WO2020134839A1 - Procédé et appareil de recherche d'image - Google Patents

Procédé et appareil de recherche d'image Download PDF

Info

Publication number
WO2020134839A1
WO2020134839A1 PCT/CN2019/121607 CN2019121607W WO2020134839A1 WO 2020134839 A1 WO2020134839 A1 WO 2020134839A1 CN 2019121607 W CN2019121607 W CN 2019121607W WO 2020134839 A1 WO2020134839 A1 WO 2020134839A1
Authority
WO
WIPO (PCT)
Prior art keywords
identifier
image
identification
data
structured data
Prior art date
Application number
PCT/CN2019/121607
Other languages
English (en)
Chinese (zh)
Inventor
刘国伟
Original Assignee
深圳云天励飞技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳云天励飞技术有限公司 filed Critical 深圳云天励飞技术有限公司
Publication of WO2020134839A1 publication Critical patent/WO2020134839A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/51—Indexing; Data structures therefor; Storage structures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content

Definitions

  • the present invention relates to the Internet field, and in particular, to an image search method and device.
  • Embodiments of the present invention provide an image search method and apparatus.
  • structured data and unstructured data of an image are stored separately. Similarity calculation of structured data and screening of structured data can quickly determine the result image, while improving query efficiency and saving storage space.
  • the first aspect of the present invention discloses an image search method, the method includes:
  • the screening conditions screening the first structured data according to the screening conditions to obtain second structured data;
  • the second structured data includes a third identifier, and the third identifier is part or all of the second identifier;
  • the obtaining the result image according to the third identifier and the preset model includes:
  • the fourth logo is input into the preset model to obtain a result image.
  • the method before receiving the user's query request, the method further includes:
  • the storing of the unstructured data of the batch of image data in the preset model includes:
  • the determining the second identifier corresponding to the first identifier includes:
  • mapping table stores the first identification and the second identification Mapping relationship
  • the first identifier is an identifier of an image stored in a preset model
  • the second identifier is a sequence identifier of the image in the target server.
  • the second aspect of the present invention discloses an image search device, the device includes:
  • a receiving unit configured to receive a query request from a user, where the query request includes a target image and a screening condition
  • an acquisition unit configured to acquire the feature value of the target image
  • an input unit for inputting the feature value of the target image into a preset model to obtain a similarity ranking
  • the first identification of the first N data where, N is a positive integer
  • a determining unit configured to determine a second identifier corresponding to the first identifier
  • the input unit is configured to input the second identifier into the target server to obtain the first structured data
  • a screening unit for screening the first structured data according to the screening conditions to obtain second structured data; wherein the second structured data includes a third identifier, and the third identifier Part or all of the second logo;
  • an acquisition unit configured to acquire a result image according to the third identification and the preset model
  • a return unit configured to return the result image to the user.
  • the obtaining unit is specifically configured to determine a fourth identifier corresponding to the third identifier, where the fourth identifier is part or all of the first identifier; the fourth identifier Input into the preset model to obtain the resulting image.
  • the device further includes a first storage unit and a second storage unit;
  • the acquiring unit is further configured to acquire the structured data and unstructured data of the batch of images when receiving a request to increase the batch of images;
  • the first storage unit is configured to store the structured data of the batch of images in the target server
  • the second storage unit is configured to store the unstructured data of the batch of images in the preset model.
  • the second storage unit is configured to generate based on the unstructured data of the batch of images
  • the determining unit is specifically configured to match the first identifier with a pre-stored mapping table to obtain the second identifier corresponding to the first identifier; wherein, the mapping table The table stores the mapping relationship between the first identifier and the second identifier, the first identifier is an identifier of an image stored in a preset model, and the second identifier is a sequence identifier of the image in the target server .
  • a query request of a user is received, wherein the query request includes a target image and a filtering condition; acquiring the feature value of the target image, and combining the target image
  • the characteristic value of is input into the preset model to obtain the first identification of the top N data of similarity; where, N is a positive integer; determine the second identifier corresponding to the first identifier; enter the second identifier into the target server to obtain the first structured data; according to the filtering conditions, the first structured data Performing screening to obtain second structured data; wherein the second structured data includes a third identifier, and the third identifier is part or all of the second identifier; based on the third identifier and the preset model Acquire a result image, and return the result image to the user.
  • the structured data and the unstructured data of the image are separately stored, and when performing image search, the similarity calculation of the unstructured data and the structured data
  • the filtering can quickly determine the result image, while improving query efficiency, saving storage space.
  • FIG. 1 is a schematic diagram of an image search method according to an embodiment of the present invention.
  • FIG. 2 is a schematic diagram of another image search method according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of another image search method according to an embodiment of the present invention.
  • FIG. 4 is a logical structure diagram of an image search apparatus according to an embodiment of the present invention.
  • FIG. 5 is a logical structure diagram of another image search apparatus according to an embodiment of the present invention.
  • FIG. 6 is a logical structure diagram of another image search apparatus according to an embodiment of the present invention.
  • FIG. 7 is a schematic diagram of a physical structure of an image search apparatus according to an embodiment of the present invention.
  • FIG. 1 is a schematic flowchart of an image search method according to an embodiment of the present invention.
  • an image search method provided by an embodiment of the present invention includes the following content:
  • the executive body of the present invention may be a server, the server has a search engine function.
