CN101470730A - Image repetition detection method based on spectrum characteristics analysis - Google Patents
Image repetition detection method based on spectrum characteristics analysis Download PDFInfo
- Publication number
- CN101470730A CN101470730A CNA2007103042078A CN200710304207A CN101470730A CN 101470730 A CN101470730 A CN 101470730A CN A2007103042078 A CNA2007103042078 A CN A2007103042078A CN 200710304207 A CN200710304207 A CN 200710304207A CN 101470730 A CN101470730 A CN 101470730A
- Authority
- CN
- China
- Prior art keywords
- image
- gradient
- spectrum
- information
- sampling
- 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.)
- Granted
Links
- 238000001228 spectrum Methods 0.000 title claims abstract description 41
- 238000001514 detection method Methods 0.000 title claims abstract description 16
- 238000004458 analytical method Methods 0.000 title claims abstract description 9
- 238000010183 spectrum analysis Methods 0.000 claims abstract description 21
- 238000005070 sampling Methods 0.000 claims abstract description 19
- 238000013519 translation Methods 0.000 claims abstract description 16
- 238000000034 method Methods 0.000 claims abstract description 7
- 239000004744 fabric Substances 0.000 claims description 15
- 238000013507 mapping Methods 0.000 claims description 15
- 238000004040 coloring Methods 0.000 claims description 2
- 230000004927 fusion Effects 0.000 abstract 1
- 238000006243 chemical reaction Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 5
- 230000004044 response Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 240000008042 Zea mays Species 0.000 description 1
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 1
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 235000005822 corn Nutrition 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 230000036651 mood Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to an image repeated detection method based on the frequency spectrum analysis, which comprises conducting the down-sampling for images through a down-sampling module, conducting the frequency spectrum analysis for the images through a frequency spectrum analyzing module, extracting the color information of image pixel, and conducting the gradient information analysis for the images through the frequency spectrum analyzing module, acquiring the image gradient distribution characteristic after the down-sampling, wherein the characteristic comprises the performances of rotation, translation and unchangeable dimension, and generating module fusion image color and gradient information through indexes as image indexes. The method after adopted can greatly reduce the storage redundancy of a database, and increase the retrieval performance and efficiency of an existing retrieval system.
Description
Technical field
The present invention relates to the multimedia retrieval technology in Computer Applied Technology field, is a kind of image duplicate detection method based on spectrum sigtral response.
Background technology
Along with the develop rapidly of multimedia technology and computer network, the whole world comprises that the multi-medium data of digital picture, audio frequency, video increases with surprising rapidity.View data particularly because it is directly perceived, the content expressed in abundance, can carry out factor such as personalization editor, is subjected to user's favor very much.But the view data of the magnanimity level that every day is newly-generated owing to its huge redundancy, has unfeelingly been engulfed our finite storage space, and has been flooded a lot of Useful Informations, thereby has brought very big trouble for user's view data search.How just can make the user effectively utilize information and to the quick location of required multimedia resource, conveniently obtain and effectively management be a problem demanding prompt solution, particularly those wish can safeguard at low cost and the update image database for internet hunt company and individuation data suppliers that the user serves especially.Yet, commercial graphic search engine such as maturations such as google, yahoo and Baidu nearly all comes thumbnail with text at present, will cause like this a large amount of text index differences occurring and content image about the same, thereby take the storage space of a large amount of preciousnesses at database; More seriously, the serious decline of quality meeting of user search service because there is very big repeatability in the image that retrieval is come out, directly can influence user's use mood, causes being discontented with of user.Therefore, the duplicate detection problem in the image data base seems particularly important, yet duplicate detection the very corn of a subject is an image expression.As long as the image index in the database is well-done, just can carry out the image duplicate detection effectively.
