262results about How to "Achieve retrieval" patented technology

The invention discloses a modified local sensitive hash vehicle retrieval method based on multitask deep learning. A multitask end-to-end convolution neural network is used to identify a vehicle model, a vehicle system, a vehicle logo, a color and a license plate simultaneously in a subsection parallel mode. A network module for extracting vehicle image example features based on a characteristic pyramid and an algorithm by using a modified local sensitive hash sorting algorithm to sort the vehicle characteristics in a database, and a cross-modal text retrieval method when a retrieval vehicle image can not be acquired are included. The multitask end-to-end convolution neural network and the modified local sensitive hash vehicle retrieval method are provided, the automation level and the intelligence level of vehicle retrieval can be improved effectively, little storage space is used, and image retrieval requirements in a big data era are met by using a quicker retrieval speed.

The invention discloses a hierarchical landmark identification method integrating global visual characteristics and local visual characteristics. High-dimensional characteristic vectors of landmark images are obtained and are used as the global visual characteristics of the landmark images; the local visual characteristics of the landmark images are obtained; the global visual characteristics andthe local visual characteristics are stored by adopting a hierarchical tree-shaped structure, and a visual characteristic set is obtained; each image is characterized according to the visual characteristic set; the images are pre-retrieved according to the global visual characteristics xi, and first candidate images are obtained; the first candidate images are further retrieved according to statistical characteristics vi of local outstanding points, and second candidate images are obtained; and the second candidate images are further retrieved according to a characteristic set yi of the localoutstanding points, and final candidate images are obtained and are fed back to a user. By adopting the hierarchical landmark identification method, the images to be identified can be rapidly and accurately retrieved, so the requirement of the user for convenient information acquisition is satisfied; and besides, through removing certain mismatching points, the landmark identification accuracy isimproved and the landmark identification complexity is reduced.

The invention discloses a fuzzy multi-keyword retrieval method of encrypted data in cloud environment. A file is subjected to set encryption by a data owner to generate a ciphertext file; keywords are extracted from each file; the keywords are subjected to binary segmentation and vectorization to obtain a binary vector group; the binary vector group is subjected to dimensionality reduction and is then inserted into a counting type bloom filter to generate index vectors; each index vector is encrypted to obtain a security index; the ciphertext file and the security indexes are sent to a cloud server; a pre-authorized data user or the data owner extracts the keywords from query data; binary segmentation and vectorization are performed to generate a query vector; encryption is performed to obtain a trap door; the trap door is sent to the cloud server; the cloud server obtains a certain number of files with the highest relevancy degree through query according to the trap door and the security index; after sorting, the files are returned to the data owner. The large data volume of ciphertext multi-keyword retrieval is supported; compared with the prior art, the method has the advantages that the index building and query processes are more efficient; the sorting result is more accurate; the data privacy is effectively protected.

The invention relates to a voice recognition method, a voice recognition device and a voice recognition system. The voice recognition method comprises the following steps: preprocessing picked voice data to obtain preprocessed voice data; extracting voice features in the preprocessed voice data; matching the voice features with a local voice feature database, and if the matching is unsuccessful, sending a first voice recognition request to a target server; receiving a first voice recognition response which includes the matching result of the voice features and the local voice feature database through the target server returned by the target server and sending a second voice recognition request to the target server when the first voice recognition response presents that the matching is unsuccessful, and receiving a second voice recognition response returned by the target server, wherein the second voice recognition response comprises a voice recognition result obtained after the preprocessed voice data is sent to a manual translation terminal to be manually translated. According to the technical scheme of the embodiment of the invention, the voice recognition rate of an intelligent terminal is greatly improved.

The invention discloses an intelligent Operation and Maintenance Analysis System Based on Multi-source Heterogeneous Data Fusion, Machine Learning and Customer Service RobotBased on multi-source heterogeneous data fusion, In the intelligent operation and maintenance analysis system of machine learning and customer service robot, the analysis and processing module uses the multi-source heterogeneous data fusion technology to preprocess and normalize the real-time data. After the correlation analysis, the model matching is used to output the results, and the root cause of the problem is determined by root cause analysis. The model management module is used for acquiring data for machine learning and outputting various reasoning models to the analysis processing module and the recommendationdecision module. The recommendation decision module recognizes the context and matches the corresponding subsequent processing behavior according to the policy. The invention is based on the multi-source heterogeneous data fusion technology and the operation and maintenance mode of machine learning, which gives consideration to the compatibility problem of the big data, improves the automation degree of the operation and maintenance, guarantees the network safety and the service quality, simultaneously has certain prediction ability, and reduces the related operation and maintenance cost. Theoutput mode of the customer service robot further realizes the intelligence of operation and maintenance.

