CN111782849B - Image retrieval method and device - Google Patents
Image retrieval method and device Download PDFInfo
- Publication number
- CN111782849B CN111782849B CN201911182241.1A CN201911182241A CN111782849B CN 111782849 B CN111782849 B CN 111782849B CN 201911182241 A CN201911182241 A CN 201911182241A CN 111782849 B CN111782849 B CN 111782849B
- Authority
- CN
- China
- Prior art keywords
- image
- neighbor
- weight
- vector
- vectors
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 239000000463 material Substances 0.000 claims abstract description 138
- 239000013598 vector Substances 0.000 claims abstract description 108
- 238000012545 processing Methods 0.000 claims description 8
- 238000003860 storage Methods 0.000 claims description 6
- 230000009467 reduction Effects 0.000 claims description 2
- 239000000284 extract Substances 0.000 abstract 1
- 238000011946 reduction process Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 206010033307 Overweight Diseases 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 235000019800 disodium phosphate Nutrition 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000013077 target material Substances 0.000 description 1
- 238000013024 troubleshooting Methods 0.000 description 1
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
- G06F16/58—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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Library & Information Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present disclosure provides an image retrieval method and apparatus. The image retrieval device extracts all feature vectors in the image after receiving the image sent by the user; retrieving neighbor vectors associated with each feature vector in the index file, and determining weights of the neighbor vectors; determining the weight of each material according to the corresponding relation between the neighbor vector and the material; calculating the material matching rate by using the weight of each material; and taking the material with the maximum matching rate as a search result. The method and the device can quickly and effectively search the material which is most matched with the image to be searched.
Description
Technical Field
The present disclosure relates to the field of image processing, and in particular, to an image retrieval method and apparatus.
Background
Currently, an advertisement or information display position is set on a platform home page, such as an e-commerce, for presenting corresponding picture materials. Abnormal coverage of auditors and forbidden categories or failure of later landing pages can cause unsatisfactory advertising materials to be played on the advertising space. Operators usually feed back the problem to maintenance personnel after finding the problem, so that the maintenance personnel can process the problem correspondingly.
Disclosure of Invention
The inventor finds through research that the feedback information received by the maintainer generally includes a problem advertisement position, a problem time, a problem material link, or an APP screenshot or a PC screenshot of the problem material. If the advertisement slot is sold in CPD (Cost Per Day, pay Per Day) mode, the advertisement slot plays little material every Day, and the troubleshooting is easy. However, if the bid advertisement is in the RTB (Real Time Bidding) mode, and the information distribution technology of thousands of people and thousands of sides is added, the amount of materials which are successfully bid and played is very large when the advertisement is recalled, and the abnormal advertisement can not be brushed by the inspector after the inspector always refreshes the page.
Therefore, the scheme capable of conveniently and rapidly retrieving the target material is provided.
According to a first aspect of an embodiment of the present disclosure, there is provided an image retrieval method, including: after receiving an image sent by a user, extracting all feature vectors in the image; retrieving a neighbor vector associated with each feature vector in an index file and determining the weight of the neighbor vector; determining the weight of each material according to the corresponding relation between the neighbor vector and the material; calculating the material matching rate by using the weight of each material; and taking the material with the maximum matching rate as a search result.
In some embodiments, determining the weights of the neighbor vectors includes: the number of times each neighbor vector is retrieved is taken as the weight of each neighbor vector.
In some embodiments, determining the weight of each material according to the correspondence between the neighbor vector and the material includes: and taking the sum of the weights of the neighbor vectors belonging to the same material as the weight of the corresponding material according to the corresponding relation between the neighbor vectors and the material.
In some embodiments, using the weights of the materials to perform the material matching rate calculation includes: for the ith material, judging whether the corresponding weight is larger than a weight threshold, wherein i is larger than or equal to 1 and smaller than or equal to N, and N is the total number of the materials; if the weight is larger than the weight threshold, inserting the identification of the ith material into a result list; and dividing the weight of each material by the total number of feature vectors in the image in the result list to obtain the matching rate of each material.