The essence of image expression is to seek a suitable feature mapping function, and this function can be with image mapped to the also low higher dimensional space of similarity between similarity height but also class in the class not only.In the image expression research field, roughly there is two types Feature Mapping function, they are based on bottom visual signature and high-level semantic feature respectively, and we claim that they are low-level image feature mapping function and high-level semantic Feature Mapping function.The low-level image feature mapping function mainly is some bottom-up informations that obtain image, and these information spinners will comprise color, texture, gradient etc.The major advantage of this function is easy to operate, the flexibility ratio height, and computation complexity is low etc., and its major defect is the high-layer semantic information that lacks image.By contrast, high-level semantic Feature Mapping function mainly is to obtain the target of the existence in the image or the semantic information of entire image scene, its major advantage is effectively to carry out image understanding, thereby express image more exactly, but its main shortcoming is the computation complexity height, needs setup parameter more, flexibility ratio is low, can not be used on a large scale.The relative merits of comprehensive above two kinds of mapping functions and the characteristic of image data base duplicate detection problem itself, we determine to adopt the low-level image feature mapping function to go to catch image information, its main cause is as follows: the scale of (1) image data base is bigger, and is very strict to the requirement of computation complexity; If adopt high-level semantic Feature Mapping function, the cost of database maintenance and renewal is very huge.(2) this problem of image duplicate detection itself to picture material express less demanding.Usually, the image that two width of cloth repeat can be changed mutually by some conversion, and these conversion mainly comprise translation, rotation and yardstick, and its conversion amplitude is very little.The low-level image feature mapping function can be handled the influence that above these conversion cause fully.(3) low-level image feature mapping function flexibility ratio height is easy to Computer Processing; And high-level semantic Feature Mapping function is just in time opposite, and it is subjected to all multifactor restrictions, and a lot of empirical parameters need be set in advance, is unfavorable for Computer Processing.So under this background, we have proposed the feature of a kind of fused images color and gradient information.In this feature, color of image information is that the frequency spectrum by image embodies, and if the frequency spectrum of image gradient information spinner by the image gradient direction histogram embody.In following components, we will introduce this feature in detail.
Summary of the invention
The present invention proposes a kind of image duplicate detection method based on spectrum sigtral response, this method has adopted color of image integrated and the image index of gradient information,, be applied to image duplicate detection field.
The image duplicate detection method based on spectrum analysis that the present invention proposes comprises:
Adopt down sample module that image is carried out down-sampling, be used to shorten the computing time that image spectrum is analyzed;
Adopt spectrum analysis module that image is carried out spectrum analysis, be used to extract the low frequency spectrum information of the image behind the down-sampling, obtain the colouring information of image pixel;
Adopt spectrum analysis module that image is carried out the gradient information analysis; Be used to obtain the image gradient distribution characteristics behind the down-sampling;
Adopt index generation module fused images color and gradient information, as image index;
With the image index is foundation, and whether detected image repeats.
Further, the low frequency spectrum information of the image behind the described down-sampling comprises that rotation, translation and yardstick are constant.
Further, the image gradient distribution characteristics behind the described down-sampling has the constant character of rotation, translation and yardstick.
Further, described spectrum analysis step comprises:
Image is carried out Fourier transform, obtain its amplitude spectrum;
Amplitude spectrum is carried out log-polar coordinate mapping form the new image of a width of cloth;
The image new to this width of cloth carries out Fourier transform, obtains amplitude spectrum;
Only keep low frequency spectrum information, be used for the feature of picture engraving color distribution.
Further, described gradient information analytical procedure comprises:
Extract the gradient of each pixel of image, add up the gradient direction of all pixels;
Director space is quantified as n grade;
The gradient direction of each pixel is mapped to corresponding grade, has been built into a gradient orientation histogram;
Histogram is carried out Fourier transform, get its amplitude spectrum;
With the feature of amplitude spectrum as the picture engraving gradient information.
The image duplicate detection method that the present invention proposes based on spectrum sigtral response, the image low-level image feature of use has been expressed structure and the detailed information in the image preferably; It can detect image data base Chinese version index difference effectively and content image about the same, thereby greatly save valuable storage space as image index, thereby has improved the quality of user search service.
Description of drawings
Fig. 1 is a system architecture diagram of the present invention;
Fig. 2 detects the application example of repetition mountains and rivers image for the present invention;
Fig. 3 detects repetition aircraft image example for the present invention;
Fig. 4 detects repetition streams image example for the present invention.
Embodiment
Below in conjunction with accompanying drawing the present invention is described in detail, be to be noted that described embodiment only is intended to be convenient to the understanding of the present invention, and it is not played any qualification effect.
