Searching Large Image Databases using Color Information

Size: px
Start display at page:

Download "Searching Large Image Databases using Color Information"

Transcription

1 Searchng Large Image Databases usng Color Informaton Ioan Racu CMSC350: Artfcal Intellgence Wnter Quarter 2004 Department of Computer Scence Unversty of Chcago ABSTRACT The goal of ths project s to mplement an exploraton system for large mage databases n order to help the user fnd smlar mages to a query mage n a very effcent method. The methods proposed n ths system should be scalable to the largest databases currently n use; for example, has an mage search database va keywords search of about 1,000,000,000 mages. Whle the proposed system works wth any type of low-level feature representaton of mages, we mplement our system usng color nformaton. The system s bult n three stages: 1) the feature extracton stage n whch mages are represented n a way that allows effcent storage and retreval results closer to the human percepton; 2) the second stage conssts of clusterng the mage database va k-means clusterng n whch the clustrod would allow quck human comprehenson of the type of mages wthn the correspondng cluster; 3) the thrd stage s the vsualzaton stage n whch the results are dsplayed to the user. We evaluated the performance of our system based on the retreval accuracy and on the perceptual smlarty order among retreved mages. Experments usng a general purpose database of 2100 mage were used to show the practcalty of the proposed system n both accuracy and tme complexty. Two features could be added to the current system as future work: 1) buld a herarchy of clusters whch would allow even faster retreval results; 2) mplement mult-dmensonal scalng (MDS) technque to provde a tool for the vsualzaton of the database at dfferent levels of detals. Keywords: mage database, mage retreval, k-means clusterng, color nformaton, HSV 1.0 INTRODUCTION The deas of ths project, buldng an mage retreval system, was nspred from the fact that very few, f any commercal systems exst that allow users to query an mage database va mages rather than keywords. One commercal mage retreval system mplemented by Google.com s shown n Fgure 1. Google.com crawls the Internet and collects mages whch t then stores n a database. For each pcture collected, the crawler also collect textual nformaton whch surrounds the pcture and whch mght even be encoded n the pcture tself (.e. the fle name, drectory name, Page 1 of 24

2 etc ), and then creates a dctonary that assocates keywords wth mages. From ths pont, once the mages have a set of keywords that represent t, the problem gets reduced to a text retreval system whch Google.com has already mplemented and was successful n dong so. Fgure 1: Exstng mage retreval system va keyword queres mplemented by Google.com; left fgure: sunset Note n Fgure 1 the query of the keyword sunset resulted n many results from whch most of them really look lke sunsets; t s nterestng to see that the word sunset appears n each and everyone of the pctures names, so unless a pcture was mstakenly labeled by the owner of the pcture, the retreval of the system based on such keywords would yeld very good results! The mage retreval system s broken up nto two separate phases. The entre frst phase s depcted n Fgure 2. The frst phase can be consdered the tranng part of the system, n whch the raw mage database s massaged nto an mage feature database, whch s later used as the nput to k-means clusterng n order to buld a 1 level herarchy (expandable to more levels for larger databases to mprove performance) of mages organzed by clusters. The leaf cluster contans a group of smlar mages (based on the dstance metrc used, n our case color nformaton), that wll be used n the Page 2 of 24

3 retreval phase as the set of mages most smlar to the query mage. The fnal part of the tranng phase s vsualzng the mage database by dsplayng the clustrods of the database. Stage 1 Stage 2 Stage 3 Clustrod 1: Level 1 Image Database: Image Database: Raw Images Image Features Clusterng Leaf: Images Level 2 Clustrod : Level 1 Vsualzaton Clustrod sqrt(n): Level 1 Fgure 2: Image retreval system overvew: tranng phase The second phase can be depcted n Fgure 3. It conssts of an nterface from whch a user supples a query mage, after whch the mage gets decomposed nto ts representatve mage features. Based on these mage features, the query mage s compared aganst the clustrods, and once the most smlar cluster s found, the query mage s compared to each of the mages n the leaf cluster, and a dstance vector s calculated from the query mage to the rest of the mages n the leaf cluster. The mages n the leaf cluster are ranked accordng to ther dstance from the query mage, and dsplayed n rank order. Page 3 of 24

4 Stage 1 Stage 2 Stage 3 Query Image: Raw Image Query Image: Image Features Clustrod 1: Level 1 Clustrod : Level Clustrod sqrt(n): Level 1 Leaf: Images Level 2 Vsualzaton Fgure 3: Image retreval system overvew: retreval phase 2.0 RELATED WORK Ths secton wll cover related work regardng mage retreval systems, other mplementatons, and the state of the art of today s mage retreval systems. The book Algorthms for Clusterng Data [3] s an excellent source for clusterng technques. There are a few earler research efforts on mage databases exploraton. The system descrbed n Photobook [4] descrbed a content based manpulaton approach for mage databases. The QBIC [5] system was developed at IBM and allows queres by mage and vdeo content. The system n MARS [6] addresses supportng smlarty queres. The system descrbed n [7] s based on a new data structure bult on the mult-lnearzaton of mage attrbutes for effcent organzaton and fast vsual browsng of the mages. The systems proposed n [8]-[9] are based on Herarchcal Self-Organzng Map and are used to organze a complete mage database nto a 2-D grd. In [10]-[11] Mult-Dmensonal Scalng s used to organze mages returned n response to a query and for drect organzaton of a complete database. Actve browsng s proposed n [12] by ntegratng relevance feedback nto the browsng envronment and thus, users can modfy the database organzaton to sut a desred task. Page 4 of 24