In some embodiments, inserting the identification of the ith material into a results list includes: judging whether the current length of the result list is smaller than a length threshold; if the current length of the result list is smaller than the length threshold, inserting the identification of the ith material into the result list; if the current length of the result list is not smaller than the length threshold, inserting the identification of the ith material into the result list, and deleting the material identification with the minimum weight in the result list.
In some embodiments, extracting all feature vectors in the image comprises: extracting an image feature vector from the image; and performing dimension reduction processing on the image feature vector.
In some embodiments, the material library is accessed at a predetermined period to download newly added material in the material library; processing the downloaded material to extract a feature vector; and writing the extracted feature vector into the index file.
According to a second aspect of the embodiments of the present disclosure, there is provided an image retrieval apparatus including: a receiving module configured to receive an image transmitted by a user; the image retrieval module is configured to extract all feature vectors in the image after the image is received, retrieve neighbor vectors associated with each feature vector in an index file, determine weights of the neighbor vectors, determine weights of all materials according to the corresponding relation between the neighbor vectors and the materials, calculate material matching rate by using the weights of all materials, and take the material with the maximum matching rate as a retrieval result.
According to a third aspect of the embodiments of the present disclosure, there is provided an image retrieval apparatus including: a memory configured to store instructions; a processor coupled to the memory, the processor configured to perform a method according to any of the embodiments described above based on instructions stored in the memory.
According to a fourth aspect of embodiments of the present disclosure, there is provided a computer readable storage medium, wherein the computer readable storage medium stores computer instructions which, when executed by a processor, implement a method as referred to in any of the embodiments above.
Other features of the present disclosure and its advantages will become apparent from the following detailed description of exemplary embodiments of the disclosure, which proceeds with reference to the accompanying drawings.
Drawings
In order to more clearly illustrate the embodiments of the present disclosure or the solutions in the prior art, the drawings that are required for the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present disclosure, and that other drawings may be obtained according to these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a flow chart of an image retrieval method according to an embodiment of the present disclosure;
fig. 2 is a flow chart of material matching rate calculation according to an embodiment of the disclosure;
FIG. 3 is a schematic diagram of an image retrieval device according to an embodiment of the present disclosure;
fig. 4 is a schematic structural view of an image retrieval device according to another embodiment of the present disclosure;
fig. 5 is a schematic structural view of an image retrieval device according to still another embodiment of the present disclosure.
Detailed Description
The following description of the technical solutions in the embodiments of the present disclosure will be made clearly and completely with reference to the accompanying drawings in the embodiments of the present disclosure, and it is apparent that the described embodiments are only some embodiments of the present disclosure, not all embodiments. The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the disclosure, its application, or uses. All other embodiments, which can be made by one of ordinary skill in the art without inventive effort, based on the embodiments in this disclosure are intended to be within the scope of this disclosure.
The relative arrangement of the components and steps, numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present disclosure unless it is specifically stated otherwise.
Meanwhile, it should be understood that the sizes of the respective parts shown in the drawings are not drawn in actual scale for convenience of description.
Techniques, methods, and apparatus known to one of ordinary skill in the relevant art may not be discussed in detail, but should be considered part of the specification where appropriate.
In all examples shown and discussed herein, any specific values should be construed as merely illustrative, and not a limitation. Thus, other examples of the exemplary embodiments may have different values.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further discussion thereof is necessary in subsequent figures.
Fig. 1 is a flowchart of an image retrieval method according to an embodiment of the present disclosure. In some embodiments, the following image retrieval method steps are performed by an image retrieval device.
In step 101, after receiving an image sent by a user, all feature vectors in the image are extracted.
In some embodiments, image feature vectors are extracted from the image and then subjected to a dimension reduction process to preserve the more energetic feature vectors.
For example, the feature vector is subjected to a dimension reduction process by means of SVD (Singular Value Decomposition ) or the like.
In some embodiments, feature vectors may be extracted using a SIFT (Scale-Invariant Feature Transform ) operator.
At step 102, neighbor vectors associated with each feature vector are retrieved from the index file and weights for the neighbor vectors are determined.
