24 Region-Based Image Retrieval using Fuzzy Clustering

Similar documents
(VKIR) VKIR VKIR DCT (R) (G) (B) Ward DCT i

Sobel Canny i

IT,, i

Web Basic Web SAS-2 Web SAS-2 i

22 Google Trends Estimation of Stock Dealing Timing using Google Trends

Web Web Web Web i

,,,,., C Java,,.,,.,., ,,.,, i

2reN-A14.dvi

情報処理学会研究報告 IPSJ SIG Technical Report Vol.2013-CVIM-186 No /3/15 EMD 1,a) SIFT. SIFT Bag-of-keypoints. SIFT SIFT.. Earth Mover s Distance

..,,,, , ( ) 3.,., 3.,., 500, 233.,, 3,,.,, i

1 1 tf-idf tf-idf i

Web Web Web Web Web, i

17 Proposal of an Algorithm of Image Extraction and Research on Improvement of a Man-machine Interface of Food Intake Measuring System

Microsoft Word - toyoshima-deim2011.doc

, IT.,.,..,.. i

7,, i

1., 1 COOKPAD 2, Web.,,,,,,.,, [1]., 5.,, [2].,,.,.,, 5, [3].,,,.,, [4], 33,.,,.,,.. 2.,, 3.., 4., 5., ,. 1.,,., 2.,. 1,,

kut-paper-template2.dvi

DEIM Forum 2009 E

Studies of Foot Form for Footwear Design (Part 9) : Characteristics of the Foot Form of Young and Elder Women Based on their Sizes of Ball Joint Girth

paper.dvi

2 ( ) i

n 2 n (Dynamic Programming : DP) (Genetic Algorithm : GA) 2 i

28 Horizontal angle correction using straight line detection in an equirectangular image

Web Stamps 96 KJ Stamps Web Vol 8, No 1, 2004

1 Web [2] Web [3] [4] [5], [6] [7] [8] S.W. [9] 3. MeetingShelf Web MeetingShelf MeetingShelf (1) (2) (3) (4) (5) Web MeetingShelf

kut-paper-template.dvi

soturon.dvi

WebRTC P2P Web Proxy P2P Web Proxy WebRTC WebRTC Web, HTTP, WebRTC, P2P i

kut-paper-template.dvi

28 TCG SURF Card recognition using SURF in TCG play video

25 Removal of the fricative sounds that occur in the electronic stethoscope

IPSJ SIG Technical Report Vol.2011-MUS-91 No /7/ , 3 1 Design and Implementation on a System for Learning Songs by Presenting Musical St

WebRTC P2P,. Web,. WebRTC. WebRTC, P2P, i

SOM SOM(Self-Organizing Maps) SOM SOM SOM SOM SOM SOM i

20 Method for Recognizing Expression Considering Fuzzy Based on Optical Flow

(Visual Secret Sharing Scheme) VSSS VSSS 3 i

Virtual Window System Virtual Window System Virtual Window System Virtual Window System Virtual Window System Virtual Window System Social Networking

,,.,.,,.,.,.,.,,.,..,,,, i


3_23.dvi

258 5) GPS 1 GPS 6) GPS DP 7) 8) 10) GPS GPS ) GPS Global Positioning System

2 122

大学における原価計算教育の現状と課題


,,,,,,,,,,,,,,,,,,, 976%, i

日本看護管理学会誌15-2

4.1 % 7.5 %

Web ( ) [1] Web Shibboleth SSO Web SSO Web Web Shibboleth SAML IdP(Identity Provider) Web Web (SP:ServiceProvider) ( ) IdP Web Web MRA(Mail Retrieval

PC PDA SMTP/POP3 1 POP3 SMTP MUA MUA MUA i

塗装深み感の要因解析

THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS TECHNICAL REPORT OF IEICE.

25 fmri A study of discrimination of musical harmony using brain activity obtained by fmri

21 e-learning Development of Real-time Learner Detection System for e-learning

2 : Open Clip Art Library [4] Microsoft Office PowerPoint Web PowerPoint 2 Yahoo! Web [5] SlideShare Yahoo! Web Yahoo! Web

DEIM Forum 2009 C8-4 QA NTT QA QA QA 2 QA Abstract Questions Recomme

26 Development of Learning Support System for Fixation of Basketball Shoot Form

29 jjencode JavaScript

生活設計レジメ

44 4 I (1) ( ) (10 15 ) ( 17 ) ( 3 1 ) (2)

I II III 28 29


IPSJ SIG Technical Report Vol.2014-HCI-158 No /5/22 1,a) 2 2 3,b) Development of visualization technique expressing rainfall changing conditions

07_伊藤由香_様.indd

ID 3) 9 4) 5) ID 2 ID 2 ID 2 Bluetooth ID 2 SRCid1 DSTid2 2 id1 id2 ID SRC DST SRC 2 2 ID 2 2 QR 6) 8) 6) QR QR QR QR

(MIRU2008) HOG Histograms of Oriented Gradients (HOG)

