39results about How to "Guaranteed retrieval efficiency" patented technology

The invention discloses an image searching method, which comprises a training part and a searching part, wherein the training part comprises the following steps of: the extraction of characteristic points, the supplementation of the characteristic points and the determination of matching relationships, the generation of similar point set, the clustering of the characteristic point sets and the generation of characteristic vectors of each image in an image database; and the searching part comprises the following steps of: extracting the characteristic points of a picture to be retrieved and generating the characteristic point sets; calculating distances between each characteristic point descriptor vector and corresponding cluster centers, and determining a cluster where a current characteristic point belongs by using a smallest distance; calculating the frequency ni of each cluster where the characteristic points of the picture to be retrieved belong; based on the frequency ni of the clusters where the characteristic points of the picture to be retrieved belong, and the probability logarithm wi of each cluster, generating and unitizing the characteristic vector; and calculating Euler distances between the characteristic vector of the picture to be retrieved and the characteristic vectors of each image in a picture library, and selecting the image output with the smallest distance as a searching result.

The invention discloses a method for managing an image warehouse, which carries out efficient management on the image warehouse by using text-based image data description information. The management process comprises the steps of creating and uploading image data index files, searching keywords and downloading image data; the method comprises image data uploading and downloading; and when a client user uploads the image data to the image warehouse, the user is made to have the unique modification operation authority to the warehouse in a way of locking. The whole uploading operation and the whole downloading operation are performed through a data description file, and the data description file records the detail information of all image data such as the subfield number, the storage path and various attributes. The method has the advantages that: the same image data description files are kept in a client and a server, so that the operation of searching the image data by the user can beperformed at the client, the pressure of the network and the server is lightened and the efficiency of warehouse management is fundamentally improved.

The invention discloses a fuzzy clustering based image retrieval method. The fuzzy clustering based image retrieval method comprises the following steps of S11, establishing a characteristic value library for images in an image library and numbering the images ; S12, selecting N images with image separation distances larger than a distance threshold value A1 from the image library and conducting first-time classification on the residual images to form N categories of image sets; S13, conducting the step S12 on the image sets with image quantities larger than a quantity threshold value in the N categories of image sets till the image quantity of each image set is smaller than the quantity threshold value and obtaining M representative points; S14, partitioning all images in the image library into the image sets represented by highest-similarity-level representative points according to similarity levels between the images and the M representative points; S15, conducting characteristic valuing on input images to be retrieved, respectively calculating the similarity levels between the input images and all representative points and selecting a plurality of highest-similarity-level representative points to perform retrieval. The fuzzy clustering based image retrieval method narrows a retrieval range on the basis that retrieval efficiency is ensured, and reduces retrieval working amount.

The present invention discloses a keyword searchable encryption method based on a Bloom filter with a storage structure, which the following steps: step 1, system parameter initialization; step 2, user secret key generation; step 3, retrieval index generation; and step 4, search token generation; step5,cloud storage server search on cryptograph key word. The present invention allows users to use trap door search linking with a key word to encrypt a document, allows data users to encrypt their data and store the data in the cloud storage server, when needed, the needed cryptograph data can be searched by a key word search token, and then is loaded and decoded. The present invention solves the problem of low retrieval efficiency of a conventional encryption method in the prior art.

The invention provides a structured information retrieval method and system and a method and system for building indexes of structured information. The structured information retrieval method comprises the steps of presetting the indexes of structured information; receiving structured information inquiry request which is submitted by users and comprises inquiry keywords; pre-processing the inquiry keywords; matching the pre-processed inquiry keywords in inverted indexes of each domain, and obtaining candidate structured information meeting preset matching rules; extracting serialization indexes of the structured information corresponding to the candidate structured information, and calculating values of the candidate structured information according to preset scoring rules and the serialization indexes of the structured information; and selecting targeted structured information serving as inquiry results according to the values of the candidate structured information. According to the structured information retrieval method and system, structured information retrieval based on Spain and other languages can be achieved, and retrieval efficiency and system stability are guaranteed.

The invention relates to a method for big data retrieval based on time characteristics and supporting complicated conditions. According to timestamps generated during data storage or according to user defined time characteristics, data information is segmented according to the time characteristics, indexes are established on different collections on a full-text retrieval engine Solr; on the conditions that users already know the time of queried information generation, quick positioning can be performed on an index database of a collection of the same month to perform retrieval, the retrieval range can be greatly reduced, and the efficiency is improved. On the conditions that users cannot determine information generation time, results conforming to user query are provided for users for reference and selection. According to the method, not only is the shortcoming that a distributed database HBase has no secondary index overcome, but also the index establishment is more flexible, the retrieval on different demand conditions is met, and the retrieval efficiency can be guaranteed.

The invention discloses an image retrieval method based on a self-adaptive rectangular window. The method specifically comprises the following steps: step A, extracting local unchanged characteristics of database images, clustering space coordinates of characteristic descriptors in each database image using a G-means method; step B, according to classification of G-means clustering in each database image, establishing a self-adaptive rectangular window, and removing a sparse rectangular window, merging rectangular windows, and removing small rectangular windows; step C, respectively vectorizing the self-adaptive rectangular windows in the database images, establishing self-adaptive window vectors, and based on the window vectors, establishing an inverted index; step D, vectorizing a to-be-retrieved target region, performing similarity search in the inverted index, to giving a final retrieval result. Compared with the prior art, the method improves retrieval accuracy while ensuring retrieval efficiency.

The invention provides indexing equipment which comprises an irregular character string segmentation module and an indexing establishment module, wherein the irregular character string segmentation module is used for segmenting an irregular character string in a document, and the indexing establishment module is used for establishing an index based on the character string obtained by being segmented and the document. Due to the fact that the technical scheme includes that the irregular character string in the document is proposed to be segmented and the new segmented character string is used as a keyword to establish the index, the indexing equipment has the advantages of being capable of reasonably controlling the length of a file list pointed by the keyword in an index structure and the length of a keyword list, namely, controlling the size of the indexing within the acceptable range of computer memory and improving indexing speed of the character string. The invention further provides an indexing method, a search device, a search method and a search system.

The invention discloses an APP similar icon retrieval method and system based on a convolutional neural network. The method comprises the steps that a sample feature vector, sample file identification and N indexes are saved in the retrieval system; for target APP icons, a target feature vector is uniformly divided into N target parts, combination is conducted on each target part, and M target combined feature vectors are obtained; retrieval is conducted M times in the retrieval system for M combined feature vectors, M file identification collections are obtained, a unit is taken from M file identification collections, linear computation is conducted in the unit, the similarity of sample APP icons and the target APP icons is obtained, and the sample APP icons are sorted by means of the similarity. Accordingly, the defects that in traditional image retrieval, the accurate rate is low, and the retrieval efficiency is low are overcome; meanwhile, great convenience is brought to a user, and corresponding cellphone application programs can be sought according to the icons.

The invention discloses a numerically-controlled machine tool design module three-dimension model retrieval method based on a single image. A user is free of providing a three-dimension model or sketch and is required to input a similar image of the model to be retrieved, and the required numerically-controlled machine tool design module three-dimension model can be retrieved. In the training process, a three-dimension CAD model in a database can be displayed in a multi-view manner, SIFT feature extraction is performed on the multi-view images of the model, SIFT histograms are calculated, and features of feature bags are generated for rapid retrieval. The retrieval segment is divided into two steps, namely rough retrieval and accurate retrieval. The method has the advantages that the problem that conventionally the three-dimension model or sketch is difficult to acquire and serve as the retrieval condition during three-dimension model retrieval on the basis of content is solved, and the user is capable of utilizing the acquired image as the retrieval condition for retrieval easily; meanwhile, by means of the two-step retrieval method, the efficiency and accuracy of the retrieval are guaranteed.

The invention provides an Solr-based unstructured file retrieval method and device, equipment and a storage medium. The method comprises the following steps: pre-utilizing a Solr search server to index unstructured files to be indexed in a database to generate an index information table, wherein the index information table comprises an index relationship between keywords and unstructured file numbers, and the keyword in the index relationship is located in the unstructured file corresponding to the unstructured file serial number in the index relationship; when a retrieval request initiated by a user is received, determining a retrieval keyword according to the retrieval request; based on an index relationship included in the index information table, determining an unstructured file number corresponding to the retrieval keyword; and returning a retrieval result according to the determined unstructured file number. According to the scheme, the retrieval efficiency of the unstructured file can be effectively improved while the method can adapt to larger data volume of the retrieved database.

The invention relates to the technical field of process recommendation, in particular to a process recommendation model training method, a process recommendation method and electronic equipment, and the training method comprises the steps: obtaining sample processing feature sub-graph pairs and target similarity between the sample processing feature sub-graph pairs; the target similarity is determined according to a similarity measurement mode corresponding to the attribute type of each node in the sample processing feature sub-graph pair, and the attribute type comprises a quantitative attribute and a semantic attribute; inputting the sample processing feature sub-graph pair into a preset process recommendation model to obtain a prediction similarity; and based on the difference between the prediction similarity and the target similarity, adjusting parameters of a preset process recommendation model to determine a trained target process recommendation model. The proposed process recommendation model performs recommendation for the process knowledge graph, and the accuracy and recommendation efficiency of the target process recommendation model are improved by converting graph structure data of different node attribute types into a vector form and performing calculation between vectors.

The invention discloses an image retrieval method based on symbiotic SURF features in the technical field of computer image processing and pattern recognition. The method includes: firstly, detectinglocal features of a database image and an image to be retrieved, extracting a symbiotic SURF feature from each image, and then quantifying the symbiotic SURF features extracted from the database imageand the image to be retrieved, generating symbiotic visual phrases, and then creating a multi-dimensional inverted index according to the symbiotic visual phrases in the database images, in the inverted index, using the symbiotic visual phrases of the images to be retrieved for similarity searching to find the candidate database images, finally, judging the consistency of the main direction between the images to be retrieved and the candidate database images, and providing the final image retrieval results. The method improves the retrieval accuracy under the condition of guaranteeing the retrieval efficiency.

The invention discloses a medical image retrieval system based on computer processing in the technical field of image retrieval. The medical image retrieval system comprises an acquisition image subsystem; the acquisition image subsystem is in electrical output connection with a central processing unit; the central processing unit is electrically and bidirectionally connected with a retrieval subsystem; the central processing unit is in electrical output connection with a display subsystem; and the display subsystem is electrically and bidirectionally connected with a feedback subsystem. According to the medical image retrieval system disclosed by the invention, information in an input image is directly read by the acquisition image subsystem so as to ensure accuracy of retrieval, benefit to accuracy of analysis on pathology and reduce individual subjectivity and uncertainty; by the central processing unit, retrieval of the image is ordered, and rapid retrieval efficiency and accuracy are ensured; and by the retrieval subsystem, information to be retrieved can be rapidly, accurately and respectively retrieved from a computer database, the internet and an index database, and accuracy of the information and a wide retrieval range are ensured.