75 results about "Feature class" patented technology

A system and method for processing images utilizing varied feature class weights is provided. A computer system associates two or more images with a set of feature class data, such as color and texture data. The computer system assigns a set of processing weights for each of the feature classes. The two or more images are blended according to the feature class weights. The blended image can be further adjusted according to the content of the images.

In some embodiments, scripts may be used to perform parcel data acquisition, conversion, and clean-up/repair in an automated manner and/or through graphical user interfaces. The scripts may be used, for example, to repair geometries of new parcel data, convert multi-part parcel geometries to single part parcel geometries (explode), eliminate duplicate parcel geometries, append columns, create feature classes, and append feature classes. These scripts may be executed in a predetermined manner to increase efficiency. In some embodiments, different combinations of attributes may be appended to stored parcel data. In some embodiments, a tracking application may be used to track information about sources of data. In some embodiments, a tracking application may be used to track which system users are assigned to specific tasks (e.g., in a data acquisition project).

The present invention discloses a method for building, defining, and storing features in an application neutral format comprising building a feature based on a feature class, wherein the feature class comprises feature geometry, feature constraints, and feature dimensions, defining the built feature as a geometric representation of an individual feature type, and storing the representation in a binary file format.

The invention relates to an indoor map and a making method thereof. The making method comprises the steps of acquiring and obtaining cartographic data of a cartographic area, classifying and layering the cartographic data, further dividing each feature class into a plurality of feature layers, determining an adopted indoor map type, namely an indoor layout map, an indoor route map or an indoor escape map, determining map data of the cartographic area according to the feature layer corresponding to the selected indoor map type, expressing priority according to the feature layers in the map data, determining specific cartographic contents of the maps in different scales, and generating and outputting the indoor map according to the determined specific map contents. According to the indoor map and the making method thereof, the indoor data is classified and layered, so that the method is applicable to various indoor environments; carrying capacities of the maps in different scales are considered; expression priorities are set for indoor data layers; the corresponding map contents are displayed according to different map scales; the requirement of a user on map cognition is met; and the map can better serve for indoor and outdoor navigation.

Systems (100) and methods (300) for efficient video analysis. The methods involve: automatically identifying features of at least one feature class which are contained in a first video stream; simultaneously generating a plurality of first video chips (904) using first video data defining the first video stream; displaying an array comprising the first video chips within a graphical user interface window; and concurrently playing the first video chips. Each of the first video chips comprises a segment of the first video stream which includes at least one identified feature.

The invention discloses a DIS-NV feature-based emotion recognition method. The method comprises the following steps of S1, selecting an emotion word having the non-fluency feature as a DIS feature word, selecting a sound production type having the non-language feature as an NV feature class, and constructing a DIS-NV feature set; S2, obtaining training voice texts, respectively matching the training voice texts with the DIS-NV feature set, extracting corresponding DIS-NV feature values, training by adopting a BLSTM model and obtaining a BLSTM classification model; S3, obtaining a to-be-recognized voice text, matching the to-be-recognized voice text with the DIS-NV feature set, extracting a corresponding DIS-NV feature value, recognizing by using the BLSTM classification model and outputting an emotion recognition result. According to the invention, the emotion recognition is realized through fully utilizing voice texts having the non-fluency feature and the non-language feature. The method is simple in implementation, high in recognition efficiency and high in precision.

The invention relates to a high-spatial resolution remote sensing image-based road region building change extraction method and device. According to the method and device, an object-oriented image processing strategy is adopted, the spectral information and spatial information (including structural indexes and spatial relationships) of images are comprehensively utilized, and a single-class classification method is adopted. In order to avoid interference (spectral similarity) on extracted results caused by ground feature classes except road regions, it is required that existing road information such as an existing road vector diagram, is provided in advance; the existing road vector diagram is adopted to extract a road region range; and newly increased buildings are extracted within the road region range. With the high-spatial resolution remote sensing image-based road area building change extraction method and device provided by the invention, time for a traditional method to obtain newly increased buildings such as illegal buildings by using image visual interpretation can be greatly decreased, efficiency can be improved, and human resources can be saved. The method and device can be used for road maintenance and monitoring business operation systems.