General structure of the present invention is made up of following three modules: one, down sample module, the function of this module are that image is carried out down-sampling.Two, spectrum analysis module, the function of this module are to carry out image spectrum analysis and image gradient information analysis.Three, image index generation module, the task of this module are that two kinds of features that spectrum analysis module obtains are united, thereby form one to the unusual image index of robust of image rotation, translation and dimensional variation.We utilize this index to carry out the image duplicate detection.Fig. 1 has shown one-piece construction of the present invention.
Provide the explanation of each related in this invention technical scheme detailed problem below in detail.
1) by down sample module image is carried out down-sampling.
At first image is carried out down-sampling, the purpose of doing like this is to reduce computation complexity, also can obtain the structural information of image large scale simultaneously.Usually, the down-sampling rate is 0.5, and just size of images becomes original half.In addition, if image to be processed is a coloured image, we at first are converted to gray level image with coloured image so.
2) by spectrum analysis module image is carried out spectrum analysis.
Image behind the down-sampling is carried out Fourier transform, obtain its amplitude spectrum; Then amplitude spectrum is carried out log-polar coordinate mapping and form the new image of a width of cloth; Then the new image of this width of cloth is carried out Fourier transform, obtain amplitude spectrum; Only keep some low-frequency spectrum informations at last.The spectrum information that is obtained through above processing back is constant in graphical rule, rotation and translation transformation, and these attributes can be confirmed by the theoretical analysis below us.
A given width of cloth original image f
a(x, y), we are rotated, obtain the new image f of a width of cloth behind yardstick and the translation transformation it
b(x, y), wherein the anglec of rotation is α, and scale factor is σ, and the translational movement of x direction is x
0, and the translational movement of y direction is y
0, can illustrate with following formula with up-conversion process:
f
b(x，y)＝f
a[σ(xcosα+ysinα)—x
0，σ(—xsinα+ycosα)—y
0]。
We are respectively to f
a(x, y) and f
b(x y) carries out Fourier transform, obtains corresponding frequency spectrum F
a(u, v) and F
b(u, v).According to Fourier transform theory, F
a(u, v) and F
b(v) there is following relation in u:
Wherein, φ
b(u v) is f
b(this phase spectrum is relevant with the anglec of rotation, scale factor and translational movement for x, phase spectrum y); If but we only consider amplitude spectrum | F
b(u, v) |, will find | F
b(u, v) | be translation invariant, that is:
|F
b(u，v)|＝σ
-2|F
a[σ
-1(ucosα+vsinα)
0，σ
-1(—usinα+vcosα)]|。
Above formula shows, | F
a(u, v) | and | F
b(u, v) | between by getting in touch between anglec of rotation α and the scale factor σ.Then we polar coordinate system (r, express again in θ) | F
a(u, v) | and | F
b(u, v) |, i.e. u=rcos θ and v=rsin θ.Thereby can derive following relation:
f
ap(θ，r)＝|F
a(rcosθ，rsinθ)|；f
bp(θ，r)＝|F
b(rcosθ，rsinθ)|
Simplify computing through some, we can draw:
f
bp(θ，r)＝σ
-2f
ap(θ—α，r/σ)。
This sampled images rotation conversion just is converted into along the translation of angle axle θ and the change of scale of radius axle r.Then, we are further with f
Bp(r, radius axle r θ) are mapped in the logarithmic coordinate system, and I have just had following relational expression like this:
f
bpl(θ，λ)＝f
bp(θ，r)＝σ
-2f
apl(θ—α，λ—η)，
Wherein, λ=log (r), η=log (σ).Like this, the rotation of image and translation just are reduced to along the translation of λ axle and η axle; We are to f then
Bpl(θ λ) carries out Fourier transform; According to the Fourier transform theory, we have following relation:
F
bpl(ξ，ζ)＝σ
-2e
-j2π(ξη+ζλ)F
apl(ξ，ζ)。
We are right | F
Bpl(ξ, ζ) | carry out normalization, get it then as last spectrum signature, like this | F
Bpl(ξ, ζ) | be rotation, yardstick and translation invariant.
3) by spectrum analysis module image is carried out the gradient information analysis.
The gradient of each pixel of image behind the extraction down-sampling is added up the gradient direction of all pixels then; Then, we are 36 grades with the director space equal interval quantizing; We are mapped to corresponding grade with the gradient direction of each pixel then, have been built into a gradient orientation histogram like this; Then histogram is carried out Fourier transform, get its amplitude spectrum.Because gradient orientation histogram itself has yardstick and translation invariant attribute, remove amplitude spectrum after adding Fourier transform, the feature that has obtained since like this has the characteristic with invariable rotary.
4), realize that feature is integrated by index generation module fused images color and gradient information.
The gradient orientation histogram spectrum signature that spectrum signature that second step obtained and the 3rd step obtain, they all have rotation, yardstick and translation invariant attribute.Whether we work the index that is used as piece image with these two kinds of characteristics combination, use Euclidean distance to measure distance between two width of cloth images respectively, define a threshold value in addition again and differentiate two width of cloth images and repeat; If two kinds of distances and less than this threshold value, we just think that two width of cloth images of comparison are repetitions; Vice versa.
Fig. 2, Fig. 3 and Fig. 4 have provided three application examples of our image duplicate detection systems.In Fig. 2, we choose the width of cloth in the database to contain the image in the mountains and rivers, it is extracted the image low-level image feature based on spectrum analysis that we propose, utilize this feature to carry out images match then, thereby pick out 4 pictures that repeat with retrieving images in the database.In Fig. 3, we choose the width of cloth in the database to contain the image of aircraft, it is extracted the image low-level image feature based on spectrum analysis that we propose, utilize this feature to carry out images match then, thereby pick out 3 pictures that repeat with retrieving images in the database.In Fig. 4, we choose the width of cloth in the database to contain the image in streams, it is extracted the image low-level image feature based on spectrum analysis that we propose, utilize this feature to carry out images match then, thereby pick out 3 pictures that repeat with retrieving images in the database.
The above; only be the embodiment among the present invention; but protection scope of the present invention is not limited thereto; anyly be familiar with the people of this technology in the disclosed technical scope of the present invention; can understand conversion or the replacement expected; all should be encompassed in of the present invention comprising within the scope, therefore, protection scope of the present invention should be as the criterion with the protection domain of claims.
Claims (5)
1. the image duplicate detection method based on spectrum analysis is characterized in that, comprising:
By down sample module image is carried out down-sampling, be used to shorten the computing time that image spectrum is analyzed;
By spectrum analysis module image is carried out spectrum analysis, be used to extract the low frequency spectrum information of the image behind the down-sampling, obtain the colouring information of image pixel;
By spectrum analysis module image is carried out the gradient information analysis, be used to obtain the image gradient distribution characteristics behind the down-sampling;
By index generation module fused images color and gradient information, as image index;
With the image index is foundation, and whether detected image repeats.
2. method according to claim 1 is characterized in that, the low frequency spectrum information of the image behind the described down-sampling comprises that rotation, translation and the yardstick of image is constant.
3. method according to claim 1 is characterized in that, the image gradient distribution characteristics behind the described down-sampling has the constant character of rotation, translation and yardstick.
4. method according to claim 1 is characterized in that, described spectrum analysis step comprises:
Image is carried out Fourier transform, obtain its amplitude spectrum;
Amplitude spectrum is carried out log-polar coordinate mapping form the new image of a width of cloth;
The image new to this width of cloth carries out Fourier transform, obtains amplitude spectrum;
Only keep low frequency spectrum information, be used for the feature of picture engraving color distribution.
5. method according to claim 1 is characterized in that, described gradient information analytical procedure comprises:
Extract the gradient of each pixel of image, add up the gradient direction of all pixels;
Director space is quantified as n grade;
The gradient direction of each pixel is mapped to corresponding grade, has been built into a gradient orientation histogram;
Histogram is carried out Fourier transform, get its amplitude spectrum;
With the feature of amplitude spectrum as the picture engraving gradient information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007103042078A CN101470730B (en) | 2007-12-26 | 2007-12-26 | Image repetition detection method based on spectrum characteristics analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007103042078A CN101470730B (en) | 2007-12-26 | 2007-12-26 | Image repetition detection method based on spectrum characteristics analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101470730A true CN101470730A (en) | 2009-07-01 |
| CN101470730B CN101470730B (en) | 2010-12-22 |
Family
ID=40828206
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2007103042078A Active CN101470730B (en) | 2007-12-26 | 2007-12-26 | Image repetition detection method based on spectrum characteristics analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101470730B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102881186A (en) * | 2011-07-11 | 2013-01-16 | 歌乐株式会社 | Environment recognizing device for a vehicle and vehicle control system using the same |
| CN103593677A (en) * | 2013-12-02 | 2014-02-19 | 中国科学院自动化研究所 | Near-duplicate image detection method |
| CN103729368A (en) * | 2012-10-13 | 2014-04-16 | 复旦大学 | Robust voice frequency recognizing method based on local frequency spectrum image descriptors |
| CN103955496A (en) * | 2014-04-18 | 2014-07-30 | 大连恒锐科技股份有限公司 | Fast field tire trace pattern retrieval algorithm |
| CN105005565A (en) * | 2014-04-18 | 2015-10-28 | 大连恒锐科技股份有限公司 | Onsite sole trace pattern image retrieval method |
| CN105023027A (en) * | 2014-04-18 | 2015-11-04 | 大连恒锐科技股份有限公司 | Sole trace pattern image retrieval method based on multi-feedback mechanism |
| CN107423715A (en) * | 2017-07-31 | 2017-12-01 | 大连海事大学 | A kind of footprint automatic identifying method based on multiple features combining decision-making |
| CN112015929A (en) * | 2019-05-30 | 2020-12-01 | 阿里巴巴集团控股有限公司 | Raster data access method and device, electronic equipment and computer storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7386170B2 (en) * | 2000-06-30 | 2008-06-10 | Texas Instruments Incorporated | Image object ranking |
| KR100788643B1 (en) * | 2001-01-09 | 2007-12-26 | 삼성전자주식회사 | Searching method of image based on combination of color and texture |
| CN100363943C (en) * | 2004-06-21 | 2008-01-23 | 南开大学 | Color image matching analytical method based on color content and distribution |
-
2007
- 2007-12-26 CN CN2007103042078A patent/CN101470730B/en active Active
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102881186A (en) * | 2011-07-11 | 2013-01-16 | 歌乐株式会社 | Environment recognizing device for a vehicle and vehicle control system using the same |
| US8903603B2 (en) | 2011-07-11 | 2014-12-02 | Clarion Co., Ltd. | Environment recognizing device for a vehicle and vehicle control system using the same |
| CN103729368A (en) * | 2012-10-13 | 2014-04-16 | 复旦大学 | Robust voice frequency recognizing method based on local frequency spectrum image descriptors |
| CN103729368B (en) * | 2012-10-13 | 2016-12-21 | 复旦大学 | A kind of robust audio recognition methods based on local spectrum iamge description |
| CN103593677A (en) * | 2013-12-02 | 2014-02-19 | 中国科学院自动化研究所 | Near-duplicate image detection method |
| CN105023027A (en) * | 2014-04-18 | 2015-11-04 | 大连恒锐科技股份有限公司 | Sole trace pattern image retrieval method based on multi-feedback mechanism |
| CN105005565A (en) * | 2014-04-18 | 2015-10-28 | 大连恒锐科技股份有限公司 | Onsite sole trace pattern image retrieval method |
| CN103955496A (en) * | 2014-04-18 | 2014-07-30 | 大连恒锐科技股份有限公司 | Fast field tire trace pattern retrieval algorithm |
| CN105023027B (en) * | 2014-04-18 | 2019-03-05 | 大连恒锐科技股份有限公司 | Live soles spoor decorative pattern image search method based on multiple feedback mechanism |
| CN107423715A (en) * | 2017-07-31 | 2017-12-01 | 大连海事大学 | A kind of footprint automatic identifying method based on multiple features combining decision-making |
| CN107423715B (en) * | 2017-07-31 | 2020-05-19 | 大连海事大学 | Automatic footprint identification method based on multi-feature joint decision |
| CN112015929A (en) * | 2019-05-30 | 2020-12-01 | 阿里巴巴集团控股有限公司 | Raster data access method and device, electronic equipment and computer storage medium |
| WO2020238749A1 (en) * | 2019-05-30 | 2020-12-03 | 阿里巴巴集团控股有限公司 | Raster data access method and apparatus, electronic device, and computer storage medium |
| US11907283B2 (en) | 2019-05-30 | 2024-02-20 | Alibaba Group Holding Limited | Multi-resolution raster data access method, apparatus, electronic device, and computer storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101470730B (en) | 2010-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101470730B (en) | Image repetition detection method based on spectrum characteristics analysis | |
| Gauen et al. | Comparison of visual datasets for machine learning | |
| US8295651B2 (en) | Coherent phrase model for efficient image near-duplicate retrieval | |
| US20160048849A1 (en) | Method and system for clustering and classifying online visual information | |
| Yu et al. | Stratified pooling based deep convolutional neural networks for human action recognition | |
| Xu et al. | Satellite image spoofing: Creating remote sensing dataset with generative adversarial networks (short paper) | |
| CN104486585B (en) | A kind of city magnanimity monitor video management method and system based on GIS | |
| CN109344298A (en) | A kind of method and device converting unstructured data to structural data | |
| Vijayarani et al. | Multimedia mining research-an overview | |
| CN104317946A (en) | Multi-key image-based image content retrieval method | |
| CN106503112B (en) | Video retrieval method and device | |
| Kim et al. | TVDP: Translational visual data platform for smart cities | |
| Choi et al. | Multimodal location estimation of consumer media: Dealing with sparse training data | |
| MXPA04010520A (en) | Systems and methods for indexing and retrieving images. | |
| CN111968073B (en) | No-reference image quality evaluation method based on texture information statistics | |
| Jing et al. | Flood event image recognition via social media image and text analysis | |
| Zhang et al. | Automatic report generation based on multi-modal information | |
| CN103092935A (en) | Approximate copy image detection method based on scale invariant feature transform (SIFT) quantization | |
| Sun et al. | Design of a Media Resource Management System for Colleges Based on Cloud Service | |
| Xue et al. | ISD-SSD: image splicing detection by using modified single shot MultiBox detector | |
| CN105808723B (en) | The picture retrieval method hashed based on picture semantic and vision | |
| JP2008204007A (en) | Image dictionary generation method, device and program | |
| Rudrawar | Content based remote-sensing image retrieval with bag of visual words representation | |
| Harish Kumar et al. | Fingerprinting of Image Files Based on Metadata and Statistical Analysis | |
| Sebastine et al. | Semantic web for content based video retrieval |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20191203 Address after: 250101 2F, Hanyu Jingu new media building, high tech Zone, Jinan City, Shandong Province Patentee after: Renmin Zhongke (Shandong) Intelligent Technology Co.,Ltd. Address before: 100080 Zhongguancun East Road, Beijing, No. 95, No. Patentee before: Institute of Automation, Chinese Academy of Sciences |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200310 Address after: Room 201, 2 / F, Hanyu Jingu new media building, no.7000, Jingshi Road, Jinan City, Shandong Province, 250000 Patentee after: Renmin Zhongke (Jinan) Intelligent Technology Co.,Ltd. Address before: 250101 2F, Hanyu Jingu new media building, high tech Zone, Jinan City, Shandong Province Patentee before: Renmin Zhongke (Shandong) Intelligent Technology Co.,Ltd. |
|
| CP03 | Change of name, title or address | ||
| CP03 | Change of name, title or address |
Address after: 100176 1401, 14th floor, building 8, No. 8 courtyard, No. 1 KEGU street, Beijing Economic and Technological Development Zone, Daxing District, Beijing (Yizhuang group, high-end industrial area, Beijing Pilot Free Trade Zone) Patentee after: Renmin Zhongke (Beijing) Intelligent Technology Co.,Ltd. Address before: Room 201, 2 / F, Hangu Jinggu new media building, 7000 Jingshi Road, Jinan City, Shandong Province Patentee before: Renmin Zhongke (Jinan) Intelligent Technology Co.,Ltd. |