  • the server can include various handheld devices with search engine functions, in-vehicle devices, wearable devices, computing devices, or other processing devices, as well as various forms of user devices (User
  • terminal device Terminal device
  • the operating system involved in the embodiments of the present application is a software system that performs unified management of hardware resources and provides service interfaces to users.
  • the user's query request may be a picture query request, or a text or voice query request.
  • the target image is a face image
  • the filtering conditions may be parameters such as age, gender, whether to bring eyes, and time for storing the picture.
  • the method includes:
  • storing the unstructured data of the batch of images into the preset model includes: according to the batch The unstructured data of the image generates a target file; use the preset model to load the target file. It is understandable that the batch of image unstructured data is first generated into a target file for the convenience of subsequent loading, so that unstructured data is not loaded one by one, thereby improving processing efficiency.
  • the target server may be an independent search application server, and the user may submit an XML (Extensible Markup Language) file in a certain format to the search application server through an HTTP request to generate an index; or through Http Get ( (Hypertext Transfer Protocol Acquisition) operation makes a search request and obtains the return result in XML format.
  • XML Extensible Markup Language
  • Http Get Hypertext Transfer Protocol Acquisition
  • the preset model is a framework to provide efficient similarity search and clustering for dense vectors. The preset model can provide multiple retrievals and is faster.
  • the structured data and the unstructured data of the images are added to different lists, where the first list stores the structured data and the second list stores the unstructured data.
  • the structured data includes parameters such as the gender, age, and storage time of the person in the image.
  • Unstructured data is feature value data of face images.
  • hdf5 has only two basic structures: groups and datasets.
  • the group contains 0 or more data set files.
  • the data volume is generated according to 1 million secretaries to generate a file for data storage.
  • the generated hdf5 file is loaded using a preset model, which generates the center point according to the process of two K-means (clustering algorithm) cluster analysis, and generates a product quantization code for each record, and sorts based on the center point .
  • the first k-means is to generate a central point of 1024 clustering calculations for the data, relative to dividing the data into 1024 copies.
  • the second K-means is to perform a clustering operation on 1024 data again. It can be understood that in the second K-means, each image feature value data is divided into 4 segments, and 4 clustering points are generated during K-mea ns, each image feature value will generate 4 character strings pg code (ie a four-dimensional vector).
  • the image identifier is stored in hdf5 for use in search.
  • the logo of the image is marked as the first logo.
  • the target server will add a docu ment id (file ID) for each record.
  • the id is marked as the second identification. Understandably, the document id can be a serial number.
  • [0065] 102 Obtain the feature value of the target image, and input the feature value of the target image into a preset model to obtain the first identifier of the top N data of similarity ranking; wherein, N is a positive integer;
  • the center point is quickly searched according to two K-means (the two searches are performed in order, that is, the clustering is performed on the result of the first clustering Class, and then perform a quick search based on the results of the secondary clustering) to obtain the top N data with the highest similarity, which can improve query efficiency.
  • the determination of the second identifier corresponding to the first identifier includes:
  • mapping table stores the first identification and the second identification
  • the first identifier is an identifier of an image stored in a preset model
  • the second identifier is a sequence identifier of the image in the target server.
  • the second identifier is input into the target server in order to obtain the structured data corresponding to the second identifier, that is, the first structured data.
  • the first structured data is 100
  • the screening condition is male
  • the structured data that meets the condition that is, the second structured data
  • the identification of the 50 pieces of structured data is the third identification (understandably, the third identification is a set of identifications, which contains 50 identifications).
  • the third identification is the identification recorded in the target server (the target server edits the serial number for each piece of data), and also needs to be converted into an image identification to obtain the image identification corresponding to the image identification from the preset model Images (the serial number corresponding to each image is stored in the holi model).
  • the obtaining the result image according to the third identifier and the preset model includes:
  • each image in the model has its own serial number
  • an identifier will be set for each data ( serial number) .
  • the first identifier that is, the serial number of the image in the model
  • the data with the numbers 10 and 20 meet the filtering conditions (that is, 10 and 20 are the third identifiers). Since 10 and 20 are the target servers Structured data identification in the model, it is necessary to obtain the identification of the image in the model. According to the mapping relationship, the identification of the image in the model is 1 and 2 (that is, the fourth identification), then it is OK to obtain the pair of hard images of identification 1 and 2 in the model .
  • a query request of a user is received, wherein the query request includes a target image and a filtering condition; acquiring the feature value of the target image, and combining the target image
  • the characteristic value of is input into the preset model to obtain the first identifier of the top N data of similarity; determine the second identifier corresponding to the first identifier; enter the second identifier into the target server to obtain the first identifier A structured data; filtering the first structured data according to the screening conditions to obtain second structured data; wherein the second structured data includes a third identifier, based on the third identifier and all
  • the preset model acquires a result image, and returns the result image to the user.
  • the structured data and unstructured data of the image are stored separately.
  • the similarity calculation of the unstructured data and the screening of the structured data can quickly determine the resulting image. While improving query efficiency, it saves storage space.
  • FIG. 2 is a schematic flowchart of another image search method according to another embodiment of the present invention.
