261results about How to "Reduce retrieval time" patented technology

The invention discloses a method for recommending personalized information, which is suitable for the wireless Internet. The method comprises the following steps of: recording user behaviors of accessing the wireless Internet and transmitting to a server by using a behavior recording module on a mobile phone; analyzing the contents of pages accessed by a user to obtain a page set interesting to the user; analyzing the user behaviors of accessing the pages to obtain the interestingness of the user for the pages; combining the interested page set and the interestingness of the pages to establish a user interest model; and carrying out dimensional simplification to a user term matrix and calculating the similarity of term sets to generate recommended personalized information. By adopting the technical scheme, the method can automatically record and analyze the user behavior and excavate user attributes and preferences to realize the matching of the user attributes and information contents and actively recommend the information to the user.

The invention discloses an intelligent inspection system and an inspection method for an electric transmission line by an unmanned aerial vehicle. The intelligent inspection system comprises an inspection mission planning system, a dispatching system, a monitoring system, an inspection result processing platform and mobile branch stations, wherein the inspection mission planning system communicates with the inspection result processing platform and the dispatching system respectively; the inspection result processing platform and the dispatching system communicates with the monitoring system respectively; the monitoring system is connected and communicate with the mobile branch stations. The inspection method has the benefits that factors of the inspection unmanned aerial vehicle, inspection equipment, an inspector, and the like accessing the dispatching platform through a dispatching terminal can be dispatched effectively to realize equipped configuration of unmanned aerial vehicle inspection resource; the usage rates of personnel and equipment can be improved; the no-load rate and vacancy rate of the inspection equipment are reduced; through centralized monitoring of condition information inspected by the unmanned aerial vehicle, unmanned aerial vehicle inspection dispatching personnel can master unmanned aerial vehicle inspection in the large effectively; the service life of the unmanned aerial vehicle is prolonged.

The invention discloses a patrolling result data processing system and method used for unmanned aerial vehicle patrolling. The system comprises an unmanned aerial vehicle, a plurality of ground monitoring stations and a patrolling result processing platform, wherein the unmanned aerial vehicle is used for patrolling a power transmission line, shooting image information of the power transmission line and transmitting the image information to the ground monitoring stations; the ground monitoring stations are used for communicating with the unmanned aerial vehicle, reading flight state data and patrolling result data of the unmanned aerial vehicle and sending the collected information to the patrolling result processing platform; the patrolling result processing platform is used for reading the flight state information and the patrolling result data of the unmanned aerial vehicle from the ground monitoring stations, and carrying out defect diagnosis on the patrolling result data and carrying out CRUD and playing management on defects diagnosed through defect diagnosis. Image defect identification information and image files are stored separately, corresponding defect information is associated with images during defect displaying, and positions, annotation or instructions and the like of the defects of the power transmission line in a patrolled image are displayed, so that efficient and quick processing of the patrolling result data of the unmanned aerial vehicle is achieved.

The invention discloses a method for recommending books. The method comprises: step 1, after acquisition and pre-processing of data, using historical access data and on books by a reader and Web log data to analyze and establish a reader preference mining model to mine current potential preference of the reader, and establishing a preference matrix of the reader on commodities; step 2, according to the established preference matrix of the reader on commodities, in consideration of sparsity of matrix data, using a project scoring and prediction method based on bidirectional association rules to fill missing values, to reduce influence of data sparsity on recommending results, and then using a collaborative filtering algorithm, according to book preference of a neighbor reader whose preference is similar to a target reader, providing recommendation for the target reader. The method and the system for recommending books can recommend book information which satisfies consumer preference to consumers in priority, so as to achieve objective of improving efficiency of retrieving book information by consumers. The method and the system save retrieval time of consumers, and improve retrieval efficiency.

The invention, which belongs to the computer vision field, relates to a method for enhancing a seasonal feature of an image, wherein the method is based on color conversion and editing propagation. The method comprises the following steps that: a target image is input and a specific seasonal image is selected from a scene database and is used as a reference image; and then color conversion is carried out on the target image; foreground extraction is carried out on the target image to obtain outline information of a foreground object; a specific foreground object is selected as a template and a repeated foreground object that is similar to the template is extracted from the target image; an insertion object that is in accordance with a specific seasonal feature is selected from an object database and an insertion position is assigned in a target scene; the insertion object is rotated and zoomed to ensure visual rationality and then editing propagation is automatically carried out on the processed insertion object; at last, fusion is carried out on all inserted objects and a scene image to form a final scene having an assigned seasonal feature. According to the invention, the method can be widely applied and popularized to fields like film and television manufacture and city planning and the like.

