WO2015172721A1 - Procédé et dispositif permettant de rechercher et de classer des images et de fournir une recherche d'images - Google Patents

Procédé et dispositif permettant de rechercher et de classer des images et de fournir une recherche d'images Download PDF

Info

Publication number
WO2015172721A1
WO2015172721A1 PCT/CN2015/078881 CN2015078881W WO2015172721A1 WO 2015172721 A1 WO2015172721 A1 WO 2015172721A1 CN 2015078881 W CN2015078881 W CN 2015078881W WO 2015172721 A1 WO2015172721 A1 WO 2015172721A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
images
module
family
webpage
Prior art date
Application number
PCT/CN2015/078881
Other languages
English (en)
Chinese (zh)
Inventor
陶哲
Original Assignee
北京奇虎科技有限公司
奇智软件（北京）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
Priority claimed from CN201410203342.3A external-priority patent/CN103995856B/zh
Priority claimed from CN201410203700.0A external-priority patent/CN103995857A/zh
Application filed by 北京奇虎科技有限公司, 奇智软件（北京）有限公司 filed Critical 北京奇虎科技有限公司
Publication of WO2015172721A1 publication Critical patent/WO2015172721A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor

Definitions

  • the present invention relates to the technical field of image data processing, and in particular, to a method and apparatus for implementing image search sorting, and a method and apparatus for providing image search.
  • the search engine is a software system applied on the network, which can be certain.
  • the method realizes the search and discovery of information on the network, and displays the search result after processing the searched information.
  • search results that can be provided to users are no longer just the search results of text information searched according to user input commands, but also can search for network images according to user needs, and will search out The results of the picture are presented to the user.
  • the present invention has been made in order to provide a method for realizing image search sorting and a corresponding apparatus for realizing image search sorting that overcomes the above problems or at least partially solves the above problems.
  • the present invention also provides a method of providing image search and a corresponding apparatus for providing image search.
  • a method for implementing image search ordering including: creating a plurality of image families corresponding to a plurality of source images; calculating a reference weight value of each image family; and using reference weights of each image family The size of the search results as a parameter for the search results of the search query feedback.
  • a method for providing an image search includes: after receiving an image query request, screening a plurality of image families associated with the query request; searching for each image family in the image family Corresponding source image and reference weight value of each image family; according to the order of the reference weights, the source images in each image family are sorted to draw search results corresponding to the query request.
  • an apparatus for implementing image search ordering comprising: a creating unit adapted to create a family of images corresponding to a plurality of source images; and a calculating unit adapted to calculate a reference weight of each image family And a sorting unit adapted to be a parameter for sorting the search results fed back by the search query according to the size of the reference weight of each image family.
  • an apparatus for providing an image search comprising: a screening unit adapted to After receiving the image query request, filtering a plurality of image families related to the query request; the searching unit is configured to search for a source image corresponding to each image family in the image family filtered by the screening unit, and each image family a reference weight; a rendering unit, adapted to receive the search result of the search unit, and sort the source images in each image family according to the order of the reference weights to draw a search result corresponding to the query request.
  • a program comprising readable code, when said readable code is run on a computing device, causing said computing device to perform image search ordering as described in an embodiment of the present invention method.
  • a program comprising readable code, when the readable code is run on a computing device, causing the computing device to perform the method of providing image search as described in an embodiment of the present invention
  • the embodiment of the present invention creates a plurality of image groups corresponding to the source image, and calculates a reference weight value of each image family, and then uses the size of the reference weight of each image family as a parameter for sorting the search results of the search query feedback. It can obtain more accurate and accurate sorting results, and make the image sorting result have priority order in the number of citations, which greatly improves the accuracy of the search and effectively improves the search efficiency.