The invention discloses an active detection and recognition system and method for a human gait behavior based on semantic folding. Embedded gait behavior detection and recognition system hardware with low power consumption is built by using a three-dimensional sound field positioning system, a sole force field positioning system, an HDMI high-definition camera, a high-definition video acquisition system, a microcomputer Raspberry and the like. Gait semantic energy diagrams with timing characteristics provided by the invention contain gait time information under different situations, and the learning and prediction ability of a gait behavior cognition system can be enhanced according to a lot of gait semantic energy diagrams with timing characteristics. Meanwhile, the active detection and recognition technology for a human gait behavior based on semantic folding has extensive application prospects in various fields which mainly comprises the fields such as long-distance identity recognition, abnormal gait behavior detection, pedestrian behavior prediction and mass video retrieval, and has excellent economical and social benefits.

The invention relates to a tone-similarity-based song retrieval method, aims at content-based music retrieval technology and provides a method of personal song modeling, calculation, retrieval, and matching based on MFCC (Mel frequency cepstrum coefficient) and GMM (Gaussian mixture model) and implementation of the method by comprehensively utilizing features such as song background music tone and voice characteristics of singers. Experimental results show that a system using the method is fast in retrieval, high in stability and highly expansible. The method is especially suitable for retrieval of audios with high requirement on tone, such as instrumental music audio retrieval systems and multimedia audio management systems.

The invention relates to the field of flight devices, and provides a flight vehicle searching device and a searching method thereof. The searching device is arranged on a flight vehicle, and works independent of the flight vehicle. The searching device comprises a controller, and a power supply module, a wireless communication module and a positioning module which are connected with the controller. When the flight vehicle is working or stops working, the controller judges whether or not to send a distress signal to the outside. After judging that there is a need to send a distress signal to the outside, the controller acquires the current location information of the flight vehicle from the positioning module, and sends a distress signal containing the acquired location information to a handheld terminal of a user through the wireless communication module. The handheld terminal determines the location of the flight vehicle according to the location information in the received distress signal. Compared with the prior art, as the searching device is an independent device and can work independently, the searching device is not affected by flight vehicles, the possibility that a flight vehicle is found is improved, and the economic loss of users is reduced.

The invention relates to an intelligent search based quick searching method of a power grid enterprise information integrating system. The method comprises the following steps of: A. carrying out full-text search by using a search engine based on keywords; B. establishing indexes by using an open source Chinese word segmentation algorithm and a word base; distinguishing the data type, listing the attribute information to be indexed for the attribute information of the core data of the power grid model and the property model, establishing the indexing and uniform designating method, and only indexing the measuring type element data for the measuring data amount; carrying out indexing maintenance on off-line global index, on-line index and proper optimum index; C. providing a keyword searching input field, and indexing the specific data attribute of the index file according to the input keywords; D. realizing the classification index on the common information by the classification index search. The classification index search can realize the index of the common information, such as substations, electric power plants, lines and running data, and then the user can visually index the data.

The invention provides a steel pipe magnetic powder inspection fluorescent image detection apparatus and detection method. The detection apparatus comprises a steel pipe motion control system, a camera, an ultraviolet flash lamp containing an ultraviolet optical filter, a fluorescent optical filter installed before the lens of the camera, a magnetic powder image electric control system connected with the steel pipe motion control system and a monitoring computer connected with the magnetic powder image electric control system, wherein the magnetic powder image electric control system is also connected with the camera and the lens, and the monitoring computer is connected with the camera and a production management system. With the detection apparatus and detection method, the problem of poor imaging quality caused by weak magnetic powder fluorescence, insufficient brightness, complex steel pipe shapes, insufficient depth of field and motion smear of a steel pipe is overcome, fluorescent steel pipe images with high brightness, good size resolution, distinct gradation, low noise and high quality are obtained, and displaying, storage, retrieval and playback of fluorescent steel pipe images, detection of suspected defects and further confirmation of flaw detection defects of the steel pipe can be realized.