This invention discloses one index result page compression display method, which comprises the following steps: a, according to user order index to get result page and imbedding the index engine AJAX leg into the page; b, starting the note browsing page super link aim page; c, according to user trigger information displaying the aim page compression pattern. This invention method adopts AJAX technique to get index result aim page to generate aim page compression without result page to browse aim page.

The invention provides a personnel identification method and system combining video identification and UWB (ultra-wideband) positioning technologies. The method comprise the steps of: A, in the simulation filed of a same supervision area, obtaining geographical positions of mobile positioning objects and the first quantity of the geographical positions according to video identification positioning, and obtaining positioning coordinates of positioning terminals and the second quantity of the positioning coordinates according to wireless positioning; B, comparing the first quantity and the second quantity at a same time, prompting abnormity and entering C if the first quantity is not equal to the second quantity, and prompting normality if the first quantity is equal to the second quantity; and C, obtaining abnormal geographical positions according to the matching between the positioning coordinates and the geographical positions. The method and system can efficiently monitor specific objects and specific occasions such as a campus, analyze and determine in real time position monitoring and video image monitoring information, and timely provide alarm response service, thereby enhancing campus safety service.

The invention discloses an interactive video retrieval method based on unsupervised learning and semantic matching features, which comprises the steps of extracting image bottom layer features and model matching sequence features from the video image frame level of a video database; extracting semantic matching features from the high-level semantic level of image bottom layer features; performing unsupervised learning on the extracted model matching sequence features and semantic matching features, establishing retrieval and direct retrieval based on unsupervised learning, and forming an interactive interface by related feedback. The middle layer features, high layer features, unsupervised retrieval mechanism and interactive mechanism of an integrated video compose a set of new complete video retrieval system, which precisely measures the time-space sequence information of a video object, therefore better retrieval effect is achieved, the semantic understand of a sports video theme is developed, the online calculating complexity and retrieval time of a system are reduced, and the interactive interface greatly improves the retrieval performance of the system.

The invention discloses a commodity purchasing system and method. The commodity purchasing method comprises the following steps of: (S101) reading a target commodity name input by a user and generating a commodity list; (S102) looking for commodity information, which matches with the target commodity name in the commodity list, from a commodity database; and (S103) generating a first display page including all commodity information obtained through matching. The commodity purchasing system comprises a first unit, a second unit and a third unit, wherein the first unit is used for reading the target commodity name input by the user and generating the commodity list; the second unit is used for looking for the commodity information, which matches with the target commodity name in the commodity list, from the commodity database; and the third unit is used for generating the first display page including all commodity information obtained through matching. According to the invention, through simultaneously providing retrieval information of multiple different commodities to user, the actions within the user retrieval time are reduced and the purchasing frequency of the user in purchasing fast-moving consumer goods and other commodities is increased.

The invention discloses a large-scale face identification method based on GPU accelerated retrieval, which relates to the field of computer vision. The large-scale face identification method comprisesthe steps of face detection and alignment, face feature extraction, Hash feature acquisition, face index database establishment, multi-GPU accelerated Rough matching, hash-based candidate set acquisition, precise matching based on distance metric, voting to obtain a best-matched person and the like. The large-scale face identification method based on GPU accelerated retrieval is based on two-stage feature matching of Hash index and multi-GPU accelerated computing, can accelerate the screening of eigenvectors by utilizing the powerful parallel computing capacity of the GPU, greatly reduces time consumption on large-scale data set retrieval, and can meet various application requirements taking the implementation of a deep convolutional neural network as the basis and having high demand on real-time performance.

The invention relates to a trademark image retrieval method which includes: performing color and local feature extraction on sample trademark images; categorizing the sample trademark images according to grey level frequency; performing foreground and background segmentation; determining the edge criteria according to foreground and background pixel gray level statistical values; extracting edge features according to the edge criteria; judging the complexity of the sample trademark images according to the edge features so as to divide the sample trademark images into at least one group; using the same steps to extract the color features, edge features and local features of a to-be-retrieved trademark image; judging the complexity of the to-be-retrieved trademark image according to the edge features of the to-be-retrieved trademark image to determine the group which the to-be-retrieved trademark image belongs to; performing similarity retrieval in the group which the to-be-retrieved trademark image belongs to by synthesizing the color and edge features of the to-be-retrieved trademark image; performing similarity retrieval in other groups when the fact that the to-be-retrieved image is not similar to each sample trademark image in the group which the to-be-retrieved trademark image belongs to is determined; performing similarity retrieval according to the local features of the to-be-retrieved trademark image when retrieval needs to be performed again.