25 D Effects of viewpoints of head mounted wearable 3D display on human task performance

paper.dvi

卒業論文2.dvi

& Vol.5 No (Oct. 2015) TV 1,2,a) , Augmented TV TV AR Augmented Reality 3DCG TV Estimation of TV Screen Position and Ro

Takens / / 1989/1/1 2009/9/ /1/1 2009/9/ /1/1 2009/9/30,,, i

21 A contents organization method for information sharing systems

ron.dvi

23 The Study of support narrowing down goods on electronic commerce sites

Study on Application of the cos a Method to Neutron Stress Measurement Toshihiko SASAKI*3 and Yukio HIROSE Department of Materials Science and Enginee

23 Study on Generation of Sudoku Problems with Fewer Clues

Web Web ID Web 16 Web Web i

08-特集04.indd

DEIM Forum 2010 A Web Abstract Classification Method for Revie

[2] OCR [3], [4] [5] [6] [4], [7] [8], [9] 1 [10] Fig. 1 Current arrangement and size of ruby. 2 Fig. 2 Typography combined with printing

21 Key Exchange method for portable terminal with direct input by user

, (GPS: Global Positioning Systemg),.,, (LBS: Local Based Services).. GPS,.,. RFID LAN,.,.,.,,,.,..,.,.,,, i

013858,繊維学会誌ファイバー1月/報文-02-古金谷

Vol.54 No (July 2013) [9] [10] [11] [12], [13] 1 Fig. 1 Flowchart of the proposed system. c 2013 Information

Q-Learning Support-Vector-Machine NIKKEI NET Infoseek MSN i

kiyo5_1-masuzawa.indd

Fig. 1. Example of characters superimposed on delivery slip.

24 Depth scaling of binocular stereopsis by observer s own movements


Web [1] [2] [3] [4] [5] SupportVectorMachine SVM [6] [7] Google [11] Web

29 Short-time prediction of time series data for binary option trade

(a) 1 (b) 3. Gilbert Pernicka[2] Treibitz Schechner[3] Narasimhan [4] Kim [5] Nayar [6] [7][8][9] 2. X X X [10] [11] L L t L s L = L t + L s

2017 (413812)

220 28;29) 30 35) 26;27) % 8.0% 9 36) 8) 14) 37) O O 13 2 E S % % 2 6 1fl 2fl 3fl 3 4

workshop_func_exp_ pptx

( )

知能と情報, Vol.30, No.5, pp

,,.,,., II,,,.,,.,.,,,.,,,.,, II i

) 2) , , ) 1 2 Q1 / Q2 Q Q4 /// Q5 Q6 3,4 Q7 5, Q8 HP Q9 Q10 13 Q11

BOK body of knowledge, BOK BOK BOK 1 CC2001 computing curricula 2001 [1] BOK IT BOK 2008 ITBOK [2] social infomatics SI BOK BOK BOK WikiBOK BO

Transcription:

24 Region-Based Image Retrieval using Fuzzy Clustering 1130323 2013 3 9

Visual-key Image Retrieval(VKIR) k-means Fuzzy C-means 2 200 2 2 20 VKIR 5 18% 54% 7 30 Fuzzy C-means i

Abstract Region-Based Image Retrieval using Fuzzy Clustering Yohei Ono Region-based image retrieval for example visual-key image retrieval is one of important techniques for image retrieval without using keywords. In visual-key image retrieval data clustering is important. Data clustering is used for dividing all sub-images to several clusters which contain similar sub-images. However conventional VKIR uses k-means or Ward method which are so called hard clustering In this thesis we apply fuzzy clustering algorithm to classify sub-images to clusters Using fuzzy clustering sub-images which are ambiguous to classify specific cluster can be classified to all clusters with membership values. Therefore users sense can be reflect to the result of clustering and the precision of VKIR improves. The experimental result shows that the precision improves to 11% and recall improves to 24% while those of conventional method are 18% and 54% respectively key words Image Retrieval System Clustering Visual-Key Fuzzy C-means Fuzzy Clustering ii

1 1 2 3 2.1 Content-Based Image Retrieval(CBIR).................. 3 2.2 Region-Based Image Retrieval(RBIR)................... 3 2.3.......................... 4 2.3.1...................... 5 2.3.2................... 6 2.3.3................... 7 2.4............................. 7 2.4.1......................... 7 2.4.2 k-means.............................. 8 2.4.3 Ward................................ 9 2.5 c-means( )................. 10 2.6....................... 11 3 12 3.1.............. 12 3.2............................. 13 3.3............... 14 4 16 4.1................................... 16 4.2....................... 18 4.3................................... 18 iii

4.4...................................... 19 5 21 22 24 A 5 25 B 5 29 iv

1 Web Web Web PC Content-Based Image Retrieval(CBIR) Region-Based Image Retrieval(RBIR) RBIR Visual-key Image Retrieval(VKIR) k-means Ward VKIR Fuzzy C-means 2 200 2 1