The invention discloses a vehicle image retrieval method and a vehicle image retrieval device for a traffic block port. The vehicle image retrieval method comprises the following steps: determining block port vehicle reference images and various block port vehicle retrieval images as retrieval objects; acquiring textural feature codes of the block port vehicle reference images; aiming to the block port vehicle retrieval images, respectively acquiring the textural feature codes of the block port vehicle reference images, and calculating the similarity between the textural feature codes of the block port vehicle retrieval image and the textural feature codes of the block port vehicle reference image; sorting the block port vehicle retrieval images according to the similarity sequence from high to low, and displaying to a user. With application of the scheme, the accuracy of the retrieval result can be improved.

The invention belongs to the technical field of information hiding, and aims to provide a high-capacity encrypted domain information hiding method capable of separating a carrier image decryption process from a secret data extraction process. According to the specific technical scheme, the high-capacity image encryption domain information hiding method is composed of the following four parts: 1, calculation of a prediction difference value and generation of a label map; 2, encrypting a carrier image; 3, storing a prediction difference value and embedding secret information; 4, reversible recovery of the carrier image and lossless extraction of secret information; the whole technical scheme is simple and easy to implement, information such as user copyright, identity authentication or content retrieval can be hidden in the encrypted image according to different requirements of a user for image protection, and management of copyright protection, security authentication, retrieval, classification and the like of the image is achieved.

The invention belongs to the technical field of digital image processing and particularly relates to a remote sensing image classification and retrieval method. The remote sensing image classificationand retrieval method comprises the following technological processes: constructing a training data set, constructing a scale space, constructing local features, constructing global features, constructing regularized classification features, constructing a to-be-classified data set, constructing a retrieval visual word bag and retrieving remote sensing images. The remote sensing image classification and retrieval method provided by the invention has the benefits that the local features of classification are found on different scale spaces of the remote sensing images by adopting scale-invariant feature transform; the global features of the remote sensing images are constructed by adopting a generalized search tree, and a Gaussian weight function is introduced for the regularization fusionof the local features and the global features; the classification of the remote sensing images is developed on the basis of the regularized fusion of the local features and the global features, so that the retrieval of the images is finally realized; a principle is scientific and reasonable; the fusion of the local features and the global features of the remote sensing images can comprehensively depict the multi-scale, spatial and textural features of the remote sensing images, so that ground features of different sizes and orientations are completely classified.

The invention discloses a connection keyword retrieval method based on attribute encryption in a hybrid cloud. The method comprises the following steps: the registration step; the initialization step;the keyword index generation step; the file retrieval party authorization step; the search trap door generation step; and the retrieval and result returning step. Through the method disclosed by theinvention, the connection keyword retrieval under the ciphertext environment in the hybrid cloud is realized, the user privacy information is guaranteed, and massive calculation tasks are transferredto a private cloud, so that this system is suitable for resource-limited mobile equipment.

The invention discloses an information system relating to the index, searches, storage and control displaying of the related data. The invention index system at least comprises two data blocks with relating relationship, each data block consists of position information blocks and content information blocks, the content information blocks are the other data besides the position information blocks in the data blocks, the position information blocks record the position information of the other data blocks related to the data blocks and generate index data according to the prescribed steps; the index method is that: the method generates data block aggregation according to the information input by the users, uses the method of generating key words by corresponding index system to generate index words, and then proceeds index to the index words and displays the result. The invention proceeds the displaying control based on the data blocks and the geometrical relationships in the existing displaying district. The invention is utilized to directly index the relating information in the related data, so as to conveniently realize the concurrent processing of multi-users and commonly share the information which is managed.

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 forage detection method based on machine vision. The method comprises the following steps: dividing a pasture feed tough region to be monitored to obtain a plurality of forage detection regions; based on the edge detection on the minimum rectangular region of a region of interest in the forage detection regions, acquiring a forage detection region with livestock; based on RGB color components, detecting the minimum rectangular region of the region of interest in the forage detection region without livestock to obtain relevant forage detection information; displaying the obtained forage detection information. The invention further discloses a forage detection system based on machine vision. The forage detection system can automatically detect the surplus of forage in pasture feed toughs and can set alarm conditions with varying degrees to remind managers according to the consumption condition of forage. The forage detection system can also record the historical data of forage consumption to realize inquiry and retrieval, and further realizes intelligent monitoring and management on the forage condition based on machine vision on the basis.

The invention provides an applet page acquisition method, a server, a client and electronic equipment. The method comprises the following steps: receiving retrieval information sent by a first clientthrough a host program; determining page key information matched with the retrieval information, and determining page path information corresponding to the page key information according to a mappingrelation; wherein the mapping relationship comprises a corresponding relationship between the page key information and the page path information; and sending the page path information to the first client, so that the first client obtains a page comprising the page key information corresponding to the page path information. Therefore, the server retrieves the small program page according to the retrieval information and sends the page path information corresponding to the retrieved small program page to the client, so that the page content of the small program is retrieved, and the retrieval efficiency is improved.

The invention relates to an interactive query method suitable for multiple big data management systems. The method comprises the steps of 1) building an associated document model, which comprises a document set and an association set, wherein the association set is a set composed of associations among documents; 2) converting different original data models into associated document models, and connecting different data sources integrally through the associated document models; 3) based on the associated document models, establishing a unified query language suitable for multivariate data; and 4) by utilizing the unified query language suitable for the multivariate data, realizing unified query of a relational database, a graph database and a file system. According to the method, the unified query language suitable for the multivariate data management system is proposed for the first time, so that the unified query of the relational database, the graph database and the file system can be realized.

The invention relates to a video processing method based on a geographic position. The method is applied to a mobile photographing terminal. The mobile photographing terminal comprises a camera. The method comprises following steps of invoking the camera to photograph, thus obtaining video data; obtaining current to-be-sent video frames according to the video data; obtaining additional information corresponding to the current video frames, wherein the additional information at least comprises current positioning information of the mobile photographing terminal; adding the additional information to the data packets of the current video frames; and sending the data packets of the current video frames added with the additional information to a cloud server or a remote client, thus enabling the cloud server or the remote client to process the video frames according to the positioning information in the received data packets. According to the method, each video frame comprises the positioning information and also comprises the other additional information; the video data can be searched based on the additional information and the geographic position, thus providing rich video applications on this basis.

The invention discloses a video monitoring image storage and retrieval method. The method comprises the following steps that position parameter information and optical parameter information of a camera are synchronously recorded in a video monitoring image file; after the video monitoring image file is transmitted to a server, the server obtains a space-time coverage range corresponding to a video monitoring image and a camera motion state according to the position parameter information and the optical parameter information, wherein the X axis of the space-time coverage range serves as a longitude, the Y axis of the space-time coverage range serves as a latitude, and the Z axis of the space-time coverage range serves as the time, and the camera motion state comprises rotating and stretching operations. A monitoring video space-time database is formed by the space-time coverage range, the motion state and the corresponding video monitoring image; inquiry of the monitoring video on the basis of time and space is achieved through the monitoring video space-time database. According to the video monitoring image storage and retrieval method, the defect of video data retrieval based on image content is overcome, and retrieval based on the time-space coverage range is achieved.

The invention discloses a power information retrieval method and a power information retrieval system based on cloud storage. The power information retrieval method comprises the following steps of carrying out participle processing on each document prearranged in a power database according to a participle library of the prearranged power theme word list; establishing an index database according to the obtained participle results; memorizing the power database and the index database in a cloud storage system; obtaining the retrieval information of a user; determining the key word of the retrieval information and the logic relationship among the key words according to the retrieval information; searching and obtaining the document set corresponding to each key word from the retrieval database; processing the document set according to the logical relationship among the key words and obtaining the document information corresponding to the retrieval information; and exhibiting the document information to the user. The invention also provides a power information retrieval system based on cloud storage. The method and the system can realize the retrieval on the power information and the retrieval speed can be improved.

The invention relates to a system used for retrieving and sharing large data center platform encryption files. The system comprises a data acquisition module used for carrying out data acquisition of user files of a user initiating a transmission request to a large data center platform and guaranteeing confidentiality and integrity of the user data, a data pre-processing module used for carrying out primary processing on the user data in the user files acquired by the data acquisition module to construct corresponding indexes and encrypting the indexes, a data storage module used for encrypting and storing the user files after data processing carried out by the data pre-processing module and storing the indexes, a sharing and exchanging module used for realizing encryption searching for the encrypted files and realizing safe sharing and exchange among different user files, and a secret key management module used for managing secret keys for encrypting the user files and secret keys for encrypting the indexes. The invention further discloses a method used for retrieving and sharing the large data center platform encryption files.

The invention provides a method for retrieving contents of application software. The method comprises: generating an execution script according to operation records of an App; instructing a test device to operate the corresponding execution script on the App; receiving the contents shown in the App in the process the test device operates the execution script; and searching a retrieval object in the shown contents. The technical scheme realizes retrieval of contents of the App. The methods and the devices are suitable to be used for all Apps, and have no limitation on development of the App.

Provided is a vehicle search method based on multiple views. The method includes that a direction dividing model of potential gestures of vehicles is built, the direction dividing is defined with the front camera shooting point of a camera as a reference point in the direction dividing model, a plurality of directions are divided evenly, a local block of each direction is set in advance, the corresponding relation between the local blocks of two directions is built, a vehicle search base is built, vehicle gesture estimation is conducted on images for vehicle detection and search vehicle pictures in the vehicle search base, the directions of the vehicle gestures are obtained according to a direction dividing model, visual characteristics of the local blocks of the direction are extracted according to the divided vehicle gestures, target vehicles are shown by adopting a local two-dimensional model according to the visual characteristics of the local blocks, visual similarity comparison is conducted according to the corresponding relation of the local blocks of the corresponding relation of the local two-dimensional model to obtain similarity and acquire a vehicle search result.

The invention discloses a method and device for retrieving film and television data and relates to the field of information retrieval. The method and device is used for solving the problems that in the prior art, film and television data retrieval is tedious, and accuracy is low. The method includes the following steps: a data base is set and used for storing the film and television data; a first key word and / or label information is obtained; according to the first key word and / or the label information, retrieval is carried out in the data base, a retrieved result is obtained, and the film and television data corresponding to the retrieved result are displayed. The method and device is suitable for the field of the information retrieval and used for retrieving the film and television data.

The invention discloses a semantic sparse Web service discovery method based on Gaussian LDA and word embedding, and specifically relates to the technical field of service computing. The method specifically comprises the following steps in order: collecting a Web service description document, preprocessing the collected Web service description document, obtaining characteristic vocabularies in the Web service description document, preprocessing the vocabularies in the Web service description document, and obtaining a set of a group of prototype words; using a word embedded training model Word2Vec to train the set of the prototype words obtained in step 1, and obtaining a continuous vector representation of each word in the set of the prototype words; using a Gaussian LDA model to train the set of the prototype words obtained in step 1, and acquiring each Web service hierarchical structure; enriching user query by using a set of continuous vectors obtained in step 2 by training, and obtaining extended user query; and using the service hierarchical structure obtained in step 2 and using a probability ranking method to obtain query output corresponding to the extended user query.

The invention provides a capsule endoscopy image retrieval method based on visual vocabularies and local descriptors. The method concretely comprises the following steps of: building an image library, wherein the image library comprises standard case images and images to be retrieved; converting capsule endoscopy images from an RGB (Red, Green and Blue) color space into an HSI (Hue, Saturation and Intensity) color space, performing quantization processing and extracting a color histogram; extracting an LBP (Local Binary Patterns) texture histogram of the capsule endoscopy images; performing weighing blending on color features and texture features to obtain the visual vocabularies of the images; extracting Sift (Scale invariant feature transform) feature descriptors of the capsule endoscopy images; and performing similarity testing to obtain retrieval results. The method provided by the invention is used for capsule endoscopy image processing; the work intensity of film reading doctors is effectively reduced; and the diagnosis efficiency is improved.

The invention provides a method and a device for obtaining remotely sensed data based on grids. The method comprises the following steps of: receiving an inquiring request sent by a request module through a source catalog service of a local node; according to inquiring conditions in the inquiring request, obtaining a physical address of a target node of the remotely sensed data meeting the inquiring conditions, and returning the physical address to the request module, wherein the inquiring conditions at least comprise part metadata information of the remotely sensed data; receiving a downloading request sent by the request module; and according to the downloading request, starting a document transmission service of the local node and the target node to transmit the remotely sensed data, wherein the downloading request at least carries the physical address of the local node and the physical address of the target node. The device comprises an obtaining module, a request module, a starting module and a receiving module. The invention provides the method and the device for obtaining the remotely sensed data based on grids, realizing the share of the remotely sensed data and simultaneously solving the problems of the safety, the heterogeneity and the distribution in the share process of the remotely sensed data.