A system and method for changing the weighting accelerometer filtering based upon motion sensor measurements. Using measurements from motion sensors, one or more features can be extracted. This information is used by an algorithm to assign a most appropriate test feature class to be the class of the state of motion at a particular moment in time. This information is used to provide more accurate positioning information, and can also be used to power down the motion sensor at various times, providing an improvement in battery efficiency.

The invention discloses a surface feature symbolization method for realizing map library integration. The method comprises the following steps: S1) configuring a symbol library template; S2) defining surface feature symbols, and describing shapes of the surface feature symbols by description statements; S3) uploading and analyzing the symbol library template on a drawing platform; S4) symbolizing the drawn surface features in real time; S5) editing continuation attribute values of the surface feature symbols by using an attribute control panel; S6) fine adjusting the shapes of the symbols by setting characteristic points; S7) when data is input to a database, creating feature classes and inputting information according to the corresponding relation of the surface feature symbols and the feature classes in a GIS (Geographic Information System) table. The invention adopts the symbolization protocol rules in the symbolization technology: all the surface feature symbols are formed by nesting basic entities, and in the instantiation process of the surface features, only one feature is needed for precise expression. According to the method, the storage space needed by the symbols is greatly reduced, the generation of extra auxiliary symbols and entities is avoided, and the map library integration is truly realized.

The invention provides a multi-data center based natural resource geospatial data organization method and system. The method comprises a data type identification step, a feature class model establishment step, a relation class data model establishment step and an auxiliary information acquisition step. Firstly, the type of natural resource geospatial data is identified and data centers, a database, a data set and an auxiliary information set are established; secondly, according to the identified data type, establishing a feature class data integration organization model and establishing a relation class data model; and finally, according to the identified data type and a relation class data calculation result, acquiring auxiliary information and storing the auxiliary information in the auxiliary information set. According to the organization method provided by the invention, unification of different data types in a conception layer based on a plurality of data centers and integration organization management of different data types are realized; orderly data management and data tracking and backtracking are realized; and the application convenience and flexibility of the natural resource geospatial data are satisfied.

The invention discloses a face recognition algorithm training method based on multi-objective learning, comprising the following steps: randomly initializing neural network parameters, using a loss function based on face identity and a loss function based on face feature point position constraint to minimize the learning objective to train the deep convolution network; when the prediction accuracyrate of the face identity reaches the threshold, calculating the loss function based on the distance within the face feature classes and the loss function based on the distance between the face feature classes, and performing calculation of the loss function based on the face identity and the loss function of the face feature point position constraint on each sample; based on the artificial weight setting, weighting each loss function to obtain the total loss function, and realizing the back propagation based on the total loss function. Therefore, the update of the network parameters is realized, and the network training is stopped when the accuracy is stable to get the trained face recognition model.

A networking system may comprise a web site incorporating a combination of one or more of the following: a crediting system, a chat bidding system, a performance bidding system, a telephony switching system, a media interaction system, a display system, a photo management system, and a messaging system. The networking system may categorize each user of the web site into classes, including a featured class, a common class, a sponsored class, and a sponsoring class. The networking system may facilitate user interactions, some of which may be partially dependent on user classes. A featured user may opt to temporarily appear as a common user, thus activating an alternative user profile and hiding his featured status. A sponsored user may participate in interactions on the web site if a sponsoring user offers his credits to pay for the sponsored member's participation. Other aspects of the networking system are also disclosed herein.

The invention relates to an incremental naive Bayes text classification method based on lifelong learning. The method comprises: obtaining an initial text; extracting feature words of the text, storing a feature word list of a text set and text number; calculating word frequency of each feature word in the text, dividing the text set into two parts of a training set and a validation set; through anaive Bayes classifier, training a training set vector model generated in a previous step, to obtain prior probability and feature class condition probability of the naive Bayes model and storing; ifa new text exists, training the new text in an incremental manner, and updating the prior probability and feature class condition probability of the naive Bayes model, if no new text exists, selecting a test corpus from the validation set, and according to the naive Bayes model, obtaining a prediction text category of the text corpus, and calculating accuracy rate of prediction. The method can guide learning of new tasks in an incremental manner by using knowledge learned in previous tasks, and has new feature processing ability and field adaptive ability.

The invention provides a pedestrian image feature classification method and system, and the method comprises the steps: carrying out the data expansion of a pedestrian image sample in a sample dataset; carrying out the grouping of the pedestrian image sample in the sample dataset after expansion, and obtaining a plurality of pedestrian sample groups; selecting samples, building a multi-channel convolution neural network, and extracting the global and local features of the sample data through the multi-channel convolution neural network; setting a loss function, calculating a loss value of the multi-channel convolution neural network, and optimizing the multi-channel convolution neural network; carrying out the feature classification of each global-local feature through the optimized multi-channel convolution neural network, and obtaining the feature class of each pedestrian sample group. The method enables the sample data to be expanded, meets the condition that triple loss exerts strict requirements for an input sample, can guarantee the robustness through employing multi-loss to optimize the multi-channel convolution neural network, and is suitable for the processing of pedestrian image features of a plurality of scenes.

The invention discloses a method for monitoring ageing conditions of asphalt pavements based on remote sensing images. Multispectral satellite remote sensing images are subjected to inversion based on a multiple endmember spectral mixture analysis model and are extracted to obtain an asphalt abundance distribution diagram of the asphalt pavements with the different ageing conditions, so as to obtain monitoring information of the ageing of the asphalt pavements. The method comprises the following steps: selecting the multispectral satellite remote sensing images containing asphalt ageing characteristic spectrum wavebands and classifying endmembers in the remote sensing images; establishing an initial endmember spectrum database, optimizing an endmember spectrum of each surface feature class, and running a multiple endmember linear spectral mixture analysis model to obtain a pixel abundance value of each type of endmember. The method disclosed by the invention can be used for rapidly and conveniently extracting asphalt pavement ageing information and estimating healthy conditions of the pavements, can be used for effectively improving the precision and applicability of the monitoring of the ageing conditions of the asphalt pavements, and can be applied to maintenance survey and planning of the pavements of expressways or other grades of roads.

Systems (100) and methods (300) for efficient spatial feature data analysis. The methods involve simultaneously generating two or more chip images (904) using image data defining a first image (508); concurrently displaying an array (906) comprising all or a portion of the chip images in a plug-in window (702); and displaying in the plug-in window information relating to an attribute (a1, a2) of a feature (A₅) contained in a selected one of the displayed chip images (1002). The chip images are generated in response to a user selection of a feature (A₁) contained in at least a portion of the first image displayed in an application window (504). Each of the chip images comprises a panned view, a zoomed view, or a panned-and-zoomed view of the first image including one or more features of a user-selected feature class.

The invention discloses a collaborative filtering method based on user project classes and scoring reliability, and the method comprises the steps: building a user scoring matrix and a user project class feature matrix through user scoring data and project information, and adding a common scoring penalty factor according to the user scoring matrix to calculate the similarity of user scores; according to the user project feature class matrix, calculating the similarity of user project preference is calculated by accessing the number of times of project feature classes and enabling the project feature class score to be larger than the average value number, on the basis, calculating the total similarity of users and then predicting on the basis of neighbor scoring reliability, unscored projects to generate recommended projects by establishing a nearest neighbor set. According to the method, the real interest preference of the user can be obtained more accurately, personalized recommendation is conducted on the designated user, and therefore the quality of a recommendation system is improved.

The invention relates to a transformer fault state evaluation method comprising the steps of: firstly, constructing a data set of a transformer insulating oil characteristic class, an insulating papercharacteristic class, a gas characteristic class, a temperature characteristic class, a capacitance characteristic class, and a partial discharge feature class, and calculating a corresponding probability value; then, calculate the average value of oil, paper, gas, temperature, capacitance, and partial discharge characteristic values reflecting the fault state of a transformer; further, constructing a transformer fault state evaluation function; and finally, performing transformer fault state evaluation. The transformer fault state evaluation method can evaluate the fault state of the transformer, reflect the uncertainty of the characteristic value of the transformer fault state, provides theoretical guidance for the transformer fault state evaluation, and provide a necessary technical support for the operation and maintenance of a distribution network.

The invention relates to an image processing method and device and a storage medium, and the method comprises the steps: obtaining the first image information of a target object; according to the character features of the first image information, determining character feature classes to which the first image information belongs, wherein each character feature class corresponds to a target object;and determining behavior characteristics of the target object according to the archive information corresponding to the character characteristic class. According to the embodiment of the invention, the behavior of the target object can be analyzed by utilizing the collected image information, so that manpower resources are saved.

The invention discloses a method for creating ETL scripts from a relational database to Hive. The method comprises the following steps of: obtaining database metadata information; rapidly recognizinga feature class of each data table in the relational database; automatically creating an ETL script for each data table according to the class; and registering the generated ETL scripts in a scheduling platform through a simulated access technology or a service calling manner. According to the method, two dimensionality features, related to ETL, of the data tables are self-defined. Features of thedata tables can be rapidly classified, and when ETL operation from the relational database to the Hive is carried out in batches, a defined script template is utilized to efficiently and automatically create a corresponding ETL script for each data table and the ETL scripts are submitted to the scheduling platform, so that the data migration efficiency is greatly improved.

The invention discloses an aircraft telemetry data feature extraction and hierarchical classification method and device, and the method comprises the steps: carrying out the time sequence analysis ofregularized preprocessing data, and constructing a plurality of basic feature classes; constructing a high-order statistic of the sample, and extracting a feature dimension with the maximum weight invarious types of data as a category feature; identifying a first-level category, and marking a category identifier on the sample sequence; carrying out primary classification, carrying out secondary classification on the steady-state slowly-varying data by adopting a linear regression classifier, and carrying out secondary classification on the dynamic rapidly-varying data by adopting an autocorrelation characteristic analysis classifier to obtain a classification result. According to the method, feature extraction and effective classification of aircraft telemetry data can be realized; a classification result covers the main category of telemetry data behavior characteristics in the field of aircraft measurement and control, classifier parameters can be used as input conditions of a compression transmission bandwidth or other signal processing methods, and technical support can be provided for development of future deep space exploration tasks and spatial information networks in China.

The invention provides an energy-saving electrical appliance load type classification and identification method. The method comprises the steps of judging the type of a current energy-saving electrical appliance, acquiring a feature class center vector of a single-body energy-saving electrical appliance, acquiring an SVM kernel function according to an SVM algorithm, acquiring a single-body energy-saving electrical appliance training set in the single-body energy-saving electrical appliance load type, identifying the load type of the single-body energy-saving electrical appliance according to an AdaBoost algorithm so as to acquire a single-body energy-saving electrical appliance training model of the single-body energy-saving electrical appliance, judging the type of the energy-saving electrical appliance, acquiring a variable working condition load identification model, combining the single-body energy-saving electrical appliance training model of each single-body energy-saving electrical appliance and the variable working condition load identification model so as to acquire a combined energy-saving electrical appliance training model. The method provided by the invention can identify the load type of an energy-saving electrical appliance quickly and accurately, and an electric energy measurement algorithm is improved so as to ensure the performance of the energy-saving electrical appliance; a basis is provided for targeted algorithm improvement research on electric energy measurement; and thus the running stability and the reliability of the energy-saving electrical appliance are ensured.

The invention discloses a method for exporting high-capacity geographic spatial data to be VCT files, belongs to the field of geographic information system data exchange, and aims at solving the problem that spatial data with big data sizes cannot be exported to be VCT files at present. The method mainly comprises the operation steps of: pre-defining a group of data structure through referring toa VCT file structure; setting a cache folder; reading basic information of a database; reading basic information of to-be-exported feature classes; writing file header/feature class definition/attribute table definition into a cache file; processing the to-be-exported feature classes; and splicing the to-be-exported feature classes into a big file. The method is capable of thoroughly solving the problem that the operation of exporting high-capacity spatial data to be VCT files is limited by memories, is ingenious in conception and high in reliability, is capable of efficiently realizing rapidconversion between high-capacity spatial data and VCT files, has strong practicability and has significance in the field of geographic information system exchange.