Here, the neighbor vector is a vector that is more similar to the feature vector. For example, the length and direction of the neighbor vector and the degree of difference between the feature vector are within a preset range.
In some embodiments, the number of times each neighbor vector is retrieved is taken as the weight of each neighbor vector.
For example, if the neighbor vectors of the feature vector q0 in the picture to be retrieved in the index file are p0, p1, p45, the hit number of p0, p1, p45 is added by 1. And finally, taking the final hit times of each neighbor vector as the weight value of each neighbor vector.
In step 103, the weight of each material is determined according to the corresponding relation between the neighbor vector and the material.
Here, the index file includes a plurality of feature vectors, and each feature vector belongs to a corresponding material.
In some embodiments, according to the corresponding relation between the neighbor vector and the material, the sum of the weights of the neighbor vectors belonging to the same material is used as the weight of the corresponding material.
For example, the neighbor vectors p0 and p1 belong to the material 1, and the sum of the weights of all the vectors belonging to the material 1, including p0 and p1, is used as the weight of the material 1.
In step 104, the material matching rate is calculated using the weight of each material.
In some embodiments, the weight of each material may be directly used for material matching rate calculation. In other embodiments, the material may be filtered first to perform the match rate calculation only for high-weight material.
Fig. 2 is a flow chart of material matching rate calculation according to an embodiment of the disclosure.
In step 201, an i-th material, i=1, is selected from the determined materials.
In step 202, it is determined whether the weight corresponding to the ith material is greater than a weight threshold.
If the weight is greater than the weight threshold, step 203 is executed; if the weight is not greater than the weight threshold, then step 206 is performed.
In step 203, it is determined whether the current length of the result list is less than a length threshold.
If the current length of the result list is less than the length threshold, step 204 is performed; if the current length of the result list is not less than the length threshold, step 205 is performed.
At step 204, the identification of the ith material is inserted into the results list.
In step 205, the identification of the ith material is inserted into the result list, and the material identification with the smallest weight in the result list is deleted.
Therefore, the materials with high weights can be reserved for matching rate calculation.
In step 206, i is updated, i.e., i=i+1.
In step 207, it is determined whether i is greater than the total number of materials N.
If i < N, then step 202 is performed. If i=n, then step 208 is performed.
In step 208, the weight of each material is divided by the total number of feature vectors in the image in the result list to obtain the matching rate of each material.
For example, if the weight of a certain material is count and the number of feature vectors of the image to be retrieved is n, the matching rate of the material is (count/n) 100%.
Returning to fig. 1. In step 105, the material with the largest matching rate is used as the search result.
In the image retrieval method provided by the embodiment of the disclosure, the related materials are retrieved by utilizing the image to be retrieved, and then the matching degree of each material is calculated, so that the material which is most matched with the image to be retrieved can be retrieved.
In some embodiments, the newly added material in the material library is downloaded by accessing the material library at predetermined periods. And processing the downloaded material to extract the feature vector. And writing the extracted feature vector into an index file so as to update the index file.
The present disclosure obtains material by delta and crops compression. And deleting the deleted material compared with the last timing task, and newly adding the material. Then cutting and compressing the advertisement graph according to the auditing standards of different advertisement positions to highlight the advertisement graph main body, and reducing the dimension of the feature vector in a SVD mode and the like to reserve the feature vector with stronger energy.
In addition, by coupling the index file generation module and the index file retrieval module, read-write separation is realized.
Fig. 3 is a schematic structural diagram of an image retrieval device according to an embodiment of the present disclosure. As shown in fig. 3, the image retrieval apparatus includes a receiving module 31 and an image retrieval module 32.
The receiving module 31 is configured to receive an image transmitted by a user.
The image retrieval module 32 is configured to extract all feature vectors in the image after receiving the image, retrieve neighbor vectors associated with each feature vector in the index file, determine weights of the neighbor vectors, determine weights of the materials according to correspondence between the neighbor vectors and the materials, perform material matching rate calculation by using the weights of the materials, and take the material with the largest matching rate as a retrieval result.