5 3.0 IMPLEMENTATION I mplemented the mage retreval system n C++ under Mcrosoft Wndows XP n Vsual Studo.NET. Due to poor support for mage lbrares n C++, I wll be usng a thrd party API called Magck++ [1]. My mplementaton had the followng major components, all mplemented by me, unless otherwse stated: Feature extracton o o o o Readng of varous fle formats from dsk (Magck++) Image segmentaton RGB to HSV converson HSV hstograms Smlarty Metrc K-means clusterng Vsualzaton: buldng the results n rank order o o Wrtng results nto mages on dsk (Magck++) Dsplayng the results to the user n a graphcal nterface (Magck++) In total, I ended up wth about 2500 lnes of source code, and a very effcent and automated CBIR system that could search for mages n relatvely large mage databases. A screen shot of the CBIR system, whch outlnes all the above components can be seen n the fgure below. Fgure 4: Screen shot of the mplemented CBIR system Page 5 of 24

6 3.1 Feature Extracton We used color as the low-level feature representaton n the mages of the database. Because the HSV color model s closely related to the human vsual percepton, we decded to convert all mages from the RGB to the HSV color space. We encode the spatal dstrbuton of features n mages va a fxed parttonng scheme; each mage dvded nto M N overlappng blocks, for whch 3 separate local hstograms (H, S, and V) are calculated for every block. Ths provdes a compact and effcent storage and retreval soluton to an otherwse very hard problem. The locaton of area-peak for every local hstogram determnes the value of the correspondng hstogram. Fgure 5 depcts the process of feature extracton for 1 partcular block; the pxel values (hue, saturaton, or value) are taken and a hstogram s computed, from whch the maxmum s taken; ths maxmum becomes the representatve value for that block for the correspondng component Block Hstogram [ K K ] e11e e e K m1 m2 e mn Occurence Pxel value Fgure 5: Depcts the process of feature extracton for 1 block Table 1 depcts 2 examples of mages decomposed nto ther feature vector and what they look lke when converted back nto ther RGB format. Page 6 of 24

7 Table 1: Examples of mages decomposed nto ther feature vector and what they look lke when converted back nto ther RGB format HSV Color Space It has been verfed expermentally that color s perceved through three ndependent color receptors whch have peak responses at approxmately, red, green and blue wavelengths and thus, t can be represented by a lnear combnaton of the three prmary colors (R, G, B) (Fgure 6). Fgure 6: Mxture of lght (Addtve prmares) Page 7 of 24

8 The characterstcs used to dstngush one color from another are: Hue s an attrbute assocate wth the domnant wavelength n a mxture of lght waves. It represents the domnant color as perceved by observer (for example, orange, red, pnk, etc) Saturaton refers to the relatve purty or the amount of whte lght mxed wth a hue. Pure colors are fully saturated. Colors such as pnk (red and whte) are less saturated wth the saturaton beng nversely proportoned to the amount of whte lght added. Brghtness (or value) embodes the chromatc noton of ntensty. Hue and saturaton together are called chromatcty. A color can be descrbed n terms of ts chromatcty and brghtness. The formulae from Fgure 7 represent the transformaton from RGB to HSV space. Snce the RGB space s not perceptually unform (the proxmty of colors n ths space does not ndcate the color smlarty), ths space s used as a startng pont for generatng other color spaces havng the propertes above. Snce RGB s a complete space, the generated color spaces are all complete even f the transformaton from one space to another s not necessarly lnear (such as n the case of the HSV space). Of the all spaces, only the HSV quantzed color space satsfes the propertes of unformty, compactness, completeness, and naturalness that s needed for workng wth mages. Fgure 7: The relatonshps between prmary colors and hue, saturaton, and brghtness components Page 8 of 24

9 3.2 Smlarty metrc Clusterng methods requre that an ndex of proxmty or assocatons be establshed between pars of patterns; a proxmty ndex s ether a smlarty or dssmlarty. Snce our retreval system s desgned to retreve the most smlar mages wth a query mage, the proxmty ndex wll be defned wth respect to smlarty. Dfferent smlarty measures have been suggested n the lterature to compare mages [14, 13]. Let q and t represent the block number n two mages Q and T, respectvely. Let ( h,s, v ) and ( h,s, v ) represent the domnant hue-saturaton par of the selected block n the two mages Q and T. The block smlarty s defned by the followng relatonshp [15]: q q q t t t S ( q,t ) = 1+ a Dh ( h,h ) + b* D ( s,s ) + c D ( v,v ) q t 1 s q t v q t Here D h, Ds and Dv represent the functons that measure smlarty n hue, saturaton and value. The constants a, b and c defne the relatve mportance of hue, saturaton and value n smlarty components. Snce human percepton s more senstve to hue, a hgher value s assgned to a than to b. The followng functon [15] was used to calculate D h : The functon D h k 1 π ( h,h ) = 1 cos h h q t 256 q Dh explctly takes nto account the fact that hue s measured as an angle. Through emprcal evaluatons, t 2 when k=2 provdes a good non-lnearty n the smlarty measure to approxmate the subjectve judgment of the hue smlarty. The saturaton smlarty [15] s calculated by: D s ( s,s ) q t = s q s 256 t Page 9 of 24

10 The value smlarty s calculated by usng the same formula as for saturaton smlarty. The range for D s and D v s [ 0..1] snce the saturaton and value smlarty measures are normalzed by dvdng wth ther maxmum values; snce the cos functon belongs to [ 1..1], D h belongs to [ 0..1]. Takng nto account the ranges for D h, Ds and D v, the range for the functon S that measures the smlarty between two mage blocks s a + b + c. In other words, a smlarty value of 1 means that two mages are dentcal, values close to 1 ndcate smlar mages, whle values close to 0 ndcate dssmlar mages. Note that the smlarty metrc can never reach 0 because the lower bound s 1 over a fnte number; however, ths does not pose any lmtatons on our approach snce we are nterested n measurng the smlarty between mages and therefore we are lookng for smlarty values close to 1. Usng the smlartes between the correspondng blocks from mage Q and mage T, the smlarty between two mages s calculated by the followng expresson [15]: S ( Q,T ) = M N m S = 1 M N m = 1 ( q,t ) The quantty blocks. m n the above expresson represents the maskng bt for block and M N stands for the number of Page 10 of 24