  • the embodiment of the present invention creates a plurality of image groups corresponding to the source image, and calculates a reference weight value of each image family, and then uses the size of the reference weight of each image family as a parameter for sorting the search results of the search query feedback. It can obtain more accurate and accurate sorting results, and make the image sorting result have priority order in the number of citations, which greatly improves the accuracy of the search and effectively improves the search efficiency.
  • FIG. 1 is a flow chart showing the steps of a method for implementing image search ordering according to an embodiment of the present invention
  • FIG. 2 is a flow chart showing another method step of implementing image search sorting according to an embodiment of the present invention
  • FIG. 3 is a flow chart showing the steps of a method for providing image search according to an embodiment of the present invention.
  • FIG. 4 illustrates a flow chart of another method of providing image search in accordance with one embodiment of the present invention
  • FIG. 5 is a block diagram showing the structure of an apparatus for performing image search sorting according to an embodiment of the present invention.
  • FIG. 6 is a block diagram showing another structure of an apparatus for performing image search sorting according to an embodiment of the present invention.
  • FIG. 7 is a structural block diagram of another acquisition module in an apparatus for performing image search sorting according to an embodiment of the present invention.
  • FIG. 8 is a block diagram showing the structure of an apparatus for providing image search according to an embodiment of the present invention.
  • FIG. 9 is a block diagram showing another structure of an apparatus for providing image search according to an embodiment of the present invention.
  • FIG. 10 is a block diagram showing the structure of a pre-acquisition module in another apparatus for providing image search according to an embodiment of the present invention.
  • Figure 11 shows a block diagram of a computing device for performing an image search method in accordance with the present invention
  • Fig. 12 shows a storage unit for holding or carrying program code implementing the image search method according to the present invention.
  • Embodiment 1 of the method for performing image search sorting according to an embodiment of the present invention is shown, which may specifically include the following steps:
  • Step 110 Create an image family corresponding to multiple source images.
  • the same image family refers to images that are visually consistent from human beings. These images are images modified from a source image, since multiple images in the image family are modified by a source image. Therefore, each image in an image family should have the same source image; based on this, in the embodiment, it is proposed that the webpage corresponding to the source image is captured from the resource site by step S111, and the webpage is obtained by parsing the webpage.
  • the plurality of images corresponding to the source image; since each image in the same image family is derived from the same source image, after acquiring a plurality of images, the transmission between the multiple images corresponding to the source image may be acquired by step S112: Relationship, and using the propagation relationship between the plurality of images to create a plurality of image families.
  • the propagation relationship between multiple images corresponding to the source image is mainly included: reprinting, copying, and modifying, but is not limited thereto, and may have other propagation relationships. This is not described again; specifically, the embodiment obtains the propagation relationship by:
  • the corresponding relationship between the webpage URL (Uniform Resource Locator) and the plurality of image URLs is obtained by parsing the webpage page; Wherein, if the plurality of webpage URLs correspond to the same image URL, determining that the plurality of webpages including the image and the image are reprinted; or
  • Step 120 Calculate a reference weight of each image family
  • the value of different propagation relationships is different.
  • the value of the above three propagation relationships may be: modification > copy > reprint; wherein, the workload required for modification is greater than the simple preservation, The cost of saving the image and then providing the image service is greater than the reprint behavior; therefore, the cost means that the value of each image is different, that is, the basis weight of each propagation relationship; at the same time, after analysis, it is known that the difference is different.
  • the value of the image referenced by the site is also different. The site with a large amount of access has a large value of the image. Therefore, the site weight parameter is set in this embodiment. Specifically, this embodiment proposes to calculate each image by the following method. Family weights, including but not limited to:
  • the size of the modified relationship weight, the copy relationship weight, and the retransmission relationship weight are successively decremented; here, a formula is calculated to calculate the reference weight of the image family as follows;
  • a is the modification relationship weight
  • b is the replication relationship weight
  • c is the retransmission relationship weight
  • a>b>c is the weight of the site channel.