In some embodiments, image feature vectors are extracted from the image and then subjected to a dimension reduction process to preserve the more energetic feature vectors.
For example, the feature vector is subjected to a dimension reduction process by means of SVD (Singular Value Decomposition ) or the like.
In some embodiments, feature vectors may be extracted using a SIFT (Scale-Invariant Feature Transform ) operator.
In some embodiments, the number of times each neighbor vector is retrieved is taken as the weight of each neighbor vector.
For example, if the neighbor vectors of the feature vector q0 in the picture to be retrieved in the index file are p0, p1, p45, the hit number of p0, p1, p45 is added by 1. And finally, taking the final hit times of each neighbor vector as the weight value of each neighbor vector.
In some embodiments, according to the corresponding relation between the neighbor vector and the material, the sum of the weights of the neighbor vectors belonging to the same material is used as the weight of the corresponding material.
In some embodiments, using the weights of the materials to perform the material matching rate calculation includes: for the ith material, judging whether the corresponding weight is larger than a weight threshold, wherein i is larger than or equal to 1 and smaller than or equal to N, and N is the total number of the materials; if the weight is larger than the weight threshold, inserting the identification of the ith material into the result list; in the result list, the weight of each material is divided by the total number of feature vectors in the image to obtain the matching rate of each material.
In some embodiments, inserting the identification of the ith material into the results list includes: judging whether the current length of the result list is smaller than a length threshold; if the current length of the result list is smaller than the length threshold, inserting the identification of the ith material into the result list; if the current length of the result list is not smaller than the length threshold, inserting the identification of the ith material into the result list, and deleting the material identification with the minimum weight in the result list.
In some embodiments, the weight of a certain material is count, and the number of feature vectors of the image to be retrieved is n, and then the matching rate of the material is (count/n) 100%.
Fig. 4 is a schematic structural diagram of an image retrieval device according to another embodiment of the present disclosure. Fig. 4 differs from fig. 3 in that in the embodiment shown in fig. 4, the image retrieval device further comprises an index generation module 33.
The index generation module 33 downloads newly added material in the material library by accessing the material library at a predetermined period. And processing the downloaded material to extract the feature vector. And writing the extracted feature vector into an index file so as to update the index file.
Fig. 5 is a schematic structural view of an image retrieval device according to still another embodiment of the present disclosure. As shown in fig. 5, the apparatus includes a memory 51 and a processor 52.
The memory 51 is for storing instructions and the processor 52 is coupled to the memory 51, the processor 52 being configured to perform a method as referred to in any of the embodiments of fig. 1 or 2 based on the instructions stored by the memory.
As shown in fig. 5, the apparatus further comprises a communication interface 53 for information interaction with other devices. Also, the apparatus includes a bus 54, and the processor 52, the communication interface 53, and the memory 51 communicate with each other via the bus 54.
The memory 51 may comprise a high-speed RAM memory or may further comprise a non-volatile memory (non-volatile memory), such as at least one disk memory. The memory 51 may also be a memory array. The memory 51 may also be partitioned and the blocks may be combined into virtual volumes according to certain rules.
Further, the processor 52 may be a central processing unit CPU, or may be an application specific integrated circuit ASIC, or one or more integrated circuits configured to implement embodiments of the present disclosure.
The present disclosure also relates to a computer readable storage medium having stored thereon computer instructions which, when executed by a processor, implement a method as referred to in any of the embodiments of fig. 1 or 2.
In some embodiments, the functional unit blocks described above may be implemented as general-purpose processors, programmable logic controllers (Programmable Logic Controller, abbreviated as PLCs), digital signal processors (Digital Signal Processor, abbreviated as DSPs), application specific integrated circuits (Application Specific Integrated Circuit, abbreviated as ASICs), field programmable gate arrays (Field-Programmable Gate Array, abbreviated as FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or any suitable combination thereof for performing the functions described in the present disclosure.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and the storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The description of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the disclosure in the form disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments were chosen and described in order to best explain the principles of the disclosure and the practical application, and to enable others of ordinary skill in the art to understand the disclosure for various embodiments with various modifications as are suited to the particular use contemplated.