2 20 VKIR 5 18% 54% 7 30 2

2 CBIR RBIR 2.1 Content-Based Image Retrieval(CBIR) Text-Based Image Retrieval(TBIR) Content based Image Retrieval(CBIR) CBIR 2.2 Region-Based Image Retrieval(RBIR) CBIR RBIR(Region-Based Image Retrieval) RBIR RBIR 3

2.3 RBIR 2.3 1 4

2.3 クラスタ 1 クラスタ 2 2.1 2.3.1 4 1. 2. 3. 4. 1 4 2.2 5

2.3 画像の分割 代表的な画像をビジュアルキーとして選定 類似している特徴ベクトル同士をクラスタ分別 Visual key1 Visual key2 2.2 2.3.2 1. 2. 3. 4. (2 2 ) 6

2.4 2.3.3 2.4 2.4.1 2 1. [ ] 2. [ ] 7

2.4 Ward 2.4.2 k-means k-means (k ) k 2 k-means k k k-means 8

2.4 2.4.3 Ward Ward Ward 2.3 9

2.5 c-means( ) 2.5 c-means( ) ( ) Fuzzy C-means(FCM) cluster fuzzy value 1 0.1 5 0.17 10 0.21 total 1 20 0.42 2.4 10

2.6 2.6 k-mean Ward 0 1 ハードクラスタリング ファジィクラスタリング クラスタ1 クラスタ2 クラスタ3 クラスタ1 1 クラスタ1 0.5 クラスタ2 0 クラスタ2 0.5 クラスタ 1 0 クラスタ 2 1 2.5 11

3 3.1 id ( ) fvalue class 200 img 20% 12

3.2 3.1 3.2 Web Apache SQLite PHP Web Web SQLite Web Web 13

3.3 ユーザ (Web ブラウザ ) Web サーバ (Apache) 画像情報の通信 SQL 文インデックスの通信 データベース (Sqlite) PHP 画像情報の取得 ファイル 画像インデックス 画像データベース 3.2 3.3 3.3 14

3.3 3.4 15

4 DCT 4.1 ArtExplosion 10 (Agriculture Ballet Castles Flowers Food Landscape Panoramic Texas Water Wilderness) 001 020 20 200 1 10 16

4.1 4.1 4.2 20 17

4.2 200 5 = = (4.1) (4.2) (4.3) 4.2 200 2 2 20 M.Serata k-means 4.3 11% 18% 24% 54% 18

4.4 0.6 検索精度の比較 0.5 0.4 0.3 0.2 平均適合率 平均再現率 0.1 0 k-means( 従来 ) Fuzzy C-means( 提案 ) 4.3 4.4 k-means Fuzzy C-means 11% 18% 7% 24% 54% 30% k mean 7% 30% Fuzzy C-means 0.2 0.2 19

4.4 0.2 5 20

5 Region-Based Image Retrieval(RBIR) VKIR k-means Fuzzy C-means k-means 11% 24% Fuzzy C-means 18% 54% 7% 30% 30% 20% 21

4 3 22

23

[1] 1999 10 30 [2] M.Serata Y.Hatakeyama and K.Hirota Designing Image Retrieval System with the Concept of Visual Keys, Journal of Advanced Computational Intelligence and Intelligent Informatics vol. 10 no. 2 pp 136-144 2006 [3] K.Okamoto and S.Yoshida. DCT Domain Color Feature Extraction for Visual Key Image Retrieval, 22 pp.322-326 2007. 24

A 5 0.5 0.45 0.4 0.35 適 0.3 合 0.25 率 0.2 0.15 0.1 0.05 0 被験者 A 1 2 3 4 5 6 7 8 9 10 画像番号 A.1 A A 5 6 10% 25

0.6 被験者 B 0.5 0.4 適合 0.3 率 0.2 0.1 0.0 1 2 3 4 5 6 7 8 9 10 画像番号 A.2 B B 6 50% 10% 0.5 被験者 C 0.4 適合率 0.3 0.2 0.1 0.0 1 2 3 4 5 6 7 8 9 10 画像番号 A.3 C C 5 6 26

0.6 被験者 D 0.5 適合率 0.4 0.3 0.2 0.1 0.0 1 2 3 4 5 6 7 8 9 10 画像番号 A.4 D D 3 5 6 10 D 適合率 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0.0 被験者 E 1 2 3 4 5 6 7 8 9 10 画像番号 A.5 E E 5 6 27

5 5 6 28

B 5 再現率 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 被験者 A 1 2 3 4 5 6 7 8 9 10 画像番号 B.1 A A 10 7 50% 1 29

再現率 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 被験者 B 1 2 3 4 5 6 7 8 9 10 画像番号 B.2 B B 3 6 80% 30% 60% 1.0 被験者 C 0.8 再現率 0.6 0.4 0.2 0.0 1 2 3 4 5 6 7 8 9 10 画像番号 B.3 C C 7 8 20% 30

再現率 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 被験者 D 1 2 3 4 5 6 7 8 9 10 画像番号 B.4 D D 3 6 10 7 15% 再現率 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 被験者 E 1 2 3 4 5 6 7 8 9 10 画像番号 B.5 E E 3 6 31

5 3 6 32