  • Step 130 According to the size of the reference weight of each image family, as a parameter of the search result feedback of the search query feedback.
  • the search as the feedback of the search query may be determined according to the size of the reference weight of each image family.
  • the sorted parameters of the result for example, the sorting of the search results is proportional to the size of the reference weights of the respective image families; of course, those skilled in the art can easily understand that the reference weights of the image families are used as search results.
  • the other parameters may be introduced as the sorting reference, which is not specifically limited in this embodiment.
  • Step 210 Create an image family including Figure A, Figure B, Figure C, Figure D, Figure E, and Figure F;
  • Step 220 Grab the webpage 7 to the webpage 14 from the resource site
  • Step 230 Parsing the webpage 7 to the webpage 14 to obtain a correspondence between the image url and the webpage url, that is, (picture url, webpage url) is: (A, 13), (B, 14), (C, 11), (D, 12), (F, 10), (E, 7), (E, 8), (E, 9); wherein the picture E corresponds to the web page 7, the web page 8, and the web page 9, so the web page 7, the web page 8 and the image in the web page 9 and the image E are reproduced relationships;
  • Step 240 After calculating the md5 values of the pictures A to F, it is known that the picture b, the picture E, and the md5 of the picture F are the same, so it can be determined that the picture B, the picture E, and the picture F are in a replication relationship;
  • Step 250 Perform an "approximate copy" calculation on picture A, picture B, picture C, and picture D having different md5 values, and determine that picture A, picture B, picture C, and picture D are an approximate copy, thereby determining picture A. , picture B, picture C and picture D are modified relationships;
  • Step 270 The size of the reference weight R of each image family is used as a parameter for sorting the search results fed back by the search query.
  • Embodiment 2 of the method for providing image search according to an embodiment of the present invention is shown, which may specifically include the following steps:
  • Step 310 After receiving the image query request, screening a plurality of image families related to the query request;
  • the method provided by the embodiment may further include: a step of pre-creating a plurality of image families corresponding to the source image and a step of calculating a reference weight for each image family; wherein the same image family refers to images that are visually consistent from a person, the images being a source image
  • the same image family refers to images that are visually consistent from a person, the images being a source image
  • the modified image since a plurality of images in the image family are modified by a source image, each image in one image family should have the same source image; based on this, in the present embodiment, a step can be adopted.
  • step S311 Grab the webpage corresponding to the source image from the resource site, and obtain multiple images corresponding to the source image by parsing the webpage page; since each image in the same image family is derived from the same source image, After the image is uploaded, the propagation relationship between the plurality of images corresponding to the source image may be acquired through step S312, and a plurality of image families are established by using the propagation relationship between the plurality of images.
  • the propagation relationship between multiple images corresponding to the source image is mainly included: reprinting, copying, and modifying, but is not limited thereto, and may have other propagation relationships.
  • the embodiment obtains the propagation relationship by: A.
  • the webpage URL is obtained by parsing the webpage page and Corresponding relationship between the plurality of image URLs; wherein, if the plurality of webpage URLs correspond to the same image URL, determining that the plurality of webpages including the image are in a reprint relationship with the image; or, B, grasping the source from the resource site
  • the plurality of images are obtained by parsing the webpage, and the information summary MD5 value of the plurality of images is calculated, wherein if the MD5 values of the multiple images are the same, it is determined that the plurality of images of the same MD5 are between To copy the relationship; otherwise, it is determined whether the multiple images are the same approximate copy, and if so, it is determined that the plurality of images having different MD5 values are modified.
  • the image family can be pre-created by other means, and the details are not described herein again.
  • the value of different propagation relationships is different.
  • the value of the above three propagation relationships may be: modification > copy > reprint; wherein the amount of work required for modification is greater than simple preservation, and the same image is saved. Then the cost of providing the image service is greater than the reprint behavior; therefore, this cost means that the value of each image is different, that is, the basis weight of each propagation relationship; at the same time, after analysis, it is known that different sites refer to The value of the image is also different. The site with a large amount of access has a large value of the image. Therefore, the site weight parameter is set in this embodiment. Specifically, this embodiment proposes to calculate the reference of each image family by the following method.