Claims (8)
1. An image retrieval method comprising:
after receiving an image sent by a user, extracting all feature vectors in the image;
retrieving a neighbor vector associated with each feature vector in an index file, and determining the weight of the neighbor vector, wherein the number of times each neighbor vector is retrieved is taken as the weight of each neighbor vector, and the degree of difference between the neighbor vector and the associated feature vector is within a preset range;
determining the weight of each material according to the corresponding relation between the neighbor vector and the material, wherein the sum of the weights of the neighbor vectors belonging to the same material is used as the weight of the corresponding material according to the corresponding relation between the neighbor vector and the material;
calculating the material matching rate by using the weight of each material;
and taking the material with the maximum matching rate as a search result.
2. The method of claim 1, wherein using the weights of the materials for material match rate calculation comprises:
for the ith material, judging whether the corresponding weight is greater than a weight threshold, wherein i is greater than or equal to 1 and less than or equal to N, and N is the total number of the materials;
if the weight is greater than the weight threshold, inserting the identification of the ith material into a result list;
and dividing the weight of each material by the total number of feature vectors in the image in the result list to obtain the matching rate of each material.
3. The method of claim 2, wherein inserting the identification of the i-th material into a results list comprises:
judging whether the current length of the result list is smaller than a length threshold;
if the current length of the result list is smaller than the length threshold, inserting the identification of the ith material into the result list;
if the current length of the result list is not smaller than the length threshold, inserting the identification of the ith material into the result list, and deleting the material identification with the minimum weight in the result list.
4. The method of claim 1, wherein extracting all feature vectors in the image comprises:
extracting an image feature vector from the image;
and performing dimension reduction processing on the image feature vector.
5. The method of any of claims 1-4, further comprising:
accessing a material library in a preset period so as to download newly added materials in the material library;
processing the downloaded material to extract a feature vector;
and writing the extracted feature vector into the index file.
6. An image retrieval apparatus comprising:
a receiving module configured to receive an image transmitted by a user;
the image retrieval module is configured to extract all feature vectors in the image after the image is received, retrieve neighbor vectors associated with each feature vector in an index file, determine weights of the neighbor vectors, determine weights of the materials according to the corresponding relation between the neighbor vectors and the materials, calculate a material matching rate by using the weights of the materials, and take the material with the maximum matching rate as a retrieval result, wherein the number of times each neighbor vector is retrieved is taken as the weight of each neighbor vector, the difference degree of the neighbor vectors and the associated feature vector is within a preset range, and take the sum of the weights of the neighbor vectors belonging to the same material as the weight of the corresponding material according to the corresponding relation between the neighbor vectors and the materials.
7. An image retrieval apparatus comprising:
a memory configured to store instructions;
a processor coupled to the memory, the processor configured to perform a method implementing any of claims 1-5 based on instructions stored by the memory.
