52 results about "Association mapping" patented technology

The invention discloses a dynamic gesture recognition method and system based on a deep neural network. The dynamic gesture recognition method comprises the steps of collecting dynamic gesture video clips with different gesture meanings to generate a training sample data set, wherein the sample data includes RGB images and depth information; designing a dynamic gesture recognition network model based on the deep neural network, and training the model by using the training samples; and performing dynamic gesture testing and recognition by using the trained dynamic gesture recognition model. Thedynamic gesture recognition network model is composed of a feature extraction network, a front and back frame association network and a classification recognition network, wherein the front and backframe association network is used for performing front and back time frame association mapping on feature vectors obtained through the feature extraction network of the samples of each gesture meaningand merging the feature vectors into a fusion feature vector of the gesture meaning. According to the invention, a bidirectional LSTM model is introduced into the network model to understand the correlation between continuous gesture postures, thereby greatly improving the recognition rate of dynamic gestures.

Provided herein are methods for evaluating associations between candidate genes and a trait of interest in a population. The methods include a combination of genome-wide association analysis and one or more of nested association mapping (NAM), expression QTL analysis (eQTL), and allele epistastic analysis (AEA). Markers are selected or prioritized if they are shown to be positively-correlated with a trait of interest using GWA and a combination of one or both of NAM and eQTL. Also provided are models for evaluating the association between a candidate marker and a trait in a nested population of organisms. These methods include single marker regression and multiple marker regression models. Markers identified using the methods of the invention can be used in marker assisted breeding and selection, as genetic markers for constructing linkage maps, for gene discovery, for identifying genes contributing to a trait of interest, and for generating transgenic organisms having a desired trait.

The invention relates to a wireless network problem root cause positioning method and a device. The wireless network problem root cause positioning method comprises the following steps of acquiring awireless network problem; comparing the acquired wireless network problem and association mapping between a preset wireless network problem and a root cause positioning process to determine the root cause positioning process aiming at the wireless network problem, wherein each wireless network problem corresponds to one root cause positioning process, and each root cause positioning flow corresponds to at least one root cause determination rule; and according to the at least one root cause determination rule corresponding to the determined root cause positioning process, determining whether awireless network parameter accords with a predetermined condition and determining the final root cause of the wireless network problem.

The invention discloses a secondary virtual circuit failure locating method based on differential approximate reasoning of a failure area. The secondary virtual circuit failure locating method comprises the following steps: at first, carrying out virtual link and physical link modeling on a process layer network connection of a "net sampling and net hopping " or "point sampling and net hopping" intelligent substation, and carrying out association mapping configuration on a virtual link and a physical link; using a directed graph theory to carry out model abstraction on the process layer network connection of the intelligent substation, and abstracting to obtain a directed connected graph; and then, according to communication link abnormality information sent by a receiving side, when receiving failure of the virtual link, starting a failure diagnosis locating system, judging the on-off state of the virtual link to continuously iterate a difference set of the physical link for loading the connected virtual link and the disconnected virtual link, and locating a physical link failure point position causing the failure of the virtual link. The secondary virtual circuit failure of the intelligent substation is monitored, judged, diagnosed and located to assist an operation and maintenance staff to quickly find the failure, remove the failure and recover the normality, so as to guarantee the safe and reliable operation of a power grid.

The invention relates to a method and a device for associating attachments. The method comprises establishing association mapping tables between documents and/or attribute information of users corresponding to the attachments and temporary attachments of the attachments; obtaining the current documents and/or attribute information of users when the documents are needed to be submitted; and associating the documents with corresponding temporary attachments according to the current documents and/or attribute information of users and the association mapping tables. According to the method and the device for associating attachments, the association mapping tables between the temporary attachments of the uploading attachments and documents and/or attribute information of users are established during the attachment uploading, and the initiative adsorption type association can be performed on the documents and the temporary attachments according to the association mapping tables if documents are abnormal; accordingly, the repeated uploading of attachments and the occupation of network resources are prevented after the user attachments are abnormal, and the waste of server resources is reduced; simultaneously, users can perform selection according to the matched absorption type association attachments, hence, the man-machine interaction experience is enhanced.

The design and methods for a distributed database, distributed processing implementation network management system are disclosed for packet-based networks, with emphasis on Next Generation Networks (NGN) involving numerous network devices. The invention presented in this document runs under the paradigm of a regionalized or zonal network management architecture, where the entire managed network is divided into zones. Every zone, comprising of a group (or groups) of network devices, has a designated autonomous network management substation which can periodically collect data from the network devices in its respective zone, report alerts of detected network device failures, and temporary store the current and most recent data until a defined time. A central database also exists as storage for the old and historical data. The network administrator retrieves current data and historical data from the network substations and the central database, respectively, through a central network management station which is the overseer of all the zones, automatically keeping track of every network device and zone association mapping.

The invention is carefully designed to minimize computational load, database load, and network traffic in all components of the management system while not compromising the security and speed of data retrieval. Scalability is kept in mind so that the system can support complex, large, and growing networks.

The invention discloses a virtual and real loops association mapping and diagnosis method of a virtual and real diagnosis system of an intelligent substation, and the method comprises an association mapping method of a process layer virtual and real loops and a diagnosis method of the process layer virtual and real loops; the association mapping method of the process layer virtual and real loops comprises the following steps: 1, building a real loop model; 2, building an association relation between the real loop and a message block, namely defining a sending port and a receiving port, and a sent or received message block APPID; 3, building the association relation between the port and a virtual loop broken link warning signal, port power and port state warning signal; 4, storing the configured association relation in a secondary operation and maintenance model file; 5, parsing a SCD model and the secondary operation and maintenance model, and storing the model association relation ina library. The association relation of the real loop and the virtual link is built, and technical problems, such as failure in building of process layer virtual and real loops mapping and implementation of diagnosis of process layer virtual and real loops, of the virtual and real loops of the intelligent substation are solved.

The invention relates to a micro-satellite DNA molecular marker for cervine animals. The molecular markers can be used for identifying allelic polymorphism, identify same or related cervine animals, distinguish the cervine animals and research the genetic diversity of the species group. The molecular marker can also be used for the genetic and phenotype research utilizing a statistics method such as linkage analysis, association mapping, linkage imbalance. The information can be used for breeding and/or selecting plants.

The invention discloses a method for carrying out transparent transmission on communication of a virtual private network (VPN). The method comprises the following steps that: (1), IP addresses of internal computers in a local area network are changed as a public network IP address of the internet outlet; (2), a router in the local area network receives VPN data packets sent by the internal computers with MAC addresses changed but without changing the IP addresses and forwards the IP packets to the internet; (3), an association mapping table expressing association between the internal computers and IP addresses of VPN servers that the internal computers access currently is established on the router of the local area network; (4), the router of the local area network receives IP packets sent back by the VPN servers, wherein the IP addresses are not changed and only the MAC addresses are changed, and searches the association mapping table and forwards the IP packets to the internal computers; and (5), communication isolation is carried out on the internal computers of the local area network. According to the provided method, only one public network IP address is used, so that a problem VPN communication of IPSEC between a local area network and the internet can be solved.

On the one hand, the invention provides a panorama video image fusion method. The method comprises the following steps of collecting, by multiple preset cameras, video images of an environment scene at different spatial degrees, and building a spatial spherical model of the environment scene; according to the spatial spherical model, stitching the video images of the environment scene corresponding to multiple cameras, so as to obtain a spherical pattern image corresponding to the spatial spherical model of the environment scene; according to a gesture matrix of multiple cameras, building an association mapping between image pixels in the spherical pattern image and image pixels in the video images of the environment scene corresponding to the cameras; and according to the association mapping, performing image fusion on image pixels in accordance with image fusion conditions in the spherical pattern image. One the other hand, the invention provides a panorama video image fusion device. The panorama video image fusion method and device effectively overcome the defect that the existing display device cannot play a high resolution panorama video fluently, and makes the panorama video play technology more applicable in different scenes.

The invention provides a method for realizing dynamic interface allocation by describing a system object, building an interface template for interface display, and describing two-way association mapping between the object and the interface template. The method has the advantages that 1. on-line dynamic interface allocation is realized, and on-line increment, modification and release are supported; 2. the interface template and service can be combined to construct different business support schemes; 3. a customer can experience an excellent interface, and real-time data submission and real-time interface data refreshing are supported; 4. the universality is high, and the method can be used in different systems.

The invention provides an XML file and PDF file synchronous association method and apparatus. The method comprises the steps of setting corresponding tag identifiers for text segment tags in an XML file; in a typesetting process of converting the XML file into a PDF file, obtaining the tag identifiers; obtaining typesetting position information of the tags in the PDF file; and establishing a synchronous association mapping relationship between the tag identifiers and the typesetting position information. Through the method, a user can know the specific positions, in the PDF file, of the tags in the XML file; and synchronous association of the XML file and the PDF file is realized, so that the use convenience is enhanced and the working efficiency is improved.

The invention discloses a structured query language instruction generation method and device based on an index system, wherein, the structured query language instruction generation method based on theindex system comprises the following steps: constructing the index system including a plurality of index models and a data subject market according to a data table; configuring data subjects in the data subject market to obtain configuration data corresponding to the data subjects; generating a structured query language instruction corresponding to the data subject according to the configurationdata; and obtaining the access service by using the structured query language instruction. According to the technical scheme provided by the invention, the structured query language instructions suitable for various business requirements can be generated quickly, automatically and flexibly, the association mapping relationship between the index system and the SQL service is established, various business requirements are effectively met, the uniqueness of data indicators is ensured, and the problem of index inconsistence is effectively avoided.

The invention discloses a loader working condition identification model construction and identification method. Firstly, a corresponding sensor is arranged on the loader to collect multi-source signals such as torque, pressure, gear and brake, and the like, and the data is standardized, the zero floating signal is stripped, the missing value is interpolated and compensated, and the collected signal is processed by noise reduction filter. Secondly, principal component analysis is used to select the feature attributes from the loader's multi-attribute data, and statistical analysis is used to extract the feature of the principal component. Then, the loader working condition samples are established, and the association mapping between the load signal and the pre-classified working condition patterns is established by using the supervised learning data mining algorithm, and the working condition identification model is formed by training a large number of data samples. Combining the feature extraction method of principal component analysis with KNN algorithm, the distance formula of KNN algorithm is improved to make it more consistent with the working condition identification, and theaccuracy and efficiency of the working condition identification algorithm are improved.

The invention relates to the technical field of a molecular marker, in particular to a calcutta hemp InDel molecular marker as well as a development method and application thereof. The molecular markers can be used for identifying the allele polymorphism, identifying identical or relevant calcutta hemp plants, distinguishing the calcutta hemp plants and studying the genetic diversity in populations. The molecular marker can also be used for using inheritance and phenotype study of a statistical method, such as linkage analysis, QTL positioning, genealogy confirmation, population genetic polymorphism study, association mapping and linkage disequilibrium; the information can be used for calcutta hemp plant propagation and/or selection.

The invention provides a history video file format storage method used for a monitoring industry and a system. The method includes: creating a video data file; creating an index file; and using creation time as an association word to establish an association mapping relationship of the index file and the video data file, and saving the associated video data file and index file at the same time into storage equipment. The system includes a video data creation module, an index file creation module, an association module and a storage module in communication connection. According to the method provided by the invention, more videos can be stored on the premise of the same disk space, the same video quality and the same playing duration; at the same time, file format processing speed before writing is greatly improved, and transmission time in a file storage process is shortened; and in addition, a video file of a specific time point in history video files can be positioned more quickly, video positioning is more accurate, and positioning time is shortened.

The invention provides a parallelization method for a DENCLUE algorithm. The method can be used for the environment where large-scale data clustering analysis is performed. Compared with a traditionalcentralized execution method, the improved parallelization clustering method for the DENCLUE algorithm realizes a higher data processing speed. The method mainly includes the following steps that 1,original data is divided; 2, grid association mappings in data space are established in parallel; 3, a global grid association mapping is generated; 4, mining and clustering are performed in parallel.Through the parallelization method for the DENCLUE algorithm, the data dividing speed and final clustering speed are effectively increased, and the speed of executing the algorithm is increased.

The invention discloses a data mining based text data network construction system. The system comprises a data collection module, a data mining module, a social network construction module, a visual display module and a database, wherein the data collection module receives newly added text data, and performs word segmentation on the text data sequentially; the data mining module extracts an eigenvalue of the newly added text data subjected to the word segmentation in the database, compares the newly added text data with stored text data subjected to eigenvalue extraction in the database, and performs association mapping on data with an eigenvalue closest to the eigenvalue of the newly added data; the social network construction module constructs a file by taking a sender of the text data received by the data collection module as a file name and taking a newly added text data name, a name of the data subjected to the association mapping and a sender of data corresponding to the data subjected to the association mapping as file attributes, and forms a social network ontology library; and a visual display module analyzes data corresponding to the file, and performs display in a manner of line segments or nodes.

The invention discloses a method for determining a function test range. A to-be-tested service element is obtained according to service test requirements and a service association mapping diagram is imported. Whether a service element and/or a service mapping relationship in the service association mapping diagram need/needs to be tested or not is judged according to the to-be-tested service element. If yes, the service element and/or the service mapping relationship in the service association mapping diagram are/is updated. Service modules of different levels associated with the to-be-testedservice element are searched in the updated service association mapping diagram. Meanwhile, a test topology path of the to-be-tested service element in the updated service association mapping diagramis outputted according to the service mapping relationship between the service modules of different levels. If not, the search is directly conducted in the service association mapping diagram. Meanwhile, the test topology path of the to-be-tested service element in the service association mapping diagram is outputted according to the service mapping relationship between the service modules of different levels. Therefore, the completeness of the function test is guaranteed to the maximum extent.

The invention discloses a gas extraction object and extraction measure optimal matching method. The method comprises the following steps of classifying gas extraction objects and building an extraction parameter library and an alternative extraction method library which correspond to the gas extraction objects and influence the extraction effect; establishing an association mapping model of an extraction parameter library and an extraction method library; selecting a key parameter with the maximum influence degree on each extraction object; dividing a plurality of intervals for matching with field application; by means of a fractional order diffusion model, a coal body double-pore medium permeability dynamic evolution model, a working face gushing gas quantitative traceability technology and the like, giving an extraction parameter relative importance weight sequence; calculating a weight vector of the extraction parameter and a relative importance weight of each gas extraction methodby utilizing an analytic hierarchy process; based on the relative importance weight of each extraction method and the actual situation of the mine, selecting the extraction method, matching the propergas extraction method according to the gas extraction object, the matching result with high accuracy and reliability is given, and the gas extraction amount is increased.

Simple sequence repeat (SSR) markers identified in Jatropha curcas and useful for the molecular genotyping of plants. are described. These markers may be used for identifying allele polymorphisms, identifying identical or related plants, differentiating plants and studying genetic diversity in a population. The markers may also be used in genetic and phenotype studies using statistical methods, for example, linkage analysis, association mapping, linkage disequilibrium and the like. The information may be used for breeding and/or selection of plants.

The invention relates to the technical field of computers, in particular to an interface calling method, and an interface calling device, an electronic equipment and storage medium. The method comprises the following steps: reading a current interface request is read, and judging whether a preorder interface request the same as the current interface request exists in an interface request queue ornot; if it is judged that the preorder interface request the same as the current interface request exists in the interface request queue, the current interface request serves as a repeated interface request to be stored in a repeated request queue; and establishing association mapping between the preorder interface request and a repeated interface request so as to process the preorder interface request and the corresponding repeated interface request according to the association mapping relationship. According to the method provided by the invention, repeated sending of multiple same interfacerequests can be avoided, and the terminal pressure is reduced. And the interface requests in the repeated request list can be processed together according to the association relationship. Thus, global management of the interface request is realized.

The present invention discloses a Fast-Flux Botnet detection method and system in a high-speed network, which belongs to the field of network security detection technology. Firstly, a DNS data packetis obtained, and the captured DNS data packet is parsed and filtered, to obtain a suspicious Fast-Flux Botnet. A global association mapping relationship between a domain name and an IP address of thesuspicious Fast-Flux Botnet is analyzed, and global features are extracted by using the global association mapping relationship. Time-based local features of the suspicious Fast- Flux Botnet are extracted. Finally, the global features and the local features in offline data are trained by using a machine learning algorithm to obtain a trained machine learning model, and global features and local features in unknown data are detected by using the trained learning model to obtain a classification result of the suspicious Fast-Flux Botnet. The present invention can improve detection accuracy, reduce a false alarm rate, and improve processing efficiency.

The invention provides a data storage method for law and regulation information. The method comprises the steps of collecting law and regulation data related to the law and regulation information; splitting the law and regulation data according to a preset splitting rule to obtain fine data; Matching the fine data according to a preset matching rule to form an association mapping table; and storing the association mapping table and the fine data in a database. Compared with the prior art, the association relationship among laws and regulations, cases and attachments can be quickly establishedand stored, and a multi-dimensional and multi-scheme query function can be provided for a user.