11 4.0 EMPIRICAL PERFORMANCE RESULTS I used an mage database of 2100 mages to test my proof of concept, but a larger database would prove the scalablty of the system emprcally. A real nce database that I found s the Corel Stock Photos [2]; t conssts of over 39,000 general purpose mages; the bggest dsadvantage to ths mage database s that t s prohbtvely expensve, and hence I cannot use t. 4.1 Performance Results From the performance results gven n the next few tables and fgures, t can be seen that clusterng s a very good optmzaton from the performance pont of vew, and that the mplemented system s very scalable. It should be noted that the entre search tme n the entre database s about 3 mllseconds wth the 1 level of clusterng, whle t s about 80 mllseconds wth no clusterng. The reason t can search so fast through the entre database even wthout any clusterng s because of the effcent representaton of the mages by the feature vectors wth only 192 components for each mage! The pre-processng stage takes a lttle over 1 second per mage, but when searchng through the database, there s normally only 1 mage that s the query, and thus ths does not hnder the scalablty of the system. Clusters 1 37 Query Space Feature Space Tme to Cluster (seconds) Tme to search (seconds) Tme to extract features n 1 mage (seconds) Compare mages per second Table 2: System tmng data comparng the clusterng approach wth no clusterng at all The next table shows the amount of off-lne tme the system needs to prep tself before queres can be made to the system. To compute the feature vectors and computng the dstance matrx from scratch, t would take about 30 mnutes for about 1400 mages, but snce ths s an off-lne process, ths s OK. Once these are computed, the data can be loaded n memory n less than 10 seconds for the entre mage database. Tme Readng Features Vectors (seconds) Computng Features Vectors (seconds) Readng Dstance Matrx (seconds) Computng Dstance Matrx (seconds) Page 11 of 24

12 Table 3: System Performance tmng data for loadng/computng the off-lne nformaton The next two fgures show how scalable our system s n comparson to a system that would search through the entre database. For example, has about 1,000,000,000 mages n ther database. Even wth ths large of a database, the search (query) space would be less than 100,000 mages whch can be done on the order of a few seconds. If we mprove the system by havng the herarchy of clusters, we can reduce the search space to only a few hundred mages, and hence retreve results n a matter of mllseconds! Query Space Query Space (Number of Images) E+07 1E+08 1E+09 Database Sze (Number of Images) QS - NO C QS - 1 C QS - H sq C QS - H log C Fgure 8: Query Space In Fgure 8 and Fgure 9, QS/QT - NO C represents no clusterng, QS/QT 1 C represents 1 level of clusterng, QS/QT H sq C represents herarchcal clusterng wth square root of n clusters, where n s the number of mages n the level above, and QS/QT H log C represents herarchcal clusterng wth logarthmc number of clusters n relaton to the level above. The notable feature s that even wth very large databases (hundreds of mllons of mages), the CBIR system could stll retreve the most relevant mages n less than a second, whch s actually qute amazng when thnkng of the sze of the databases we are tryng to search. Page 12 of 24

13 Query Tme Query Tme (Seconds) E+08 Database Sze (Number of Images) QT - NO C QT - 1 C QT - H sq C QT - H log C Fgure 9: Query Tme Database Sze Mem Req (MB) QS - NO C QS - 1 C QS - H sq C QS - H log C Table 4: Memory requrements (n megabytes) for the query space In the above table, we see that all three clusterng mplementatons would have a very small memory requrement, needng only on the order of dozens of MB whch can easly be done even on commodty workstatons. Lastly, we calculated the performance of the system by measurng some statstcs regardng the accuracy of the results. Table 5 depcts these fndngs. We used 1383 mages for the mage database, and we used the remanng 718 mages to test the system wth. These 718 mage were chosen randomly from the entre 2100 mages whch makes up the database. For each query, we retreved the 4 most smlar mages, and rank them by smlarty. The frst result ndcates Page 13 of 24

14 that all retreved mages were color smlar to the query mage. The second result shows that the semantc meanng mpled by the query mage was preserved across 78% of the retreved mages. The thrd results shows that the most smlar mage to the query mage retreved was 91% of the tmes semantcally smlar to the query mage. The last result shows that n all query tests done, at least 1 of the 4 mages retreved was semantcally smlar to the query mage. Descrpton Percentage % all retreved color smlar 100% % all retreved semantc smlar 78% % hghest rank semantc smlar 91% % at least 1 semantc smlar 100% 4.2 Sample Query Results Table 5: Query retreval performance results Table 6 through Table 10 show some sample queres performed on the mage database. All the rest of the queres (almost 700 more queres) can be found onlne at my web ste at The clustrods of the database can be depcted n Fgure 10. Note the varyng colors of the clustrods, rangng from black to varous shades of blue, green, orange, red, brown, yellow, and whte; t vrtually covers the entre color space. Ths s the result of a very good smlarty metrc and the success of the clusterng technque used! The retreval accuracy s very good, gven that the amount of nformaton present n the feature vectors s very small n comparson to the orgnal mages themselves. Page 14 of 24

15 Fgure 10: Clustrods of the mage database Page 15 of 24

16 Table 6: 5 sample queres; the query mage s the leftmost mage, and the value depcted above the mage name denotes the smlarty between the partcular mage and the query mage (1 most smlar and 0 least smlar) Page 16 of 24

17 Table 7: 5 sample queres; the query mage s the leftmost mage, and the value depcted above the mage name denotes the smlarty between the partcular mage and the query mage (1 most smlar and 0 least smlar) Page 17 of 24

18 Table 8: 5 sample queres; the query mage s the leftmost mage, and the value depcted above the mage name denotes the smlarty between the partcular mage and the query mage (1 most smlar and 0 least smlar) Page 18 of 24

19 Table 9: 5 sample queres; the query mage s the leftmost mage, and the value depcted above the mage name denotes the smlarty between the partcular mage and the query mage (1 most smlar and 0 least smlar) Page 19 of 24

20 Table 10: 5 sample queres; the query mage s the leftmost mage, and the value depcted above the mage name denotes the smlarty between the partcular mage and the query mage (1 most smlar and 0 least smlar) Page 20 of 24

21 5.0 CONCLUSION AND FUTURE WORK In concluson, we demonstrated that a CBIR system can be bult effcently that works on real world general mages very well. As the number of dgtal mages ncreases n our daly lves, the need for some systems wll grow and eventually, these systems wll become as common as text search engnes have become n the past 10 years! Good CBIR systems can eventually be used n many domans that currently rely on hand annotatng and hand analyzng large number of mages. As future work, the current system could be extended relatvely smple wthout any new sgnfcant theoretcal work to support a search tme complexty that s logarthmc n relaton to the sze of the database usng herarchcal clusterng. In order to support very large databases (on the order of hundreds of mllons of mages), better vsualzaton tools would be needed as well even to nterpret the large amounts of results. One effcent method that s sutable n ths context s multdmentonal scalng (MDS) whch can be mplemented n the vsualzaton stage of our system to vsualze the results n a 2-dmentonal space as the mages relate to each other va ther smlarty metrc. Stage 1 Stage 2 Stage 3 Image Database: Image Database: Raw Images Image Features Clusterng Clustrod: Level 1 Clustrod: Level 1 Clustrod: Level Clustrod: Level 2... Leaf: Images Level n MDS Fgure 11: Image retreval system overvew: tranng phase The new enhanced mage retreval system s depcted n Fgure 2. The frst phase s dentcal to our system; we use the mage feature vectors as the nput to k-means clusterng n order to buld a herarchy of mages organzed by clusters. The leaf cluster contans a group of smlar mages (based on the dstance metrc used, n our case color nformaton), that wll be used n the retreval phase as the set of mages most smlar to the query mage. The fnal part of the Page 21 of 24

22 tranng phase s the mult-dmensonal scalng (MDS) technque that wll be used to vsualze the mage database at varyng levels of detal. The second phase can be depcted n Fgure 3. It conssts of an nterface from whch a user supples a query mage, after whch the mage gets decomposed nto ts representatve mage features. Based on these mage features, the query mage s compared aganst the clustrods at the frst level of the herarchy, and the search contnues deeper nto the herarchy only along the best path, based on the dstance metrc used. The query mage s compared at each level wth the clustrods, and eventually, the query mage wll arrve at a leaf cluster. The query mage s compared to each of the mages n the leaf cluster, and a dstance vector s calculated from the query mage to the rest of the mages n the leaf cluster. The dstance vector s then the nput to the MDS module, whch graphcally wll dsplay the results n a 2- dmentonal space wth the most smlar mages close to the query mage and the most deferent mages far away from the query mage Fgure 12: Image retreval system overvew: retreval phase From a performance pont of vew (gven that I can get a larger mage database), t would be nterestng to mplement the herarchcal clusterng, whch should be relatvely smple to modfy the current code to support the mult level herarchy rather than the sngle level I currently used. From a utlty pont of vew, varous smlarty metrcs should be expermented wth for the color feature vectors the system extracts. Also, modfyng the feature vectors to contan other nformaton such as texture nformaton, shape nformaton, and annotated text would help n augmentng the results n Page 22 of 24

23 the system and create a complete end-to-end soluton for a CBIR system that would work excellent and effcent at the same tme! More query results, electronc verson of ths document, the presentaton on ths project, and source code can be found at my web ste at For more nformaton, comments, or suggestons, I can be reached at racu@cs.uchcago.edu. 6.0 REFERENCES [1] [2] [3] A. K. Jan and R. C. Dubes, Algorthms for Clusterng Data, Prentce Hall Advanced Reference Seres [4] A. Pentland, R. W. Pcard, S. Sclaroff, Photobook: Content Based Manpulaton of mage databases, n Multmeda Tools and Applcatons, edtor Burko Furht, Kluwer Academc Publsher, Boston, pp , [5] Flckner, M., and et al., Query by Image and Vdeo Content: The QBIC System, IEEE Computer, [6] M. Ortega, Y. Ru, K. Chakrabart, S. Mehrotra, and T. S. Huang, Supportng smlarty queres n MARS," ACM Conf. on Multmeda, [7] S. Craver, B.-L. Yeo, and M.M. Yeung. Image browsng usng data structure based on multple space-fllng curves. Proceedngs of the Thrty-sx Aslomar Conference on Sgnals, Systems, and Computers, Nov. 1-4, 1998 [8] I.K. Seth and I. Coman. Image retreval usng herarchcal self-organzng feature maps. Pattern Recognton Letters 20, pp , 1999 [9] H. Zhang and D. Zhong. A scheme for vsual feature based mage ndexng. Proceedngs of SPIE/IS&T Conf. on Storage and Retreval for Image and Vdeo Databases III, vol. 2420, pp , 1995 [10] J. MacCush, A. McPherson, J. Barros, and P. Kelly. Interactve layout mechansms for mage database retreval. Proceedngs of SPIE/IS&T Conf. on Vsual Data Exploraton and Analyss, III, vol.2656, pp , 1996 [11] Y. Rubner, L. Gubas, and C. Tomas. The earth s mover dstance, mult-dmensonal scalng, and color-based mage retreval. Proceedngs of the ARPA Image Understandng Workshop, May 1997 Page 23 of 24

24 [12] J.-Y. Chen, C.A. Bouman, and J.C. Dalton. Smlarty pyramds for browsng and organzaton of large mage databases. Human Vson and Electronc Imagng III, pp , 1998 [13] Faloutsos C. et al. (1993). Effcent and effectve queryng by mage content, Techncal Report, IBM. [14] Swan M.J. and Ballard D.H. (1991). Color Indexng, Internatonal Journal of Computer Vson, 7(1), pp [15] Seth I. K., Coman I., Day B. et al. (1998). Color-WISE: A system for mage smlarty retreval usng color, Proceedngs SPIE: Storage and Retreval for Image and Vdeo Databases, 3132, pp Page 24 of 24

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data A Fast Content-Based Multmeda Retreval Technque Usng Compressed Data Borko Furht and Pornvt Saksobhavvat NSF Multmeda Laboratory Florda Atlantc Unversty, Boca Raton, Florda 3343 ABSTRACT In ths paper,

More information

1. Introduction. Abstract

1. Introduction. Abstract Image Retreval Usng a Herarchy of Clusters Danela Stan & Ishwar K. Seth Intellgent Informaton Engneerng Laboratory, Department of Computer Scence & Engneerng, Oaland Unversty, Rochester, Mchgan 48309-4478

More information

Content Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers

Content Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers IOSR Journal of Electroncs and Communcaton Engneerng (IOSR-JECE) e-issn: 78-834,p- ISSN: 78-8735.Volume 9, Issue, Ver. IV (Mar - Apr. 04), PP 0-07 Content Based Image Retreval Usng -D Dscrete Wavelet wth

More information

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps Vsual Thesaurus for Color Image Retreval usng Self-Organzng Maps Chrstopher C. Yang and Mlo K. Yp Department of System Engneerng and Engneerng Management The Chnese Unversty of Hong Kong, Hong Kong ABSTRACT

More information

A Binarization Algorithm specialized on Document Images and Photos

A Binarization Algorithm specialized on Document Images and Photos A Bnarzaton Algorthm specalzed on Document mages and Photos Ergna Kavalleratou Dept. of nformaton and Communcaton Systems Engneerng Unversty of the Aegean kavalleratou@aegean.gr Abstract n ths paper, a

More information

Fuzzy C-Means Initialized by Fixed Threshold Clustering for Improving Image Retrieval

Fuzzy C-Means Initialized by Fixed Threshold Clustering for Improving Image Retrieval Fuzzy -Means Intalzed by Fxed Threshold lusterng for Improvng Image Retreval NAWARA HANSIRI, SIRIPORN SUPRATID,HOM KIMPAN 3 Faculty of Informaton Technology Rangst Unversty Muang-Ake, Paholyotn Road, Patumtan,

More information

A Clustering Algorithm for Key Frame Extraction Based on Density Peak

A Clustering Algorithm for Key Frame Extraction Based on Density Peak Journal of Computer and Communcatons, 2018, 6, 118-128 http://www.scrp.org/ournal/cc ISSN Onlne: 2327-5227 ISSN Prnt: 2327-5219 A Clusterng Algorthm for Key Frame Extracton Based on Densty Peak Hong Zhao

More information

FEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur

FEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur FEATURE EXTRACTION Dr. K.Vjayarekha Assocate Dean School of Electrcal and Electroncs Engneerng SASTRA Unversty, Thanjavur613 41 Jont Intatve of IITs and IISc Funded by MHRD Page 1 of 8 Table of Contents

More information

A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION

A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION 1 THE PUBLISHING HOUSE PROCEEDINGS OF THE ROMANIAN ACADEMY, Seres A, OF THE ROMANIAN ACADEMY Volume 4, Number 2/2003, pp.000-000 A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION Tudor BARBU Insttute

More information

Cluster Analysis of Electrical Behavior

Cluster Analysis of Electrical Behavior Journal of Computer and Communcatons, 205, 3, 88-93 Publshed Onlne May 205 n ScRes. http://www.scrp.org/ournal/cc http://dx.do.org/0.4236/cc.205.350 Cluster Analyss of Electrcal Behavor Ln Lu Ln Lu, School

More information

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices Steps for Computng the Dssmlarty, Entropy, Herfndahl-Hrschman and Accessblty (Gravty wth Competton) Indces I. Dssmlarty Index Measurement: The followng formula can be used to measure the evenness between

More information

Background Removal in Image indexing and Retrieval

Background Removal in Image indexing and Retrieval Background Removal n Image ndexng and Retreval Y Lu and Hong Guo Department of Electrcal and Computer Engneerng The Unversty of Mchgan-Dearborn Dearborn Mchgan 4818-1491, U.S.A. Voce: 313-593-508, Fax:

More information

Object-Based Techniques for Image Retrieval

Object-Based Techniques for Image Retrieval 54 Zhang, Gao, & Luo Chapter VII Object-Based Technques for Image Retreval Y. J. Zhang, Tsnghua Unversty, Chna Y. Y. Gao, Tsnghua Unversty, Chna Y. Luo, Tsnghua Unversty, Chna ABSTRACT To overcome the

More information

TN348: Openlab Module - Colocalization

TN348: Openlab Module - Colocalization TN348: Openlab Module - Colocalzaton Topc The Colocalzaton module provdes the faclty to vsualze and quantfy colocalzaton between pars of mages. The Colocalzaton wndow contans a prevew of the two mages

More information

Detection of an Object by using Principal Component Analysis

Detection of an Object by using Principal Component Analysis Detecton of an Object by usng Prncpal Component Analyss 1. G. Nagaven, 2. Dr. T. Sreenvasulu Reddy 1. M.Tech, Department of EEE, SVUCE, Trupath, Inda. 2. Assoc. Professor, Department of ECE, SVUCE, Trupath,

More information

An Optimal Algorithm for Prufer Codes *

An Optimal Algorithm for Prufer Codes * J. Software Engneerng & Applcatons, 2009, 2: 111-115 do:10.4236/jsea.2009.22016 Publshed Onlne July 2009 (www.scrp.org/journal/jsea) An Optmal Algorthm for Prufer Codes * Xaodong Wang 1, 2, Le Wang 3,

More information

MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION

MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION Paulo Quntlano 1 & Antono Santa-Rosa 1 Federal Polce Department, Brasla, Brazl. E-mals: quntlano.pqs@dpf.gov.br and

More information

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance Tsnghua Unversty at TAC 2009: Summarzng Mult-documents by Informaton Dstance Chong Long, Mnle Huang, Xaoyan Zhu State Key Laboratory of Intellgent Technology and Systems, Tsnghua Natonal Laboratory for

More information

A Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems

A Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems A Unfed Framework for Semantcs and Feature Based Relevance Feedback n Image Retreval Systems Ye Lu *, Chunhu Hu 2, Xngquan Zhu 3*, HongJang Zhang 2, Qang Yang * School of Computng Scence Smon Fraser Unversty

More information

Parallelism for Nested Loops with Non-uniform and Flow Dependences

Parallelism for Nested Loops with Non-uniform and Flow Dependences Parallelsm for Nested Loops wth Non-unform and Flow Dependences Sam-Jn Jeong Dept. of Informaton & Communcaton Engneerng, Cheonan Unversty, 5, Anseo-dong, Cheonan, Chungnam, 330-80, Korea. seong@cheonan.ac.kr

More information

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching A Fast Vsual Trackng Algorthm Based on Crcle Pxels Matchng Zhqang Hou hou_zhq@sohu.com Chongzhao Han czhan@mal.xjtu.edu.cn Ln Zheng Abstract: A fast vsual trackng algorthm based on crcle pxels matchng

More information

Machine Learning: Algorithms and Applications

Machine Learning: Algorithms and Applications 14/05/1 Machne Learnng: Algorthms and Applcatons Florano Zn Free Unversty of Bozen-Bolzano Faculty of Computer Scence Academc Year 011-01 Lecture 10: 14 May 01 Unsupervsed Learnng cont Sldes courtesy of

More information

K-means and Hierarchical Clustering

K-means and Hierarchical Clustering Note to other teachers and users of these sldes. Andrew would be delghted f you found ths source materal useful n gvng your own lectures. Feel free to use these sldes verbatm, or to modfy them to ft your

More information

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes SPH3UW Unt 7.3 Sphercal Concave Mrrors Page 1 of 1 Notes Physcs Tool box Concave Mrror If the reflectng surface takes place on the nner surface of the sphercal shape so that the centre of the mrror bulges

More information

Classifier Selection Based on Data Complexity Measures *

Classifier Selection Based on Data Complexity Measures * Classfer Selecton Based on Data Complexty Measures * Edth Hernández-Reyes, J.A. Carrasco-Ochoa, and J.Fco. Martínez-Trndad Natonal Insttute for Astrophyscs, Optcs and Electroncs, Lus Enrque Erro No.1 Sta.

More information

Photo management applications

Photo management applications Techncal Note PR-TN 7/698 Issued: /7 Photo management applcatons M.A. Peters; P.M.F. Fonseca Phlps Research Europe PR-TN 7/698 Authors address M.A. Peters WB 4 marc.a.peters@phlps.com P.M.F. Fonseca WB

More information

Matching and Retrieval Based on the Vocabulary and Grammar of Color Patterns

Matching and Retrieval Based on the Vocabulary and Grammar of Color Patterns 38 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 9, NO. 1, JANUARY 2000 Matchng and Retreval Based on the Vocabulary and Grammar of Color Patterns Aleksandra Mojslovć, Member, IEEE, Jelena Kovačevć, Senor

More information

Outline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1

Outline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1 4/14/011 Outlne Dscrmnatve classfers for mage recognton Wednesday, Aprl 13 Krsten Grauman UT-Austn Last tme: wndow-based generc obect detecton basc ppelne face detecton wth boostng as case study Today:

More information

Enhanced Watermarking Technique for Color Images using Visual Cryptography

Enhanced Watermarking Technique for Color Images using Visual Cryptography Informaton Assurance and Securty Letters 1 (2010) 024-028 Enhanced Watermarkng Technque for Color Images usng Vsual Cryptography Enas F. Al rawashdeh 1, Rawan I.Zaghloul 2 1 Balqa Appled Unversty, MIS

More information

Outline. Type of Machine Learning. Examples of Application. Unsupervised Learning

Outline. Type of Machine Learning. Examples of Application. Unsupervised Learning Outlne Artfcal Intellgence and ts applcatons Lecture 8 Unsupervsed Learnng Professor Danel Yeung danyeung@eee.org Dr. Patrck Chan patrckchan@eee.org South Chna Unversty of Technology, Chna Introducton

More information

Hierarchical clustering for gene expression data analysis

Hierarchical clustering for gene expression data analysis Herarchcal clusterng for gene expresson data analyss Gorgo Valentn e-mal: valentn@ds.unm.t Clusterng of Mcroarray Data. Clusterng of gene expresson profles (rows) => dscovery of co-regulated and functonally

More information

Data Modelling and. Multimedia. Databases M. Multimedia. Information Retrieval Part II. Outline

Data Modelling and. Multimedia. Databases M. Multimedia. Information Retrieval Part II. Outline ALMA MATER STUDIORUM - UNIVERSITÀ DI BOLOGNA Data Modellng and Multmeda Databases M Internatonal Second cycle degree programme (LM) n Dgtal Humantes and Dgtal Knowledge (DHDK) Unversty of Bologna Multmeda

More information

Lecture 5: Multilayer Perceptrons

Lecture 5: Multilayer Perceptrons Lecture 5: Multlayer Perceptrons Roger Grosse 1 Introducton So far, we ve only talked about lnear models: lnear regresson and lnear bnary classfers. We noted that there are functons that can t be represented

More information

Query Clustering Using a Hybrid Query Similarity Measure

Query Clustering Using a Hybrid Query Similarity Measure Query clusterng usng a hybrd query smlarty measure Fu. L., Goh, D.H., & Foo, S. (2004). WSEAS Transacton on Computers, 3(3), 700-705. Query Clusterng Usng a Hybrd Query Smlarty Measure Ln Fu, Don Hoe-Lan

More information

PRÉSENTATIONS DE PROJETS

PRÉSENTATIONS DE PROJETS PRÉSENTATIONS DE PROJETS Rex Onlne (V. Atanasu) What s Rex? Rex s an onlne browser for collectons of wrtten documents [1]. Asde ths core functon t has however many other applcatons that make t nterestng

More information

Problem Set 3 Solutions

Problem Set 3 Solutions Introducton to Algorthms October 4, 2002 Massachusetts Insttute of Technology 6046J/18410J Professors Erk Demane and Shaf Goldwasser Handout 14 Problem Set 3 Solutons (Exercses were not to be turned n,

More information

Outline. Self-Organizing Maps (SOM) US Hebbian Learning, Cntd. The learning rule is Hebbian like:

Outline. Self-Organizing Maps (SOM) US Hebbian Learning, Cntd. The learning rule is Hebbian like: Self-Organzng Maps (SOM) Turgay İBRİKÇİ, PhD. Outlne Introducton Structures of SOM SOM Archtecture Neghborhoods SOM Algorthm Examples Summary 1 2 Unsupervsed Hebban Learnng US Hebban Learnng, Cntd 3 A

More information

UB at GeoCLEF Department of Geography Abstract

UB at GeoCLEF Department of Geography   Abstract UB at GeoCLEF 2006 Mguel E. Ruz (1), Stuart Shapro (2), June Abbas (1), Slva B. Southwck (1) and Davd Mark (3) State Unversty of New York at Buffalo (1) Department of Lbrary and Informaton Studes (2) Department

More information

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points; Subspace clusterng Clusterng Fundamental to all clusterng technques s the choce of dstance measure between data ponts; D q ( ) ( ) 2 x x = x x, j k = 1 k jk Squared Eucldean dstance Assumpton: All features

More information

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision SLAM Summer School 2006 Practcal 2: SLAM usng Monocular Vson Javer Cvera, Unversty of Zaragoza Andrew J. Davson, Imperal College London J.M.M Montel, Unversty of Zaragoza. josemar@unzar.es, jcvera@unzar.es,

More information

The Codesign Challenge

The Codesign Challenge ECE 4530 Codesgn Challenge Fall 2007 Hardware/Software Codesgn The Codesgn Challenge Objectves In the codesgn challenge, your task s to accelerate a gven software reference mplementaton as fast as possble.

More information

COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL

COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL Nader Safavan and Shohreh Kasae Department of Computer Engneerng Sharf Unversty of Technology Tehran, Iran skasae@sharf.edu

More information

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

Learning the Kernel Parameters in Kernel Minimum Distance Classifier Learnng the Kernel Parameters n Kernel Mnmum Dstance Classfer Daoqang Zhang 1,, Songcan Chen and Zh-Hua Zhou 1* 1 Natonal Laboratory for Novel Software Technology Nanjng Unversty, Nanjng 193, Chna Department

More information

Hybrid Non-Blind Color Image Watermarking

Hybrid Non-Blind Color Image Watermarking Hybrd Non-Blnd Color Image Watermarkng Ms C.N.Sujatha 1, Dr. P. Satyanarayana 2 1 Assocate Professor, Dept. of ECE, SNIST, Yamnampet, Ghatkesar Hyderabad-501301, Telangana 2 Professor, Dept. of ECE, AITS,

More information

An Image Fusion Approach Based on Segmentation Region

An Image Fusion Approach Based on Segmentation Region Rong Wang, L-Qun Gao, Shu Yang, Yu-Hua Cha, and Yan-Chun Lu An Image Fuson Approach Based On Segmentaton Regon An Image Fuson Approach Based on Segmentaton Regon Rong Wang, L-Qun Gao, Shu Yang 3, Yu-Hua

More information

Semantic Image Retrieval Using Region Based Inverted File

Semantic Image Retrieval Using Region Based Inverted File Semantc Image Retreval Usng Regon Based Inverted Fle Dengsheng Zhang, Md Monrul Islam, Guoun Lu and Jn Hou 2 Gppsland School of Informaton Technology, Monash Unversty Churchll, VIC 3842, Australa E-mal:

More information

Signature and Lexicon Pruning Techniques

Signature and Lexicon Pruning Techniques Sgnature and Lexcon Prunng Technques Srnvas Palla, Hansheng Le, Venu Govndaraju Centre for Unfed Bometrcs and Sensors Unversty at Buffalo {spalla2, hle, govnd}@cedar.buffalo.edu Abstract Handwrtten word