8. A computer readable storage medium storing computer instructions which, when executed by a processor, implement the method of any one of claims 1-5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911182241.1A CN111782849B (en) | 2019-11-27 | 2019-11-27 | Image retrieval method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911182241.1A CN111782849B (en) | 2019-11-27 | 2019-11-27 | Image retrieval method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111782849A CN111782849A (en) | 2020-10-16 |
CN111782849B true CN111782849B (en) | 2024-03-01 |
Family
ID=72755737
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911182241.1A Active CN111782849B (en) | 2019-11-27 | 2019-11-27 | Image retrieval method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111782849B (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20040034337A (en) * | 2002-10-14 | 2004-04-28 | 삼성전자주식회사 | Image retrieval method and apparatus using iterative matching |
WO2013159722A1 (en) * | 2012-04-25 | 2013-10-31 | Tencent Technology (Shenzhen) Company Limited | Systems and methods for obtaining information based on an image |
CN103577409A (en) * | 2012-07-19 | 2014-02-12 | 阿里巴巴集团控股有限公司 | Method and device for establishing image indexes in image searches |
CN104715449A (en) * | 2015-03-31 | 2015-06-17 | 百度在线网络技术(北京)有限公司 | Method and device for generating mosaic image |
CN108304431A (en) * | 2017-06-14 | 2018-07-20 | 腾讯科技(深圳)有限公司 | A kind of image search method and device, equipment, storage medium |
CN108664583A (en) * | 2018-05-04 | 2018-10-16 | 北京物灵智能科技有限公司 | A kind of index tree method for building up and image search method |
CN110297935A (en) * | 2019-06-28 | 2019-10-01 | 京东数字科技控股有限公司 | Image search method, device, medium and electronic equipment |
-
2019
- 2019-11-27 CN CN201911182241.1A patent/CN111782849B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20040034337A (en) * | 2002-10-14 | 2004-04-28 | 삼성전자주식회사 | Image retrieval method and apparatus using iterative matching |
WO2013159722A1 (en) * | 2012-04-25 | 2013-10-31 | Tencent Technology (Shenzhen) Company Limited | Systems and methods for obtaining information based on an image |
CN103577409A (en) * | 2012-07-19 | 2014-02-12 | 阿里巴巴集团控股有限公司 | Method and device for establishing image indexes in image searches |
CN104715449A (en) * | 2015-03-31 | 2015-06-17 | 百度在线网络技术(北京)有限公司 | Method and device for generating mosaic image |
CN108304431A (en) * | 2017-06-14 | 2018-07-20 | 腾讯科技(深圳)有限公司 | A kind of image search method and device, equipment, storage medium |
CN108664583A (en) * | 2018-05-04 | 2018-10-16 | 北京物灵智能科技有限公司 | A kind of index tree method for building up and image search method |
CN110297935A (en) * | 2019-06-28 | 2019-10-01 | 京东数字科技控股有限公司 | Image search method, device, medium and electronic equipment |
Non-Patent Citations (1)
Title |
---|
基于Spark平台的人脸图像检索系统;陈新荃 等;计算机工程;第44卷(第2期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN111782849A (en) | 2020-10-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11586664B2 (en) | Image retrieval method and apparatus, and electronic device | |
CN109634698B (en) | Menu display method and device, computer equipment and storage medium | |
US11481402B2 (en) | Search ranking method and apparatus, electronic device and storage medium | |
CN109815487B (en) | Text quality inspection method, electronic device, computer equipment and storage medium | |
EP2815335A1 (en) | Method of machine learning classes of search queries | |
CN112889042A (en) | Identification and application of hyper-parameters in machine learning | |
CN108021708B (en) | Content recommendation method and device and computer readable storage medium | |
CN109388760B (en) | Recommendation label obtaining method, media content recommendation method, device and storage medium | |
KR102222087B1 (en) | Image recognition method and apparatus based on augmented reality | |
WO2011011046A1 (en) | Ranking search results based on word weight | |
CN103218355A (en) | Method and device for generating tags for user | |
CN107291825A (en) | With the search method and system of money commodity in a kind of video | |
CN110705245A (en) | Method and device for acquiring reference processing scheme and storage medium | |
CN111324804B (en) | Search keyword recommendation model generation method, keyword recommendation method and device | |
CN109582155B (en) | Recommendation method and device for inputting association words, storage medium and electronic equipment | |
CN111552767A (en) | Search method, search device and computer equipment | |
CN111651669A (en) | Information recommendation method and device, electronic equipment and computer-readable storage medium | |
CN109992659B (en) | Method and device for text sorting | |
CN111797319A (en) | Recommendation method, device, equipment and storage medium | |
CN110309293A (en) | Text recommended method and device | |
CN109388740A (en) | A kind of monitoring method and device of spreading network information effect | |
CN111782849B (en) | Image retrieval method and device | |
CN106257449A (en) | A kind of information determines method and apparatus | |
CN110674388A (en) | Mapping method and device for push item, storage medium and terminal equipment | |
CN111738754A (en) | Object recommendation method and device, storage medium and computer equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |