way to control the feature decomposition process, which results in extraction of features that are more meaningful from an engineering viewpoint. The technique is computationally practical for use in indexing large models.

A combined truss bridge of double bridge decks relates to a novel bridge structure and the combined truss bridge of double bridge decks consists of an upper concrete bridge deck, a steel structure web member and a lower concrete bridge deck. The upper and lower bridge decks are used for bearing the load of vehicles and are used as an upper chord part and a lower chord part to share the global stress with the global structure. The steel structure web member consists of a slant web member and a vertical web member, and combines the upper and the lower concrete bridge decks as a whole to be used for resisting the shearing force of the truss bridge. The upper concrete bridge deck and the lower concrete bridge deck in tensile areas are provided with vertically stressed steel bars or section steel. The invention, compared with traditional single bridge deck combined girder bridges, enjoys greater traffic capacity and higher traffic comfort; compared with steel truss bridges, the invention remarkably reduces steel consumption by using concrete bridge decks as compression chords, and has the advantages of low noise, easy maintenance, good durability and so on.

The invention puts forward a compressed sensing reconstruction method based on image nonlocal similarity. According to the method, image nonlocal similarity, a low-rank matrix and minimum total variation (TV) are combined, and two priors, namely, the local similarity and local smoothness of images, are fully utilized. On one hand, the block effect and the loss of global structure information caused by independent processing of single image blocks in the traditional method are eliminated, and on the other hand, real details of images are retained and false details produced by unreliable information are reduced or removed while noise is suppressed. High-quality compressed sensing reconstruction of images is realized. Compared with a general reconstruction method based on transform domain sparse or TV constraints, the method of the invention is of robustness to noise, better reconstruction quality is achieved, and great improvement is achieved both in visual effect and evaluation index.

Provided is an image feature extraction method based on global and local structure fusion, characterized by comprising: 1) constructing a weight adjacent map; 2) determining laplacian matrix of similar matrix, degree matrix and images, 3) determining scatter matrix inside the kind and between the kind; 4) determining projection matrix, 5) identifying. The invention provides a feature extraction method of fusing the global structure information and the local structure information, wherein complex features fused of the global feature and the local feature are extracted, thereby the method has strong resolving power. The method not only has the characteristics of holding the reflection method locally, namely holding the characteristics of manifold structure of data; moreover has the characteristics of linear discrimination analysis method, namely assembling the date of the kind more compact to enlarge the distance between the kinds. The invention is applied in image recognition, thereby increasing identifying performance.

The invention discloses a PoLSAR image classification method based on FCN and sparse-low rank subspace expression integration; the method fully utilizes the powerful ability of the FCN which can automatically learn non-linear deep multiple dimensional space characteristics from PoLsar data, can utilize the advantages that a linear dimension reduction algorithm based on sparse-low rank map embedment can simultaneously capture local and global structure information of the PoLsar data, and can effectively integrate the FCN and sparse-low rank subspace expression, thus solving the PoLSAR image non-linear deep space feature extraction problem, the dimension reduction problem, the valid integration problem between polarized information and space information, and effectively solving the PoLSAR image classification problem. The method can obtain integrated multi-layer subspace characteristic information containing various types; the information contains linear and non-linear types, shallow anddeep types, local and global types, and polarized and multiple dimensioned space types, thus providing strong discrimination ability, and greatly improving PoLSAR image classification accuracy.

A system and method for creating a customized portal environment website that retrieves content from external websites and presents this content with a consistent and controlled look and feel. Three features of this invention that enable the consistent look and feel are clipping, scrubbing and link behavior tools. The clipping tool identifies the structure of clipped pages. The structure of the clipped content is stored in a database in the system. As the system dynamically retrieves clipped pages from the target site, the system compares the structure of the retrieved clip to the anticipated, stored structure of the clip. If the two structures match, the clipped page is rendered in the look and feel of the custom website. If the structures do not match, the dynamically clipped page is displayed in a new separate window. The scrubbing feature of the present invention allows the designer of the website to define a global structure for all pages that are retrieved from a particular domain. If a requested page does not conform to the defined structure, the system opens a new window for the page. The link behavior tool of the present invention traps a request for the embedded link upon it's execution and redirecting the request to a page that has already been approved and clipped or by opening a new window for the embedded link.

The invention discloses an affine transformation-based frontal face image super-resolution reconstruction method. In the method, a recently popular two-step method is adopted to perform super-resolution reconstruction on images, wherein a first step is used for reconstructing intermediate-low frequency information (a global structure) for testing low-resolution images, namely, firstly blocking all images, establishing mapping relationship between a high-resolution pixel space and a low-resolution pixel space at each block position respectively and performing the super-resolution reconstruction on the tested low-resolution images through an obtained mapping matrix to obtain a first-step high-resolution reconstructed image block of each block position; and a second step is used for reconstructing high-frequency information (detailed information), namely, performing error compensation on the reconstructed image blocks obtained in the first step according to LLE to obtain a residual image block at each block position, adding the high-resolution reconstructed image block at each block position and the corresponding residual image block to obtain a super-resolution reconstructed image block and finally synthesizing all the super-resolution reconstructed image blocks into a complete image which is a super-resolution reconstruction result corresponding to the tested images.

The invention discloses an image super resolution (SR) reconstruction method based on subspace projection and neighborhood embedding. The method is characterized by: using first and secondary subspace projection methods to project original high-dimensional data to a low-dimensional space, using dimension reduction feature vectors to show a feature of a low-resolution image block so that global structure information and local structure information of original data can be maintained; comparing a Euclidean distance between the dimension reduction feature vectors in the low-dimensional space, finding a neighborhood block which is most matched with the low-resolution image block to be reconstructed, using a similarity and a scale factor between the feature vectors to construct an accurate embedded weight coefficient so that a searching speed and matching precision can be increased; then constructing the similarity and the scale factor between the feature vectors, calculating the accurate weight coefficient and acquiring more high frequency information from a training database; finally, according to the weight coefficient and the neighborhood block, estimating the high-resolution image block with high precision, reconstructing the image which has the high similarity with a real object, which is good for later-stage real object identification processing.

The invention relates to a method for repairing a three-dimensional grid model based on a global structure, which comprises the following four stages of: stage 1, detecting cavities of a three-dimensional grid model; stage 2, decomposing the three-dimensional grid model into a base model and high-frequency information by using the improved bilateral filtering algorithm; stage 3, based on the base model obtained after the decomposition in stage 2, repairing the base model by using a smooth three-dimensional model repairing method; and stage 4, based on the high-frequency information obtained after the decomposition in stage 2, repairing geometric structure details in the cavities, wherein stage 4 comprises the following steps of: (1) adding a structure curve by user's interactive operation on a graphical interface; (2) detecting interest regions based on the structure curve in (1); (3) specifying a control region by user's interactive operation on the graphical interface, and parameterizing the region into a two-dimensional plane; (4) generating a geometric detail image in the two-dimensional plane obtained after the parametrization in (3) based on the high-frequency information obtained after the decomposition in stage 2; (5) repairing the geometric detail image obtained in (4); and (6) mapping the repaired geometric detail image in (5) back to the three-dimensional grid model to ultimately finish repairing the three-dimensional model. The invention can repair the three-dimensional grid model having detail information of an obvious global structure, and the repaired three-dimensional grid model has richer geometric structure details.

The large span buildings comprising no ordinary beams and columns are formed of vertical load-bearing composite wall-panels and composite floors, both comprising two concrete layers interconnected by steel strip webs. The stiff horizontal plane formed of assembled roof/ceiling units, supported by wall-panels, connected to both gables restrains transversal movement of longitudinally arranged wall-panels attached tops, bracing them simultaneously against sideway and lessening their buckling lengths. Floors, if any applied, being rigidly connected to the vertical panels additionally improve stability of the global structure. Hereby invented composite wall-panel and floor are adapted to the same purpose. The global structure, being braced in that way, behaves as a rigid box made of slender panels.

A structure-guided transformation transforms a region of an image into a region in the structure-transformed image according to the desired structure. The invention achieves efficient and accurate structure-guided processing such as filtering, detection and comparison in the transformed domain and thereby facilitates use of simple operations to enhance or detect straight lines or edges. Structure information is used to enhance and detect image features of interest even when the shape of the image structure is not regular. Both global and local structures of objects can be inspected. Global structure inspection detects gross errors in image structure; therefore side effects caused by mismatched structure-guided processing are avoided. Subtle defects along the edge of a structure can be detected by local structure inspection. Structure information guidance provides an edge detection inspection system that tolerates significant noise and contrast variations.

The invention discloses a method for recognizing Chinese characters in a natural scene. The method recognizes characters by a modeling training stage and a character recognition stage. The modeling training stage comprises successively establishing a tree structure expression of template characters, synthesizing a training set, training a convolutional nerve network, extracting the deep template characteristic of a node, and establishing and training a Markov random field. The character recognition stage comprises preprocessing a picture to be recognized, extracting the deep characteristic of an input picture, minimizing a Markov random field energy function, and finally recognizing characters. The method takes account of the local characteristic and the global structure of the characters while recognizing the characters, overcomes an influence on a recognition effect due to fuzzy characters and large deformation in the natural scene by combining the Markov random field technique with the deep characteristic of the node in the tree structure, thereby increasing recognition efficiency.

Systems and methods consistent with some embodiments of the present invention provide data structures and methods for the automatic storing, manipulating and processing of a intermediate printable data generated from a first printable data are presented. In some embodiments, the first printable data may take the form of a PDL description of a document and the intermediate printable data may take the form of a display list of objects generated from the PDL description.

In some embodiments, a data structure for storing a intermediate printable data generated from a first printable data may comprise at least one memory pool, which may further comprise a plurality of uniformly sized segments to store the intermediate printable data; at least one global structure for storing information related to the one or more memory pools; and buffers for performing operations on the first printable data and the intermediate printable data.

The invention discloses a semi-supervised network representation learning algorithm LSDNE (Labeled Structural Deep Network Embedding) based on a deep compression self-encoder. The method comprises thefollowing steps: building a model, pre-training the input data with a deep belief network (DBN) to obtain the initial values of the model parameters, and taking the adjacency matrix and Laplace matrix of the network as inputs; encoding the network by a self-encoder with deep compression, and obtaining the global structure of the node; using Laplacian feature mapping, and obtaining the local structural features of nodes; using an SVM classifier to classify the known label nodes and optimize the whole model; using the Adam optimization model and obtaining a representation of the node. The invention can simultaneously use the structure information of the network and the label information of the node to carry out network representation learning, and the deep learning model is used, so that the performance of the representation of the node on the label classification task is better than the existing algorithm. Deep compression self-encoder can reduce the over-fitting phenomenon and make the model have better generalization performance.

The invention discloses a head-neck blood vessel center line extraction method and device. The method comprises the steps of obtaining a blood vessel segmentation result and a blood vessel thermogramresult according to a head blood vessel CTA image; inputting a blood vessel skeleton line extracted based on the blood vessel segmentation result into a preset network model for blood vessel classification and naming, and obtaining blood vessel starting and ending point data; and extracting a head and neck blood vessel center line according to the blood vessel starting and ending point data and the blood vessel thermogram result. The technical problem of low blood vessel center line extraction accuracy in related technologies is solved. Through the method and the device, the purpose of extracting the blood vessel center line by combining the local detail information and the global structure information of the head and neck blood vessels is achieved, so that the technical effect of improving the accuracy of extracting the head blood vessel center line is achieved.

The invention provides an image recognition method and device based on non-negative low-rank representation and semi-supervised learning. The method includes the following steps that: an image data set is obtained, wherein the data set contains marked data and unmarked data; an objective function is obtained according to a Gaussian field, a harmonic function and a low-rank representation function,non-negative constraint is performed on the coefficient of the low-rank representation function, the objective function is converted into a Lagrangian function, and variables, Lagrangian multipliersand a penalty factor in the Lagrangian function are updated; and iterative updating is carried out continuously until the method terminates, and the label matrix of the image data set is outputted, and test data are classified and identified according to the label matrix. According to the image recognition method and device of the invention, the semi-supervised learning and the low-rank representation are combined, and therefore, global structure information and local structure information can be well utilized, and the corruption of samples can be effectively eliminated or mitigated. The method and device have high robustness to noises and can obtain high classification performance regardless of whether training samples or test samples are damaged.

The invention discloses an image denoising method and apparatus based on global and local prior cooperative constraints. The image denoising method based on global and local prior cooperative constraints is provided, through comprehensive utilization of self-similarity prior of the global structure and the nonlinear mapping based on a local kernel function, being able to effectively recover the fine structure of the image while removing noise, and further improving the signal-to-noise ratio and the subjective objective quality of the image.

The invention discloses an image classification method based on low-rank sparse representation, and the method comprises the following steps: carrying out the normalization of each column of a training sample A and a test sample Y into a unit L2 norm; calculating a projection matrix P; calculating a projection matrix A' of the training sample A; calculating a projection matrix Y' of the test sample Y; and completing a classification task through a nearest neighbor classifier. The beneficial effects of the invention are that the method enables the sparse representation, low-rank representationand discrimination projection to be integrated in one frame, and gives consideration to the local and global structure information of the observation data; compared with other dimension-reduction methods, the method is highly robust to outlier points and noise.

The invention discloses a sparse learning RCS sequence feature extraction method, belonging to the neighborhood of radar target recognition technology. At first, sparse learning is carried out on thetrain sample, Using the sparse coefficient as the adjusting factor of the distance between the features of samples, A sparse learning projection matrix is established, which automatically selects a small number of neighborhood samples through sparse learning, and can better preserve the local structure information of the sample neighborhood, thereby improving the performance of target recognitionand overcoming the shortcoming of the traditional projection method that only global structure features can be extracted.

The invention discloses multisystem image file producing method, an image file and a multisystem start guiding method. The producing method comprises the following steps: S11, producing a bootloader, and writing the bootloader into an object file; S12, establishing a corresponding global structure for one operating system corresponding to the unestablished global structures in a plurality of operating systems of a multisystem; S13, writing the global structure established in the S12, and a root file system and core of the operating system corresponding to the global structure into the object file; and S14, judging whether the global structures are established for all global structures correspondingly or not, if yes, taking the object file as an image file and finishing the flow, otherwise, executing the S12 again. According to the invention, intelligent start guiding of the systems can be realized, and further the stability and maintainability of the system can be obviously improved.

The invention relates to an electroencephalogram (EEG) signal online identification method with data structure information being fused. The method comprises the following steps: S1) establishing a classification model based on an online sequential extreme learning machine (OS-ELM) algorithm by utilizing a small training set formed by a small number of labeled EEG samples to serve as an initial classification model in semi-supervised learning; S2) establishing a structure learning model by utilizing an on-line fuzzy clustering method, and estimating a global structure of data distribution afterbatch increase of EEG samples collected online based on prior information of the labeled EEG samples; S3)carrying out labeling on the EEG samples collected online by utilizing the classification model, and through a batch learning mode and based on the structure information estimated by a structural learning model, selecting a batch of EEG samples collected online and meeting a certain conditionsto add to a training set, and re-training the classification model by utilizing the updated training set; and S4) carrying out online identification on the collected EEG signals through the updated classification model.

The invention discloses a finite element analysis method of a curved stiffened plate in a high-temperature environment. The method comprises that following steps of: obtaining a plate stiffness matrixthrough considering the influence of temperature change on materials and combining a DST-BK plate element theory; obtaining a rib element stiffness matrix and a rib element mass matrix after rib plate displacement coordination at interfacial surfaces of rib plates according to rib plate displacement coupling conditions; applying a uniform temperature field which is equivalent to thermal load to the plates and carrying out static analysis to obtain a plate additional thermal stress stiffness matrix considering thermal stress and derive a rib additional thermal stress stiffness matrix; carryingout eigenvalue analysis on a global structure stiffness matrix and a global additional thermal stiffness matrix to obtain a buckling critical temperature; superimposing the global structure stiffnessmatrix and the additional thermal stress stiffness matrix to obtain a curved stiffened plate stiffness matrix considering the influence of the temperature on material properties and the influence ofadditional thermal stress; and carrying out eigenvalue analysis on a global mass matrix to obtain the dynamic characteristics of curved stiffened plates in a uniform-temperature environment.