  • the method includes:
  • storing the batch of unstructured data of image data in the preset model includes: generating a target file according to the batch of unstructured image data; using the preset The model loads the target file.
  • the determining of the second identifier corresponding to the first identifier includes: matching the first identifier with a pre-stored mapping table to obtain the second identifier corresponding to the first identifier
  • the mapping table stores the mapping relationship between the first identifier and the second identifier
  • the first identifier is the identifier of the image stored in the preset model
  • the second identifier is the image in the Describe the sequence identifier in the target server.
  • the acquiring the result image according to the third identifier and the preset model includes:
  • the structured data (stored by the target server) and the unstructured data (stored by the preset model) of the images added in batches are stored separately, and the preset model is used to store the
  • the structured data can improve the matching data of the image feature data, and storing the structured data through the target server can increase the screening speed and save the storage space.
  • Technical solutions provided by using embodiments of the present invention To further ensure the user's query efficiency.
  • FIG. 3 a schematic flowchart of another image search method according to another embodiment of the present invention. Wherein, as shown in FIG. 3, the method includes:
  • [0100] 304 Establish a mapping relationship between the first identifier and the second identifier.
  • the determining the second identifier corresponding to the first identifier includes: matching the first identifier with a pre-stored mapping table to obtain the second identifier corresponding to the first identifier
  • the mapping table stores the mapping relationship between the first identifier and the second identifier
  • the first identifier is the identifier of the image stored in the preset model
  • the second identifier is the image in the Describe the sequence identifier in the target server.
  • a data processing apparatus 400 provided by an embodiment of the present invention, wherein the apparatus 400 includes the following units:
  • the receiving unit 401 is configured to receive a query request from a user, where the query request includes a target image and filtering conditions;
  • an obtaining unit 402 configured to obtain the feature value of the target image
  • the input unit 403 is configured to input the feature value of the target image into a preset model to obtain the first identifier of the top N data of similarity; where N is a positive integer;
  • a determining unit 404 configured to determine a second identifier corresponding to the first identifier
  • the input unit 403 is configured to input the second identifier into the target server to obtain the first structured data
  • a screening unit 405 configured to screen the first structured data according to the screening conditions to obtain second structured data; wherein the second structured data includes a third identifier, the third The mark is part or all of the second mark;
  • an obtaining unit 406 configured to obtain a result image according to the third identifier and the preset model
  • a returning unit 407 configured to return the result image to the user.
  • the obtaining unit 406 is specifically configured to determine a fourth identifier corresponding to the third identifier, where
  • the fourth mark is part or all of the first mark; the fourth mark is input into the preset model to obtain a result image.
  • the device 400 further includes a first storage unit 408 and a second storage unit 409;
  • the obtaining unit 406 is further configured to obtain structured data and unstructured data of the batch of images when receiving a request to increase the batch of images; [0124]
  • the first storage unit 408 is used to store the batch of structured data of the image to the target server;
  • the second storage unit 409 is configured to store the unstructured data of the batch of images in the preset model.
  • the second storage unit 409 is configured to generate a target file based on the unstructured data of the batch of images; use the preset model to load the target file.
  • the determining unit 404 is specifically configured to match the first identifier with a pre-stored mapping table to obtain the second identifier corresponding to the first identifier; wherein, the mapping table The mapping relationship between the first identifier and the second identifier is stored, the first identifier is an identifier of an image stored in a preset model, and the second identifier is a sequence identifier of the image in the target server.
  • the above units 401-409 can be used to perform the methods described in steps 101-106 in Embodiment 1.
  • the above units 401-409 can be used to perform the methods described in steps 101-106 in Embodiment 1.
  • a data processing apparatus 500 provided by an embodiment of the present invention, where the apparatus 500 includes the following units:
  • the obtaining unit 501 is configured to obtain structured data and unstructured data of the batch of images when receiving a request to add images in batches;
  • a storage unit 502 configured to store the batch of structured image data to the target server
  • storing the batch of unstructured data of the image data in the preset model includes: generating a target file according to the batch of image unstructured data; using the preset The model loads the target file.
  • the receiving unit 503 is configured to receive a query request from a user, where the query request includes a target image and a filtering condition;
  • an obtaining unit 504 configured to obtain the feature value of the target image
  • the input unit 505 is configured to input the feature value of the target image into a preset model to obtain the first identification of the top N data of similarity; where N is a positive integer;
  • a determining unit 506 configured to determine a second identifier corresponding to the first identifier
  • the determining of the second identifier corresponding to the first identifier includes: Matching the stored mapping table to obtain the second identification corresponding to the first identification; wherein, the mapping table stores the mapping relationship between the first identification and the second identification, and the first identification is An identifier of an image stored in a preset model, the second identifier is a sequence identifier of the image in the target server.
  • the input unit 505 is configured to input the second identifier into the target server to obtain the first structured data
  • a screening unit 507 configured to screen the first structured data according to the screening conditions to obtain second structured data; wherein the second structured data includes a third identifier, the third The mark is part or all of the second mark;
  • the obtaining unit 501 is configured to obtain a result image according to the third identifier and the preset model, and return the result image to the user.
  • the above units 501-507 may be used to execute the method described in steps 201-208 in Embodiment 2.
  • the description of the method in Embodiment 2 please refer to the description of the method in Embodiment 2, which will not be repeated here.
  • a data processing apparatus 600 provided by an embodiment of the present invention, wherein the apparatus 600 includes the following units:
  • an obtaining unit 601 configured to obtain structured data and unstructured data of the batch of images when receiving a request to add images in batches;
  • storage unit 602 for storing the batch of structured data of the image to the target server
  • an obtaining unit 601 configured to obtain the first identifier of the structured data edited by the target service
  • the generating unit 603 is configured to generate a target file based on the unstructured data of the batch of images; load the target file using the preset model, and obtain a second identification of the image;
  • the establishing unit 604 is configured to establish a mapping relationship between the first identifier and the second identifier.
  • the receiving unit 605 is configured to receive a query request from a user, where the query request includes a target image and a filtering condition;
  • the obtaining unit 601 is further configured to obtain the feature value of the target image
  • the input unit 606 is configured to input the feature value of the target image into a preset model to obtain the first identification of the top N data of similarity ranking; where N is a positive integer; [0151] a determining unit 607, configured to determine a second identifier corresponding to the first identifier, and input the second identifier into a target server to obtain first structured data;
  • the determining the second identifier corresponding to the first identifier includes: matching the first identifier with a pre-stored mapping table to obtain the second identifier corresponding to the first identifier
  • the mapping table stores the mapping relationship between the first identifier and the second identifier
  • the first identifier is the identifier of the image stored in the preset model
  • the second identifier is the image in the Describe the sequence identifier in the target server.
  • a screening unit 608 configured to screen the first structured data according to the screening conditions to obtain second structured data; wherein the second structured data includes a third identifier, the third The mark is part or all of the second mark;
  • an obtaining unit 601 configured to obtain a result image according to the third identifier and the preset model, and return the result image to the user.
  • the above units 601-608 may be used to execute the method described in steps 301-309 in Embodiment 2.
  • the above units 601-608 may be used to execute the method described in steps 301-309 in Embodiment 2.
  • the above units 601-608 may be used to execute the method described in steps 301-309 in Embodiment 2.
  • a data processing apparatus 700 is provided.
  • the device 70 includes hardware such as a CPU 701, a memory 702, a bus 703, and a transceiver 704.
  • the above logic units shown in FIGS. 4-6 can be implemented by the hardware device shown in FIG. 7.
  • the CPU 701 executes the server program stored in the memory 702 in advance, and the execution process specifically includes:
  • the obtaining the result image according to the third identifier and the preset model includes:
  • the fourth identification is input into the preset model to obtain a result image.
  • the execution process before receiving the user's query request, the execution process further includes:
  • the storing of the unstructured data of the batch of image data in the preset model includes:
  • the determining the second identifier corresponding to the first identifier includes:
  • mapping table stores the first identification and the second identification Mapping relationship
  • the first identifier is an identifier of an image stored in a preset model
  • the second identifier is a sequence identifier of the image in the target server.
  • a query request from a user is received, wherein the query request includes a target image and a filtering condition; acquiring the feature value of the target image, and combining the target
  • the feature value of the image is input into the preset model to obtain the first identifier of the top N data of similarity; the second identifier corresponding to the first identifier is determined; the second identifier is input into the target server to obtain First structured data; screening the first structured data according to the screening conditions to obtain second structured data; wherein the second structured data includes a third identifier, and according to the third identifier and
  • the preset model acquires a result image, and returns the result image to the user.
  • the structured data and the unstructured data of the image are separately stored, and when the image search is performed, the similarity calculation of the unstructured data and the screening of the structured data can be used Quickly determine the result image, while improving query efficiency, saving storage space.
  • a computer program product in another embodiment of the present invention, contains program code; when the program code is executed, the method in the foregoing method embodiment will be carried out.
  • a chip in another embodiment of the present invention, is disclosed, and the chip includes program code; when the program code is executed, the method in the foregoing method embodiment will be executed.
  • the disclosed device may be implemented in other manners.
  • the device embodiments described above are only schematic.
  • the division of the unit is only a logical function division.
  • there may be another division manner for example, multiple units or components may be combined or may Integration into another system, or some features can be ignored, or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be electrical or other forms.
  • the unit described as a separate component may or may not be physically separated, and the component displayed as the unit may or may not be a physical unit, that is, may be located in one place, or may be distributed to multiple networks On the unit. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.
  • the above integrated unit may be implemented in the form of hardware, or in the form of a software functional unit.
  • the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium.
  • the technical solution of the present invention essentially or part of the contribution to the existing technology or all or part of the technical solution can be embodied in the form of a software product, the computer software product is stored in a storage medium , Including several instructions to enable a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention.
  • the aforementioned storage media include: U disk, read-only memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), removable hard disk, magnetic disk or CD-ROM and other media that can store program code.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Library & Information Science (AREA)
  • Software Systems (AREA)
  • Processing Or Creating Images (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