The invention provides a combined inquiring method oriented to an Hbase database and belongs to the technical field of data inquiry. The combined inquiring method includes creating an index table, namely defining Key and Value of the index table, wherein factors of the Key of the index table include N of M combining conditions, factors of the Value of the index table at least include Key corresponding to a main record table, the N is larger than or equal to 2 and smaller than or equal to the M, and the M is an integer larger than or equal to 2; creating indexes; retrieving data, namely performing main key inquiry in the index table according to first Key generated by extracting the N combining conditions, and performing main key inquiry in the main record table at least according to second Key defined by a factor of the Key of the main record table. The combined inquiring method is high in inquiring efficiency and especially suitable for data inquiry of a mass database.

The invention discloses a vertical engine searching method for domain body restraint, which comprises the following steps of: establishing a domain body library and constructing a domain body generator and a domain body importer; performing semantic analysis and body description on network resources by using a web crawler with a semantic analysis filter from various body models of the domain body library, and automatically performing resource conformity calculation and classification to form classification information with semantic features; establishing a semantic relation between the network resources, forming domain resources with semantic relation, and finishing body description of single network resource and storing the network resources; and performing semantic rewriting and mapping on retrieval according to the body, finishing semantic analysis and expansion of the retrieval, taking the body-based resource and retrieval as input, and finishing expansion of the body library and rule restraint expansion through rule learning and modes so as to form a secondary body of the domain body. The method has the advantages that the method with semantic restraint saves the retrieval time, improves the retrieval precision and realizes the advantage of semantic supporting retrieval.

An apparatus, method, and medium for retrieving a highlighted section of audio data using song lyrics. The apparatus, method and medium allow a user to quickly retrieve and recognize desired music by setting a highlighted section of audio data using song lyrics to an audio file and playing the highlighted section during retrieval of the audio file. The method includes extracting title information and lyric information from metadata related to an audio file, checking whether a character string containing title information and a repeated character string exist based on the extracted title information and lyric information, when a character string containing title information and a repeated character string are found, storing the found character strings as highlight candidates, selecting one of the stored highlight candidates as a highlighted character string, and setting a highlighted section containing the selected highlighted character string.

The invention discloses a method and system for rapid video data characteristic retrieval. The method includes the following steps: identifying video data with characteristic parameters in massive video data; extracting the characteristic parameters of the video data, and recording metadata of positions of the characteristic parameters in the video data; generating corresponding lists of the characteristic parameters, the metadata and names of the video data corresponding to the characteristic parameters and the metadata; storing the corresponding lists, extracted characteristic parameters and a characteristic metadata index database; and retrieving characteristic parameters matched with demand characteristics from the characteristic metadata index database according to the demand characteristics, and obtaining the video data corresponding to the characteristic parameters through the metadata. The method and system for the rapid video data characteristic retrieval identifies the massive video data, extracts the characteristic parameters and the positions of the characteristic parameters in the massive video data, compares demand characteristic parameters with the characteristic parameters directly to obtain the positions of the characteristic parameters in the video data rapidly, and accelerates retrieval speed.

The invention discloses a method and a system for dynamically enhancing the input / output (I / O) throughput of a server and aims to overcome the defect that the service performance of the server is limited by the input / output throughput in the prior art. The method comprises the following steps that: the server resolves a read-write request sent by a client into a data block read-write operation request; the server judges whether a target data block is in a cache or not; the server performs read-write operation on the target data block if the target data block is in the cache, and traverses a data access information linked list in a rapid-access disk if the target data block is not in the cache; and the server copies the target data block to the cache to perform data read-write operation if the target data block is found, and reads the target data block from a low-speed disk, saves the target data block in the cache, performs the data read-write operation and maps the target data block to the rapid-access disk if the target data block is not found, wherein the data access information linked list is updated when the data read-write operation is performed. The read-write operation of the server is detected dynamically, so that the hit ratio of a buffer zone is enhanced and the technical effect of high I / O throughput of the server is realized.

The invention provides a house supply recommendation method. The method includes the steps of S1, mapping user requirement information transmitted by a client to several requirement tags; using any ofthe requirement tags to represent any user requirement in the user requirement information; S2, matching house supply tags meeting user requirements according to any of the requirement tags, and extracting the house supply tags from the house supply information of each house supply by using a natural language processing technology; S3, repeating S2 until all the house supply tags matched with therequirement tags are obtained; S4, applying all the house supply tags matched with the requirement labels to obtain several recommendation house supplies, and then bringing all the recommendation house supplies back to the client. The invention provides a house supply recommendation method, by matching purchase demands with the house supply tags acquired by applying the natural language processing technology, the house supplies with higher matching degree with the demands are provided for users, the retrieval time of the users is saved, the user experience is optimized, and the improvement ofthe housing transaction efficiency is promoted.

The invention discloses a retrieval ordering method. The method comprises the following steps of: receiving query phrases, and retrieving; returning retrieval results; extracting the low-dimensional features of all documents in a document library; acquiring relevancy feedback information of a plurality of documents in the retrieval results; generating a relevancy judgment model; calculating the relevancy scores of all documents in the retrieval results; and returning again retrieval results which are ordered from high to low according to the relevancy scores. The invention also discloses a retrieval ordering system, which comprises a retrieval module, a feature extraction module, a relevancy feedback module, an ordering module and a storage module. By the retrieval ordering method and the retrieval ordering system, high-efficiency and accurate retrieval results can be provided.

The invention discloses a voice sectional song search method, which is realized through interaction of a client side and a server side; and a song database is required to be established at the server side. The method comprises the following steps of: during search, the client side prompts a user to speak out song information to be searched in sections and transmits the voice spoken by the user to the server side after the user finishes speaking; after receiving a voice signal, the server side automatically identifies a character corresponding to the voice signal, then searches in the song database in layers and transmits a search result to the client side; and finally, the client side provides the song search result to the user. According to the method, a rational voice interaction flow is designed at the client side, so that a super large song search space can be decomposed into combination of a plurality of smaller song search spaces by the server side; and thus, search efficiency and automatic voice recognition accuracy in the super large song search space are improved, user experience is improved, meanwhile, hardware cost is saved for service providers.

The invention discloses a vehicle feature extraction and matching method based on video, comprising the following steps: reading a video file; performing video image background modeling; the background subtraction method is used to obtain the foreground region containing moving objects. Vehicle detection is carried out by acquiring vehicle contour. Extracting pHash feature from vehicle image; extracting vehicle image ISIFT features; combined with pHash algorithm and ISIFT algorithm, and through BBF search algorithm and RANSAC algorithm to screen ISIFT feature matching points, so as to achievevehicle matching. The invention can realize the functions of vehicle detection, vehicle characteristic information extraction and vehicle matching based on the video, and can detect the vehicle outline of the vehicle very accurately, extract the vehicle information characteristic accurately, meet the real-time requirement with high operation speed and low cost, and can detect the vehicles traveling in different directions, and has wide application range.

The invention discloses a monitoring video multi-target classification retrieval method and system based on depth learning. The method comprises: extracting the object to be retrieved: inputting the image to be retrieved, selecting the object to be retrieved by automatic detection or manual digging, and determining the category of the object; extracting depth features: extracting depth features ofthe object to be retrieved by using depth learning feature extraction model; retrieving object: according to the category of the object to be retrieved, comparing the depth feature of the object to be retrieved with the depth feature of the same kind of object in the retrieval database to obtain the object most similar to it; displaying Retrieval results: displaying the retrieval results according to similarity from high to low. Depth features obtained by depth learning feature extraction module are used as the benchmark to match the retrieved target and the similar target in the retrieval database. Because depth features have the advantages of both strong expression and low dimension, the precision of retrieval results is greatly improved.

A communication assistance system includes a database having a plurality of listings therein, each of the listings includes a least one contact name and a contact information. A means is provided for receiving incoming communications from a plurality of requesters seeking a listing. A table stores past requested listings of the requesters, where the system reviews the table for storing past requested listings and, based on contents contained in the past requested listing table, develops a list of requesters to target a promotion to, whereupon when a requester who is among the list of requesters contacts the system, they are provided with the promotion.

