60results about How to "Narrow your search" patented technology

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 invention relates to an image retrieval method and apparatus. The method comprises the steps of: obtaining a to-be-retrieved image from a user terminal; converting the to-be-retrieved image to HSV space from RGB space; quantizing the converted to-be-retrieved image into N-dimensional color features; sorting the feature values of the N-dimensional color features in a descending order, and selecting the first M-dimensional color features as main color of the to-be-retrieved image; determining a cluster index table name of the to-be-retrieved image according to an index value corresponding to the main color; querying whether a cluster index table with the same name as the cluster index table of the to-be-retrieved image exists in the established cluster index table according to a name of the cluster index table; if existence, obtaining an image index of the corresponding image; querying main color percentage of the corresponding image according to the obtained image index; calculating the image similarity according to the main color percentage; and returning an image matched with the to-be-retrieved image to the user terminal according to the similarity. The image retrieval method and apparatus disclosed by the invention remarkably improve the image retrieval efficiency and accuracy.

The invention proposes a color video matting method based on depth foreground tracking, which relates to the field of digital image processing. The method comprises: before the starting of the matting, defining a clear search area by a user at a designated foreground area of a depth image; then in the search box, using the depth information difference of the foreground, segmenting the foreground targets; reconstructing the depth image to the resolution of the color image and performing the registration so that the pixels of the two gradually correspond to each other and generating a three-part map; according to the color information of the color image, making the refined adjustment to the three-part map; based on the three-part map and the color image, using the super-pixel gradient for fast foreground extraction; and finally according to the extracted foreground position, establishing and updating the motion equation; predicting the foreground center of the next frame; and updating the search box position. With the above steps, the video matting can be completed frame by frame. Compared with the current video matting algorithm, the invention greatly simplifies the operation, has low complexity of the algorithm, and realizes the real-time and accurate video foreground matting.

The invention discloses a mobile terminal, and a method and a device for seeking the mobile terminal, and belongs to the technical field of communication. The method for seeking the mobile terminal includes flow paths: sending an appointed communication signal trigger instruction to the mobile terminal through the device for seeking the mobile terminal so as to enable the mobile terminal to start at lease one type of communication signal; seeking the communication signal started by the mobile terminal; when the communication signal of the mobile terminal is sought, judging whether the communication signal is a unique communication signal sent by the mobile terminal; if the communication signal is the unique communication signal, using the communication signal to build communication connection between the device for seeking the mobile terminal and the mobile terminal, and if the communication signal is not the unique communication signal, building communication connection between the device for seeking the mobile terminal and the mobile terminal by selecting a communication signal shortest in effective distance; sending an indication signal to a user, and guiding the user to move towards the mobile terminal. By using the mobile terminal, and the method and the device for seeking the mobile terminal, the device for seeking the mobile terminal can switch between all the communication signals according to the distance between the device for seeking the mobile terminal and the mobile terminal, and the mobile terminal can be sought out whether the distance between the mobile terminal and the user is long or short.

The invention discloses a method for tracing a target object in a panoramic video automatically, which comprises the following steps: (1) acquiring the position of a target object in the panoramic video; (2) detecting if plane crossing appears in the motion of the target object in the cubic panorama, and performing a step (3) if plane crossing appears in the motion of the target object in the cubic panorama or performing a step (4) if plane crossing does not appear in the motion of the target object in the cubic panorama; (3) expanding an image on the original plane of the target object; (4) searching and matching along an epipolar line to obtain the motion vector of the target object; and (5) making a judgment according to the result of the step (4), if a matched target is found, ending the tracing or performing the step (2). The method can be used for tracing the target object in the panoramic video automatically and can narrow search range in a tracing process and ensure the quickness and stability of the tracing of the target object.

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.

Provided is a target tracking method based on a LabVIEW. When a target is not shielded, area template matching is adopted in target tracking, and learning and matching of a template are achieved through a machine vision module in a vision and movement function library arranged in the LabVIEW; when the target is shielded, a certain or some invariance characteristics of the target are extracted, characteristic template matching is carried out, tracking is completed, a search algorithm is brought in the tracking process to carry out estimation hypothesis on the position state of the target at the next moment, accordingly, tracking can be achieved in real time quickly and accurately, the shielded target can be tracked accurately and stably, it is guaranteed that the target cannot be lost in the tracking process, the tracking speed is accelerated, and real-time performance and robustness of tracking are improved.

The invention provides a directory service data retrieval method, which is applied to a directory server with an embedded database having a data structure which is a balanced binary tree. The method includes: obtaining directory service data, packaging the directory service data into data nodes including directory service data key values, inserting the data nodes into the balanced binary tree structure of the embedded database and storing the data nodes as tree nodes of the balanced binary tree; adjusting the balanced binary tree structure of the embedded database to be balanced; and when a user needs to retrieve the directory service data, receiving a directory service data query request which is sent by the user and includes key matching values of the directory service data to be queried, finding the data nodes corresponding to the key matching values in the balanced binary tree structure of the embedded database of the directory server according to the key matching values, and obtaining the directory service data in the data nodes.

The invention relates to a double-frequency GPS signal capturing method based on segmented and split-radix Fourier transform. The method comprises the first step of capturing an L1 signal, the second step of obtaining auxiliary information for capturing an L2C signal according to the carrier doppler frequency shift of the obtained L1 signal and the initial phase position of a C/A code, and the third step of carrying out segmentation on the L2C signal and adopting the split-radix Fourier transform to capture the double-frequency GPS signal on the basis of segments. The computing amount and the storing space are reduced to a very large extent during the capturing process, the computing speed is improved, and a theoretical foundation is laid for achieving an L1/L2C double-frequency GPS signal receiver on an embedded type system.

The invention discloses a computer data induction, arrangement and classified storage method. The method specifically comprises the following steps that firstly, people can operate a computer to opena user to operate a client; the user logs in the data input module and operates the user information login module in the client to input the account information of the user into the system for authority authentication, after authentication is successful, people can input the data to be stored into the pre-storage address in the computer through the data input module, and the computer data processing method and system relate to the technical field of computer data processing. According to the invention, a large amount of time does not need to be spent on retrieving and searching the whole stored file, the time of people is greatly saved, the input data is orderly stored in a single storage module in a classified mode, and the purposes of reducing the retrieval range of the needed file, reducing the processing burden of a processor and shortening all the time of retrieving and searching are well achieved.

The invention discloses a cross-domain vision search method based on significance detection. The method comprises the following steps: firstly, utilizing each super-pixel zone boundary connecting value to endow all the zones with different significance values, thereby acquiring a main object zone; performing multi-scale treatment on a target image in a database; extracting the feature of the main object zone, thereby acquiring a target image feature template; performing feature extraction on the main object zone of an inquired image and training a linear classifier; performing iteration training on a large quantity of negative samples, thereby acquiring an optimized inquired image feature template; finally, while searching, returning a zone with the highest response score as a final searching result according to the matching degree of each target image feature template and the inquired image feature template. According to the invention, the significance detection for the main zone is performed, so that the influence of a background zone on the searching result is reduced, the searching precision and efficiency can be effectively increased in the cross-domain vision search and the robustness is excellent.

The invention discloses a method and device for determining the answers of the question data, which relates to the field of computers, and is used for accurately determining the answers of the question data. The method comprises the steps of obtaining the first question data inputted by a user, wherein the first question data comprises at least one word; determining a word relationship between the words in the at least one word; determining a service type corresponding to the first problem data according to the at least one word, the word relationship and the intention map, wherein the intention map at least comprises a service type associated with the at least one word; and determining the answer of the first question data from at least one preset answer according to the service type corresponding to the first question data.

The invention discloses a method for recovering a file by reducing the initial cluster number of an FAT32 directory entry. The method comprises the following steps: 1, searching a data area, and distinguishing deleted files; 2, acquiring the high level of the initial cluster number of the directory entry according to the directory entry of a deleted file; 3, searching a directory entry of which the high level is zeroed, and recoding; 4, extracting the directory entry recorded in the step 3 to recover a corresponding file; and 5, judging whether the task is completed, if the task is completed, ending the operation, and if the task is not completed, executing the step 4. By adopting the method, files of which the high level of the initial cluster number is cleared can be found and recovered, the research range can be reduced according to a parent directory or the high level of a file of a same class, the program operation time is shortened, files can be recovered through three levels of judgment, and the success rate is increased.

The invention provides a geographic positioning system and method based on image recognition cloud service; according to the method, a query area is delimited in a geographic information database by combining initial positioning with the information of an obtained instant scene image, and a retrieval database of the query area is established, so that a retrieval scope is narrowed, the retrieval database is smaller, the problem that the search speed is slow due to a huge search data volume can be solved, and the image retrieval efficiency and the image matching speed are effectively improved. By use of a computer vision image matching technology, geographic positioning can be accurately realized; by use of a location information measurement unit of a mobile terminal, a scene search range isnarrowed, the validity and the real-time performance of the geographic location information positioning can be greatly improved, and the system and method are popularized and applied in modern smartmobile terminals and wireless communication technology, and can be effectively implemented and popularized.

The invention discloses a webpage verification code element identification method, which comprises the steps of obtaining website page source code through a web crawler, andchecking if the page source code has a <form> tag; retrieving a characteristic keyword in the <form> tag; searching a tag <input> adjacent to the characteristic keyword, and determining if the tag "type" attribute is "text"; taking the <input> tag as a reference, and searching the tag recursively to upward and downward adjacent tags in order; retrieving onclick event in the tag, and extracting onclick event attribute value; coupling create_code () function content in javascript code, adding a string of digital random value to a tag src attribute value to submit in a parameter form; retrieving onclick event content, coupling javascript function content to generate a random value, a date at present, and a time stamp at present; and simulating and executing the onclick event, hereby determining if the img content is the picture identifying code. By search the tags adjacent to the characteristic keyword, the object retrieval scope is greatly reduced, and searching efficiency is improved.

The invention discloses a matching search method for a stereo image, which comprises the process of performing dynamic adjustment on the search length of matching points, wherein the process judges the trend of matching error of the matching points, and finishes the matching process of current reference points if the increase trend is showed. The method determines whether the matching process continues to be executed or not by judging the variation trend of the matching error, thereby realizing the dynamic adjustment of the search length. By narrowing the search range, the unnecessary matching operation is reduced as much as possible, which not only improves matching accuracy and matching efficiency, but also ensures accurate matching of shadowed parts.

The invention discloses a river pollutant tracing method, which comprises the following steps of: enabling tracing equipment to cruise in a monitored river, and when the tracing equipment monitors that a certain pollutant index reaches an early warning value, enabling the tracing equipment to sail between two banks in a river channel in a broken line-shaped tracing line; monitoring the maximum pollutant concentration value in each line segment in the traceability line, recording the coordinates of the maximum pollutant concentration value, and when the traceability equipment monitors that themaximum pollutant concentration in a certain line segment is lower than the maximum pollutant concentration in the previous line segment, enabling the traceability equipment to sail and monitor between the two maximum pollutant concentration points in a broken line manner, until the location of a pollution source is acquired. A pollution source is traced in a river water surface two-dimensional plane, and the tracing mode is reliable and efficient; the tracing method is simple and feasible; and the tracing equipment cost is low.

The invention relates to an image searching method and system based on motion acceleration and electronic equipment. The method comprises the following steps: step a, calculating the acceleration of ato-be-tracked target in a current frame image according to the displacements of the first two frames of images; b, determining a search range rectangular frame of a to-be-tracked target in the current frame image according to the acceleration calculation result; and step c, extracting a candidate box of a to-be-tracked target in the current frame image along the diagonal line of the search rangerectangular box through an RPN network, and carrying out feature analysis on the candidate box to obtain the position of the to-be-tracked target in the current frame image. According to the method, the limited search range rectangular frame is determined in an acceleration calculation mode, candidate frames of the tracking target are selected along the diagonal line of the search range rectangular frame, a smaller search range is determined, global search is not needed, the search range is greatly narrowed, the calculated amount is reduced, and the calculation speed is increased.

The invention provides a closed large-area intelligent monitoring system based on code recognition and a recognition method thereof. The system comprises an identification subsystem and a video subsystem. The video subsystem carries out authorization registration on people to be registered in a non-closed area. During registration, the unique number is sprayed to an authorized person through ultraviolet or infrared paint, after spraying, the number and a camera shoot calibration object data, the calibration object data are sent to an image index server to load the number and a registration photo, and storage and filing are carried out for checking during identification of the identification subsystem. And the identification subsystem is used for performing identification judgment, classification and processing according to the serial numbers and the pictures shot by the AI camera in the video subsystem in a closed large area in real time. According to the system, the invisible light ismarked with a unique number, high-speed irradiation and shooting are carried out, the number is accurately identified in a long distance, high-speed checking and identification are carried out, and an AI closed area intelligent monitoring system is adopted, so that the identification accuracy and efficiency are effectively improved.

The invention provides a quick face recognition method, apparatus and device, and a readable storage medium. The method comprises the steps of preprocessing and binarizing a to-be-recognized face image to obtain a binary image; performing right lower vertex detection on the binary image to obtain a right lower vertex set; synthesizing a synthesized face area corresponding to each right lower vertex according to a preset value, calculating to obtain an evaluation array according to face area extraction features, calculating a normal form evaluation value of each right lower vertex by adopting aface evaluation normal form, and selecting the right lower vertex with the maximum normal form evaluation value as an actual right lower vertex; and generating a target face area according to the actual lower right vertex and the face width and height, and intercepting a corresponding image from the to-be-recognized face image according to the target face area to perform face recognition. According to the invention, the human face is positioned by combining the right lower vertex detection and the human face evaluation normal form, the search range is reduced, and the influence of various interferences is reduced, so that the human face area can be quickly and accurately positioned for human face recognition.

The present invention relates to a target state prediction method in a dynamic environment. For different colors in the dynamic environment, a color space with a best identification effect is extracted, then through continuously updating a next frame center point and a next frame image threshold range, an effect of adapting to target detection is achieved, threshold range estimation can be effectively shortened, a target is identified in a maximum way, through the combination of HSV and YCbCr color space, an update center point is corrected and restored to improve the accuracy of an algorithm, then the prediction of a target is tracked according to a Kalman filter and a Markoff model, the interference of discrete color features and similar objects in a complex environment is overcome, a detector search range is reduced, the tracking precision in the partial occlusion of the target is improved, and the tracking accuracy is improved.

The invention discloses a method for searching data. The method comprises the following steps: receiving an image searching message, wherein the image searching message at least comprises an image to be searched and a geographic location message related to the image to be searched; searching the Internet for a plurality of candidate data to constitute a candidate database according to the received geographic location message, wherein each candidate datum comprises at least one candidate image; and then, comparing the image to be searched with the candidate image in the candidate database to determine a search result, wherein the search result at least comprises one of the candidate data.

The invention provides a power equipment retrieval method based on deep learning, and relates to the technical field of image processing. The invention provides a power equipment retrieval method based on deep learning. The power equipment retrieval method comprises the steps of detecting target equipment, intercepting an equipment area picture, extracting equipment characteristics, calculating cosine characteristic similarity and determining a final retrieval result. According to the invention, a target detection method in deep learning is combined; according to the method, the device positioning in the device picture and the accurate cutting of the device area are realized, the noise interference of the background picture is reduced, the feature extraction is carried out by adopting thedeep convolutional neural network, the retrieval precision is improved, the retrieval range in the database is reduced through the target detection result, and the retrieval efficiency is improved. Inaddition, the calculation of the local cosine feature similarity of the equipment can more accurately distinguish the equipment type with extremely high similarity, so that the retrieval precision isfurther improved.

The embodiment of the invention discloses a network rate segment processing method and apparatus. The method comprises the following steps: selecting a target interface in a network as a demarcation point of a first segment of link and a second segment of link in an end-to-end communication process, wherein the first segment of link is a link from a transmitting end to the target interface, and the second segment of link is a link from the target interface to a receiving end; collecting interaction data passing through the target interface, the transmitting end and the receiving end, and the time of the interaction data to arrive at the target interface; obtaining the delays of the first segment of link and the second segment of link according to the interaction data and the time of the interaction data to arrive at the target interface; and obtaining the rates of the first segment of link and the second segment of link according to the delays. According to the network rate segment processing method and apparatus disclosed by the embodiment of the invention, the data of a single network interface in the network are collected to form a data set to serve as an original data source ofstatistical analysis, and compared with the prior art, the network rate segment processing method and apparatus have the advantages of avoiding the problem that integrated calculation cannot be achieved due to unsynchronized data times of different interfaces, and accurately reducing the network problem search range.

The invention discloses a robot repositioning method and apparatus, and a storage device. The method comprises: determining whether a current frame image is matched with a key frame image stored within a first search range; if NO, determining whether the current frame image comprises distance information of a robot, wherein the distance information is provided by an ultra wideband base station onthe robot; if YES, acquiring a second search range of the key frame image in combination with the distance information; determining whether the current frame image is provided with a matched key framewithin the second search range; and if matching succeeds, acquiring a current position and posture of a camera according to the matched key frame image. Through the above mode, the robot repositioning method can narrow the search range of the key frame image and reduce a search range of a subsequent bag-of-word, thereby shortening the search time and improving the subsequent matching efficiency.

The invention discloses a hybrid positioning algorithm based on an LTE and Wi-Fi heterogeneous network. The hybrid positioning algorithm comprises the steps that the CELLID of the cell in which a mobile station is located and the signal intensity indication of a nearby wireless access point are acquired to acquire the position range of the mobile station; a measurement value is acquired from the network side; based on a single base station multi-signal hybrid positioning algorithm geometric model, the minimum area of particle search is acquired; an objective function is constructed; an improved PSO algorithm is used to solve the objective function; and finally, an improved PSO algorithm based on the fitness dynamic adjustment inertia weight is used to solve the objective function. According to the invention, the single base station multi-signal positioning geometric model suitable for an NLOS environment is established; the improved PSO algorithm based on the fitness dynamic adjustment inertia weight is provided on the positioning phase; the activity of each particle is dynamically adjusted; the global search ability and the convergence ability of the PSO algorithm are improve; and the position of the mobile station is accurately acquired.

The invention provides a retrieval method and device for looking up contacts in an address book in multiple conditions. The method comprises an input step of inputting keywords and separators, wherein at least two keywords are input, and a plurality of keywords are connected by the separators; a multi-condition processing step of forming a logic relation of the plurality of keywords according to the keywords and the separators; a retrieval step of retrieving the contacts in the address book according to the logic relation and obtaining a target contact meeting the retrieval condition; and a display step of displaying the target contact according to a retrieval result of a retrieval module. The device is a device applying the method. By adopting the method and the device provided by the invention, a user inputs the plurality of keywords and the plurality of separators to carry out multiple limitation on the contact to be found, so that in the case the user lacks a clear impression on the target contact and/or the recorded information is incomplete, a large amount of information related to the target contact can also be provided for the user, so as to shrink the lookup range and screen the target contact.

The invention relates to the Chinese character input technology and provides a handwritten simplified Chinese character input method and system to overcome the defects that in the existing handwriting process, complete character patterns and recognition deviation need to be input, and the input speed is low. The method comprises the steps of receiving handwriting track information generated when handwriting is conducted on a handwriting device, converting the handwriting track information into simplified Chinese characters, accurately retrieving absolutely-correlated Chinese characters according to matching simplified Chinese characters in the system with the matching simplified Chinese characters as retrieving units, and achieving input of the current Chinese characters with reduced strokes and handwritten outlines and the Chinese characters predicted from the current Chinese characters. By means of the handwritten simplified Chinese character input method and system, the Chinese characters with a small number of strokes are input accurately, and the Chinese characters with a large number of strokes can be input accurately just through a part of the smokes, so that the input speed of handwritten Chinese characters is increased efficiently, the input accuracy of the handwritten Chinese characters is improved efficiently, and great significance is achieved.

The invention relates to an e-commerce trial evaluation and recommendation system based on user requirements. The system is characterized by being composed of a physical store trial terminal, a streetdisplay trial terminal, a marketing management center and a mobile phone terminal, and the physical store trial terminal and the street display trial terminal are respectively composed of a trial area, a keyboard input and display module, a gift area, a data processing and storage module and a wireless transmission module; the marketing management center is composed of a data preprocessor, a datamining PDM server, an FIFO sequence memory, a data exchange server, a marketing management server and a permanent memory. And the marketing management server is connected with a mobile phone terminalof a customer who participates in trial through a mobile communication network. The system solves the problem that the existing e-commerce enterprise has few means of attracting consumers and miningnew customers and is poor in efficiency, and compared with the existing telephone marketing means, the system takes the user requirements as a fundamental starting point, avoids the blindness, improves the return rate, and saves the manpower and material resources.

A method and apparatus for searching for cache data is disclosed. The method includes receiving a target data read request including a target logical block address of the target data. A logical blockaddress derive from a target logical block address and a logical block address and a physical block address corresponding to that logical block address and the physical block address storing the respective cache data. A physical block address mapping table for obtaining a target physical block address of the target data; According to the maximum physical block address and the minimum physical block address of the physical block address of the cache data included in each cache data set grouped based on the physical block address size of the respective cache data, If a target cache data set corresponding to a target physical block address is found, target data is searched in the target cache data set. That is, only when the physical block address of the data to be read corresponds to a set of Cache data can the search be performed in the set, thereby narrowing the search range and improving the search efficiency and the read performance.