L'invention concerne un procédé et un appareil de recherche d'image. Le procédé consiste à : recevoir une demande d'interrogation d'un utilisateur, la demande d'interrogation comprenant une image cible et une condition de filtrage (101) ; acquérir une valeur caractéristique de l'image cible et entrer la valeur caractéristique de l'image cible dans un modèle prédéfini de façon à acquérir un premier identifiant de données avec la similarité classée dans les N premières (102) ; déterminer un deuxième identifiant correspondant au premier identifiant (103) ; entrer le deuxième identifiant dans un serveur cible de façon à acquérir des premières données structurées (104) ; filtrer les premières données structurées selon la condition de filtrage de façon à acquérir des secondes données structurées (105), les secondes données structurées comprenant un troisième identifiant ; et acquérir une image de résultat selon le troisième identifiant et le modèle prédéfini et renvoyer l'image de résultat à l'utilisateur (106). Au moyen du procédé, l'efficacité d'interrogation peut être améliorée et l'espace de stockage peut également être préservé.
PCT/CN2019/121607 2018-12-29 2019-11-28 Procédé et appareil de recherche d'image WO2020134839A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811640366.XA CN109829073B (zh) 2018-12-29 2018-12-29 一种图像搜索的方法及装置
CN201811640366.X 2018-12-29

Publications (1)

Publication Number Publication Date
WO2020134839A1 true WO2020134839A1 (fr) 2020-07-02

Family

ID=66861411

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/121607 WO2020134839A1 (fr) 2018-12-29 2019-11-28 Procédé et appareil de recherche d'image

Country Status (2)

Country Link
CN (1) CN109829073B (fr)
WO (1) WO2020134839A1 (fr)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109829073B (zh) * 2018-12-29 2020-11-24 深圳云天励飞技术有限公司 一种图像搜索的方法及装置
CN110287346B (zh) * 2019-06-28 2021-11-30 深圳云天励飞技术有限公司 数据存储方法、装置、服务器及存储介质
CN111061888B (zh) * 2019-11-20 2023-05-16 北京明略软件系统有限公司 图像获取的方法及系统
CN111966629A (zh) * 2020-06-28 2020-11-20 电子科技大学 粒子模拟数据存储方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120209833A1 (en) * 2011-02-11 2012-08-16 Siemens Aktiengesellschaft Methods and devices for data retrieval
CN106503121A (zh) * 2016-10-19 2017-03-15 公安部第三研究所 一种x光安检图像的结构化描述方法及系统
CN107346435A (zh) * 2017-06-15 2017-11-14 浙江捷尚视觉科技股份有限公司 一种基于车辆特征库的嫌疑套牌车捕获方法
CN107729502A (zh) * 2017-10-18 2018-02-23 公安部第三研究所 一种卡口车辆个性化特征智能检索系统及方法
CN107944017A (zh) * 2017-12-11 2018-04-20 浙江捷尚视觉科技股份有限公司 一种视频中非机动车的检索方法
CN109829073A (zh) * 2018-12-29 2019-05-31 深圳云天励飞技术有限公司 一种图像搜索的方法及装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102523304B (zh) * 2011-12-29 2014-10-22 北京新媒传信科技有限公司 一种应用云平台和云平台的实现方法
CN103116643A (zh) * 2013-02-25 2013-05-22 江苏物联网研究发展中心 基于Hadoop的智慧医疗数据管理方法
CN103425780B (zh) * 2013-08-19 2016-08-17 曙光信息产业股份有限公司 一种数据的查询方法和装置
CN106886553B (zh) * 2016-12-27 2020-07-28 浙江宇视科技有限公司 一种图像检索方法及服务器

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120209833A1 (en) * 2011-02-11 2012-08-16 Siemens Aktiengesellschaft Methods and devices for data retrieval
CN106503121A (zh) * 2016-10-19 2017-03-15 公安部第三研究所 一种x光安检图像的结构化描述方法及系统
CN107346435A (zh) * 2017-06-15 2017-11-14 浙江捷尚视觉科技股份有限公司 一种基于车辆特征库的嫌疑套牌车捕获方法
CN107729502A (zh) * 2017-10-18 2018-02-23 公安部第三研究所 一种卡口车辆个性化特征智能检索系统及方法
CN107944017A (zh) * 2017-12-11 2018-04-20 浙江捷尚视觉科技股份有限公司 一种视频中非机动车的检索方法
CN109829073A (zh) * 2018-12-29 2019-05-31 深圳云天励飞技术有限公司 一种图像搜索的方法及装置

Also Published As

Publication number Publication date
CN109829073A (zh) 2019-05-31
CN109829073B (zh) 2020-11-24

Similar Documents

Publication Publication Date Title
WO2020134839A1 (fr) Procédé et appareil de recherche d'image
CN109101620B (zh) 相似度计算方法、聚类方法、装置、存储介质及电子设备
CN109617762B (zh) 一种利用网络流量识别移动应用的方法
WO2016201933A1 (fr) Procédé de traitement de données d'utilisateur, procédé de fourniture, système et dispositif informatique
WO2018010491A1 (fr) Procédé et appareil permettant de générer un trajet de service
WO2020093289A1 (fr) Procédé et appareil de recommandation de ressource, dispositif électronique et support d'informations
CN111061750A (zh) 一种查询处理方法、装置及计算机可读存储介质
US10467307B1 (en) Grouping of item data using seed expansion
WO2015081792A1 (fr) Procédé, dispositif et système de recherche étendue corrélative et personnalisée
WO2017143930A1 (fr) Procédé de tri de résultats de recherche, et dispositif associé
US20150120708A1 (en) Information aggregation, classification and display method and system
JP6664599B2 (ja) 曖昧性評価装置、曖昧性評価方法、及び曖昧性評価プログラム
WO2022068543A1 (fr) Procédé et appareil de publication de contenu multimédia, ainsi que dispositif électronique et support de stockage
KR102197739B1 (ko) 연기자 정보 제공 방법
JP2018518764A (ja) オブジェクト検索方法、装置およびサーバ
US20160342860A1 (en) Method and system for searching images
JP2019053519A (ja) 分類装置、分類方法及び分類プログラム
US20150169727A1 (en) Information processing apparatus, information processing method, and system
CN112364051B (zh) 一种数据查询方法及装置
CN114490923A (zh) 相似文本匹配模型的训练方法、装置、设备及存储介质
US11188582B2 (en) Method for resource search and related products
JP5622880B2 (ja) アイテム推薦システム、アイテム推薦方法およびアイテム推薦プログラム
WO2023206960A1 (fr) Procédé et appareil de recommandation de produit reposant sur un filtrage basé sur le contenu et un filtrage collaboratif, et dispositif informatique
WO2017097102A1 (fr) Procédé de récupération et dispositif de récupération
US10394913B1 (en) Distributed grouping of large-scale data sets

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19902418

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19902418

Country of ref document: EP

Kind code of ref document: A1