The present invention is directed to a communication assistance system comprised of a first database (18) having a plurality of listings where each of the listings has at least one contact name and a corresponding contact number. A telephony hardware (12) is provided, configured to couple a telephone device (6) of a user to the system. A past requested listing table is also provided for at least one of the users of the system, configured to store past requested listings obtained from the database, as requested by the user. When the user contacts the system with the telephone device, the system recognizes the user and forwards the past requested listing table to an operator terminal of the system.

The invention discloses a database forced access control method and system based on a policy. The method comprises the following steps of: (1) establishing a multi-dimensional index tree for a data table in a database; (2) stacking layer by layer and storing forced access control safety tag setting policies into nodes of the multi-dimensional index tree; and (3) performing user access control on the multi-dimensional index tree with stacked policies. The system comprises a front end and a rear end, the front end comprises a policy managing assembly and an index tree assembly, and the rear endcomprises an access control assembly and a policy modifying assembly. When the database forced access control system is in use, a manager only needs to manage a few of forced access control safety tag setting policies in the mass database so as to greatly reduce the wastes of time and labors. The policies are stacked on the index tree so that the safety tag calculation of the data and the access control based on the safety tag can be integrated with the retrieve process of the data to shorten the additional retrieve time. The executing efficiency and the response rate of the data access are increased.

The invention discloses an open dialogue method and system based on man-machine interaction, and the method comprises the steps: carrying out the voice input query by a user, carrying out the voice recognition by a search engine, carrying out the Lucene retrieval, and enabling the retrieval content to be a retrieval pair which is sorted through a corpus and pre-established with an index; transmitting the retrieval result to the mixed vector model, enabling the mixed vector model to generate a model through a pre-trained vector, answers are expanded, and then the expanded answers are transmitted to the sorting model; and enabling the sorting model to return a result through pre-trained parameters. According to the invention, related technologies in the field of natural language processing are utilized; The semantic and grammatical analysis of the questions and the literature library is emphasized, the questions are subjected to comprehensive and detailed semantic analysis, the method and the technology are applied to sentence analysis of the literature library, a standardized question type library is established, the types of the questions and the types of answers which should be answered are standardized, and the retrieval time is shortened.

The invention relates to a retrieval method of similar numerical control technics of electronic parts based on characters of a three-dimensional model. According to parts technics of actual example model, the technics of actual example already in the database is retrieved based on retrieval algorithm of similar technics level. The retrieved results are displayed to technics designers according to the similarity from high to low order. In similar degree calculation, traditional similarity algorithm, which is relied on the simple sum of individual character similarity between the parts, without considering the relationship between the characters, is abandoned in the invention, and the characteristic triangle is adopted as the basic unit of the similarity calculation, thus more comprehensive similarity Information between parts in topological structure is considered, to improve the accuracy of similarity retrieve; on the other hand, traditional retrieval method of actual example, considering multiple factors in single-layer at same time, is abandoned, and the retrieval method of level by level filter is used, considering few factors in every level, to shorten retrieval time in some extend and improve the retrieval efficiency.

The invention discloses a picture book recognition method. The picture book recognition method is applied to a device with a camera, and comprises the following steps of: acquiring a picture book photo through the camera according to a preset acquisition frequency; uploading the picture book photo to a server; receiving a first audio link, returned by the server, corresponding to the picture bookphoto; if the picture book photo is a picture book cover photo, receiving a picture book ID corresponding to the picture book cover photo; according to the first audio link, connecting a first audio stream in the server and playing an audio. The invention furthermore provides a picture book recognition device and system and electronic equipment. According to the picture book recognition method, device and system and the electronic equipment, the problem that the picture book recognition error rate is high in the prior art is solved.

The invention provides a metadata processing method. The metadata processing method comprises the following steps: obtaining correlation fields between a target data table and different correlation data tables; According to the correlation field, obtaining correlation information between the target data table and different correlation data tables, and constructing a correlation information set ofthe target data table; Obtaining processing task information of the target data table; Determining upstream and downstream relationships between the target data table and different data tables according to the processing task information, and determining an upstream data table and a downstream data table of the target data table based on the upstream and downstream relationships; Wherein the target data table inherits the association information of the upstream data table, and the downstream data table inherits the association information set of the target data table. The invention further provides a metadata processing device and equipment and a readable storage medium. The technical problem that in an existing database scheme, the meaning of metadata service is limited, and consequentlyit is difficult to effectively participate in specific application of a database is solved.