  • the weight including but not limited to: preset the weight of the resource site and different propagation relationships; wherein, if the propagation relationship includes reprinting, copying, and modifying, the weight relationship between the three is modified > copy > reprint , that is, setting the modified relationship weight, the copy relationship weight, and the resale relationship weight in the propagation relationship are successively decremented;
  • the formula weights reference image group as follows:
  • a is the modification relationship weight
  • b is the replication relationship weight
  • c is the retransmission relationship weight
  • a>b>c is the retransmission relationship weight
  • Step 320 Find a source image corresponding to each image family in the image family and a reference weight of each image family;
  • Step 330 Sort the source images in each image family according to the order of the reference weights to draw a search result corresponding to the query request.
  • the reference weights are first sorted, and then the matching between the source images is obtained by using a matching relationship between the source image corresponding to the image family and the reference weight of the image family, where the source image is
  • the ordering of the references is the same as the order of the weights of the reference weights, and then the ordering between the source images is one of the basis for drawing the search results of the query request; of course, those skilled in the art can easily understand that Other parameters may be added to the search result, which is not specifically limited in this embodiment.
  • Step 410 Receive an image query request, and filter an image family related to the query request; the image family includes a map A ⁇ , a diagram B ⁇ , a diagram C ⁇ , a diagram D ⁇ , a diagram E ⁇ , and a diagram F ⁇ ;
  • Step 420 Find a source image in the image family as Figure A ⁇ ;
  • Step 430 Grab the webpage 7 ⁇ page 14' from the resource site; parse the webpage 7 ⁇ page 14', and obtain the correspondence between the image url and the webpage url, that is, (picture url, webpage url) is: (A ⁇ , 13 ⁇ ), (B ⁇ , 14 ⁇ ), (C ⁇ , 11 ⁇ ), (D ⁇ , 12 ⁇ ), (F ⁇ , 10 ⁇ ), (E ⁇ , 7 ⁇ ), (E ⁇ , 8 ), (E ⁇ , 9 ⁇ ); wherein the picture E ⁇ corresponds to the web page 7 ⁇ , the web page 8' and the web page 9', so the image in the web page 7', the web page 8' and the web page 9' and the image E ⁇ is the reprint relationship;
  • Step 440 After calculating the md5 value of the picture A' ⁇ F', it is known that the picture B ⁇ , the picture E ⁇ , and the picture f'md5 are the same, so the picture B', the picture E', and the picture F' can be determined. Inter-replication relationship;
  • Step 450 Perform an "approximate copy" calculation on the picture A ⁇ , the picture B', the picture C', and the picture D' having different md5 values, and determine that the picture A', the picture B', the picture C', and the picture D' are an approximation. Copy, which determines the picture The relationship between A ⁇ , picture B ⁇ , picture C ⁇ and picture D ⁇ is modified;
  • the method may further include: a creating unit 510, configured to create an image family corresponding to multiple source images;
  • the calculating unit 520 is adapted to calculate a reference weight value of each image family;
  • the sorting unit 530 is adapted to use the size of the reference weight value of each image family as a parameter of the search result feedback of the search query feedback.
  • the method may include the following module: a creating unit 610, configured to create a family of images corresponding to multiple source images.
  • the calculation unit 620 is adapted to calculate a reference weight value of each image family;
  • the sorting unit 630 is adapted to use the size of the reference weight value of each image family as a parameter of the search result feedback of the search query feedback.
  • the creating unit 610 includes: a crawling module 6102, configured to capture the source image pair from a resource site.
  • the image processing module 6104 is configured to acquire a plurality of images corresponding to the source image by parsing the webpage page captured by the crawling module, and the acquiring module 6106 is configured to acquire the multiple images corresponding to the source image.
  • the propagation relationship; the family building module 6108 is adapted to establish a plurality of image families by using the propagation relationship between the plurality of images.
  • the method may further include the following modules: the obtaining module 6106 in the embodiment may further include The first processing module 610602 is configured to receive the analysis result of the parsing module, and obtain a correspondence between the webpage uniform resource locator URL and the image URL according to the parsing result; the first comparing module 610604 is adapted to compare the Corresponding relationship between the plurality of webpage URLs and the plurality of image URLs, and when the plurality of webpage URLs correspond to the same image URL, determining that the plurality of webpages including the image and the image are reprinted relationships.
  • the obtaining module 6106 further includes: a second processing module 610606, configured to calculate a message digest MD5 value of the plurality of images parsed by the parsing module; and a second comparing module 610608, configured to compare the plurality of images The MD5 value, and when the MD5 values of the plurality of images are the same, it is determined that the plurality of images having the same MD5 are in a copy relationship.
  • a second processing module 610606 configured to calculate a message digest MD5 value of the plurality of images parsed by the parsing module
  • a second comparing module 610608 configured to compare the plurality of images The MD5 value, and when the MD5 values of the plurality of images are the same, it is determined that the plurality of images having the same MD5 are in a copy relationship.
  • the obtaining module further includes: a third processing module 610610, configured to calculate a message digest MD5 value of the plurality of images parsed by the parsing module; and a third comparing module 610612, configured to compare the plurality of The MD5 value of the image, and when the MD5 values of the plurality of images are different, whether the plurality of images having different MD5 values are modified by the approximate copying manner.
  • a third processing module 610610 configured to calculate a message digest MD5 value of the plurality of images parsed by the parsing module
  • a third comparing module 610612 configured to compare the plurality of The MD5 value of the image, and when the MD5 values of the plurality of images are different, whether the plurality of images having different MD5 values are modified by the approximate copying manner.
  • the calculating unit 620 may further include: a setting module 6202, configured to preset a resource site of the crawling module to capture a webpage, and the different propagation relationship acquired by the acquiring module.
  • the weighting ratio matching module 6204 is adapted to calculate the reference weight of the image family by using the resource site in the same image family and the different propagation relationship weights.
  • FIG. 8 is a block diagram showing an embodiment of an apparatus for providing an image search according to an embodiment of the present invention.
  • the method may include the following module: a screening unit 810, configured to: after receiving an image query request, screening and a plurality of image families related to the query request; the searching unit 820 is adapted to search for a source image corresponding to each image family in the image family filtered by the filtering unit 810 and a reference weight of each image family; the drawing unit 830, The method is adapted to receive the search result of the searching unit 820, and sort the source images in each image family according to the order of the reference weights to draw a search result corresponding to the query request.
  • the method may include a module: a pre-creation unit 940, which is pre-configured to create an image corresponding to multiple source images.
  • a pre-computation unit 950 adapted to calculate a reference weight for each image family.
  • the filtering unit 910 is configured to: after receiving the image query request, filter a plurality of image families related to the query request; the searching unit 920 is configured to search for each image family in the image family selected by the screening unit 910. Source image and reference weight for each image family; drawing
  • the unit 930 is adapted to receive the search result of the searching unit 920, and sort the source images in each image family according to the order of the reference weights to draw a search result corresponding to the query request.
  • the pre-creation unit 940 includes: a pre-fetching module 9402, configured to capture a webpage corresponding to the source image from a resource site; and a pre-parsing module 9404, configured to parse the webpage page captured by the crawling module. Acquiring a plurality of images corresponding to the source image; the pre-acquisition module 9406 is adapted to acquire a propagation relationship between the plurality of images corresponding to the source image; and the pre-built family module 9408 is adapted to utilize the propagation between the multiple images Relationships establish multiple image families.
  • FIG. 10 is a block diagram showing a structure of a pre-acquisition module in an apparatus for providing an image search according to an embodiment of the present invention.
  • the pre-acquisition module 9406 may also be used in this embodiment.
  • the first pre-comparison module 940604 is adapted to receive the parsing result of the parsing module, and obtain the correspondence between the webpage uniform resource locator URL and the image URL according to the parsing result; the first pre-comparison module 940604 is adapted to Comparing the correspondence between the plurality of webpage URLs and the plurality of image URLs, and when the plurality of webpage URLs correspond to the same image URL, determining that the plurality of webpages including the image and the image are reprinted relationships.
  • the pre-acquisition module 9406 further includes: a second pre-processing module 940606, configured to calculate an information digest MD5 value of the plurality of images parsed by the parsing module; and a second pre-comparison module 940608, configured to compare the plurality of The MD5 value of the image, and when the MD5 values of the plurality of images are the same, it is determined that the plurality of images having the same MD5 are in a copy relationship.
  • a second pre-processing module 940606 configured to calculate an information digest MD5 value of the plurality of images parsed by the parsing module
  • a second pre-comparison module 940608 configured to compare the plurality of The MD5 value of the image, and when the MD5 values of the plurality of images are the same, it is determined that the plurality of images having the same MD5 are in a copy relationship.
  • the pre-acquisition module further includes: a third pre-processing module 940610, configured to calculate an information digest MD5 value of the plurality of images parsed by the parsing module; and a third pre-comparison module 940612, suitable for comparing the The MD5 value of the plurality of images is described, and when the MD5 values of the plurality of images are different, whether the plurality of images having different MD5 values are modified by the approximate copying manner is determined.
  • a third pre-processing module 940610 configured to calculate an information digest MD5 value of the plurality of images parsed by the parsing module
  • a third pre-comparison module 940612 suitable for comparing the The MD5 value of the plurality of images is described, and when the MD5 values of the plurality of images are different, whether the plurality of images having different MD5 values are modified by the approximate copying manner is determined.
  • the pre-calculation unit 950 may further include: a pre-setting module 9502, configured to preset a resource site of the crawling module to capture a webpage, and the different propagation acquired by the acquiring module.
  • the weighting of the relationship; the pre-matching module 9504 is adapted to calculate the reference weight of the image family by using the resource site in the same image family and the different propagation relationship weights.
  • modules in the devices of the embodiments can be adaptively changed and placed in one or more devices different from the embodiment.
  • the modules or units or components of the embodiments may be combined into one module or unit or component, and further they may be divided into a plurality of sub-modules or sub-units or sub-components.
  • any combination of the features disclosed in the specification, including the accompanying claims, the abstract and the drawings, and any methods so disclosed, or All processes or units of the device are combined.
  • Each feature disclosed in this specification (including the accompanying claims, the abstract and the drawings) may be replaced by alternative features that provide the same, equivalent or similar purpose.
  • the various component embodiments of the present invention may be implemented in hardware, or in a software module running on one or more processors, or in a combination thereof.
  • a microprocessor or digital signal processor may be used in practice to implement some or all of the functionality of some or all of the components of the web page loading device in accordance with embodiments of the present invention.
  • the invention can also be implemented as a device or device program (e.g., a computer program and a computer program product) for performing some or all of the methods described herein.
  • a program implementing the invention may be stored on a computer readable medium or may be in the form of one or more signals. Such signals may be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.
  • FIG. 11 illustrates a computing device that can implement an image search method according to the present invention, wherein the image search method includes the method of implementing image search ordering described in the above embodiments, and the above-described embodiments.
  • the computing device conventionally includes a processor 1110 and a program product or readable medium in the form of a memory 1120.
  • the memory 1120 may be an electronic memory such as a flash memory, an EEPROM (Electrically Erasable Programmable Read Only Memory), an EPROM, or a ROM.
  • the memory 1120 has a function for performing the above
  • the storage space 1130 for program code may include respective program codes 1131 for implementing various steps in the above methods, respectively.
  • These program codes can be read from or written to one or more program products.
  • These program products include program code carriers such as memory cards.
  • Such a program product is typically a portable or fixed storage unit as described with reference to FIG.
  • the storage unit may have a storage segment, a storage space, and the like that are similarly arranged to the storage 1120 in the computing device of FIG.
  • the program code can be compressed, for example, in an appropriate form.
  • the storage unit includes readable code 1131 ', ie, code that can be read by a processor, such as, for example, 1110, which when executed by a computing device causes the computing device to perform various steps in the methods described above .

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)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

L'invention concerne un procédé et un dispositif permettant de rechercher et de classer des images et de fournir une recherche d'images. Le procédé consiste : à créer des ensembles d'images correspondant à une pluralité d'images sources (110) ; à calculer le poids de référence de chaque ensemble d'images (120) ; et à utiliser la valeur du poids de référence de chaque ensemble d'images comme paramètre de classement d'un résultat de recherche renvoyé par l'interrogation de recherche (130). Ce procédé et ce dispositif permettent d'obtenir des résultats de classement meilleurs et plus précis, et les résultats du classement d'images ont un ordre de priorité dépendant du nombre de références, ce qui améliore nettement la précision de la recherche et accroît considérablement son efficacité.
PCT/CN2015/078881 2014-05-14 2015-05-13 Procédé et dispositif permettant de rechercher et de classer des images et de fournir une recherche d'images WO2015172721A1 (fr)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201410203342.3A CN103995856B (zh) 2014-05-14 2014-05-14 一种提供图像搜索的方法和装置
CN201410203342.3 2014-05-14
CN201410203700.0A CN103995857A (zh) 2014-05-14 2014-05-14 一种实现图像搜索排序的方法和装置
CN201410203700.0 2014-05-14

Publications (1)

Publication Number Publication Date
WO2015172721A1 true WO2015172721A1 (fr) 2015-11-19

Family

ID=54479338

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/078881 WO2015172721A1 (fr) 2014-05-14 2015-05-13 Procédé et dispositif permettant de rechercher et de classer des images et de fournir une recherche d'images

Country Status (1)

Country Link
WO (1) WO2015172721A1 (fr)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101398832A (zh) * 2007-09-30 2009-04-01 国际商业机器公司 利用人脸检测的图像搜索方法和系统
CN102713902A (zh) * 2009-12-02 2012-10-03 萨基姆通讯宽带公司 用于生成使用搜索引擎执行的搜索的结果的方法
WO2013075310A1 (fr) * 2011-11-24 2013-05-30 Microsoft Corporation Reclassement à l'aide d'échantillons d'images fiables
CN103995857A (zh) * 2014-05-14 2014-08-20 北京奇虎科技有限公司 一种实现图像搜索排序的方法和装置
CN103995856A (zh) * 2014-05-14 2014-08-20 北京奇虎科技有限公司 一种提供图像搜索的方法和装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101398832A (zh) * 2007-09-30 2009-04-01 国际商业机器公司 利用人脸检测的图像搜索方法和系统
CN102713902A (zh) * 2009-12-02 2012-10-03 萨基姆通讯宽带公司 用于生成使用搜索引擎执行的搜索的结果的方法
WO2013075310A1 (fr) * 2011-11-24 2013-05-30 Microsoft Corporation Reclassement à l'aide d'échantillons d'images fiables
CN103995857A (zh) * 2014-05-14 2014-08-20 北京奇虎科技有限公司 一种实现图像搜索排序的方法和装置
CN103995856A (zh) * 2014-05-14 2014-08-20 北京奇虎科技有限公司 一种提供图像搜索的方法和装置

Similar Documents

Publication Publication Date Title
US9008433B2 (en) Object tag metadata and image search
JP6785921B2 (ja) ピクチャ検索方法、装置、サーバー及び記憶媒体
CN112136123B (zh) 表征文件以进行相似性搜索
WO2015070673A1 (fr) Procédé pour une recherche de réseau côté navigateur et navigateur
CN105608159B (zh) 数据缓存的方法和装置
CN106855952B (zh) 基于神经网络的计算方法及装置
JP6932360B2 (ja) オブジェクト検索方法、装置およびサーバ
WO2016000507A1 (fr) Procédé de service de recherche de mode d'économie de trafic, serveur, client et système associés
US8861896B2 (en) Method and system for image-based identification
KR102361112B1 (ko) 유사 그룹 요소 추출
CN109190062B (zh) 目标语料数据的爬取方法、装置及存储介质
CN106777201B (zh) 搜索结果页上的推荐数据的排序方法及装置
US20120166412A1 (en) Super-clustering for efficient information extraction
CN108665459A (zh) 一种图像模糊检测方法、计算设备及可读存储介质
CN109635072B (zh) 舆情数据分布式存储方法、装置、存储介质和终端设备
CN109685069B (zh) 图像检测方法、装置及计算机可读存储介质
WO2015172721A1 (fr) Procédé et dispositif permettant de rechercher et de classer des images et de fournir une recherche d'images
CN106452855B (zh) 一种文章标签添加方法和装置
CN111782945B (zh) 书籍搜索方法、计算设备及存储介质
CN105589863B (zh) 一种搜索方法及数据处理方法、装置及系统
CN103995856B (zh) 一种提供图像搜索的方法和装置
US20160378839A1 (en) Searching method, searching apparatus and device
JP2018072873A (ja) 情報処理装置、情報処理方法、およびプログラム
JP5538459B2 (ja) 情報処理装置及び方法
WO2016033155A2 (fr) Mécanisme de tri de résultats de recherche sélectionné en fonction de l'étendue de requête

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: 15791932

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: 15791932

Country of ref document: EP

Kind code of ref document: A1