More information

Solving two-person zero-sum game by Matlab

Solving two-person zero-sum game by Matlab Appled Mechancs and Materals Onlne: 2011-02-02 ISSN: 1662-7482, Vols. 50-51, pp 262-265 do:10.4028/www.scentfc.net/amm.50-51.262 2011 Trans Tech Publcatons, Swtzerland Solvng two-person zero-sum game by

More information

Local Quaternary Patterns and Feature Local Quaternary Patterns

Local Quaternary Patterns and Feature Local Quaternary Patterns Local Quaternary Patterns and Feature Local Quaternary Patterns Jayu Gu and Chengjun Lu The Department of Computer Scence, New Jersey Insttute of Technology, Newark, NJ 0102, USA Abstract - Ths paper presents

More information

Optimal Workload-based Weighted Wavelet Synopses

Optimal Workload-based Weighted Wavelet Synopses Optmal Workload-based Weghted Wavelet Synopses Yoss Matas School of Computer Scence Tel Avv Unversty Tel Avv 69978, Israel matas@tau.ac.l Danel Urel School of Computer Scence Tel Avv Unversty Tel Avv 69978,

More information

Relevance Feedback for Image Retrieval

Relevance Feedback for Image Retrieval Vashal D Dhale et al, / (IJCSIT Internatonal Journal of Computer Scence and Informaton Technologes, Vol 4 (2, 203, 39-323 Relevance Feedback for Image Retreval Vashal D Dhale, Dr A R Mahaan, Prof Uma Thakur

More information

CMPS 10 Introduction to Computer Science Lecture Notes

CMPS 10 Introduction to Computer Science Lecture Notes CPS 0 Introducton to Computer Scence Lecture Notes Chapter : Algorthm Desgn How should we present algorthms? Natural languages lke Englsh, Spansh, or French whch are rch n nterpretaton and meanng are not

More information

12/2/2009. Announcements. Parametric / Non-parametric. Case-Based Reasoning. Nearest-Neighbor on Images. Nearest-Neighbor Classification

12/2/2009. Announcements. Parametric / Non-parametric. Case-Based Reasoning. Nearest-Neighbor on Images. Nearest-Neighbor Classification Introducton to Artfcal Intellgence V22.0472-001 Fall 2009 Lecture 24: Nearest-Neghbors & Support Vector Machnes Rob Fergus Dept of Computer Scence, Courant Insttute, NYU Sldes from Danel Yeung, John DeNero

More information

HIGH-LEVEL SEMANTICS OF IMAGES IN WEB DOCUMENTS USING WEIGHTED TAGS AND STRENGTH MATRIX

HIGH-LEVEL SEMANTICS OF IMAGES IN WEB DOCUMENTS USING WEIGHTED TAGS AND STRENGTH MATRIX HIGH-LEVEL SEMANTICS OF IMAGES IN WEB DOCUMENTS USING WEIGHTED TAGS AND STRENGTH MATRIX P.Shanmugavadvu 1, P.Sumathy 2, A.Vadvel 3 12 Department of Computer Scence and Applcatons, Gandhgram Rural Insttute,

More information

MPEG-7 Pictorially Enriched Ontologies for Video Annotation

MPEG-7 Pictorially Enriched Ontologies for Video Annotation MPEG-7 Pctorally Enrched Ontologes for Vdeo Annotaton C. Grana, R.Vezzan, D. Bulgarell, R. Cucchara Dpartmento d Ingegnera dell Informazone Unverstà degl Stud d Modena e Reggo Emla Abstract. A system for

More information

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz Compler Desgn Sprng 2014 Regster Allocaton Sample Exercses and Solutons Prof. Pedro C. Dnz USC / Informaton Scences Insttute 4676 Admralty Way, Sute 1001 Marna del Rey, Calforna 90292 pedro@s.edu Regster

More information

Self-tuning Histograms: Building Histograms Without Looking at Data

Self-tuning Histograms: Building Histograms Without Looking at Data Self-tunng Hstograms: Buldng Hstograms Wthout Lookng at Data Ashraf Aboulnaga Computer Scences Department Unversty of Wsconsn - Madson ashraf@cs.wsc.edu Surajt Chaudhur Mcrosoft Research surajtc@mcrosoft.com

More information

An Improved Image Segmentation Algorithm Based on the Otsu Method

An Improved Image Segmentation Algorithm Based on the Otsu Method 3th ACIS Internatonal Conference on Software Engneerng, Artfcal Intellgence, Networkng arallel/dstrbuted Computng An Improved Image Segmentaton Algorthm Based on the Otsu Method Mengxng Huang, enjao Yu,

More information

ANALYSIS OF ADAPTIF LOCAL REGION IMPLEMENTATION ON LOCAL THRESHOLDING METHOD

ANALYSIS OF ADAPTIF LOCAL REGION IMPLEMENTATION ON LOCAL THRESHOLDING METHOD Nusantara Journal of Computers and ts Applcatons ANALYSIS F ADAPTIF LCAL REGIN IMPLEMENTATIN N LCAL THRESHLDING METHD I Gust Agung Socrates Ad Guna 1), Hendra Maulana 2), Agus Zanal Arfn 3) and Dn Adn

More information

Private Information Retrieval (PIR)

Private Information Retrieval (PIR) 2 Levente Buttyán Problem formulaton Alce wants to obtan nformaton from a database, but she does not want the database to learn whch nformaton she wanted e.g., Alce s an nvestor queryng a stock-market

More information

Unsupervised Learning

Unsupervised Learning Pattern Recognton Lecture 8 Outlne Introducton Unsupervsed Learnng Parametrc VS Non-Parametrc Approach Mxture of Denstes Maxmum-Lkelhood Estmates Clusterng Prof. Danel Yeung School of Computer Scence and

More information

Lecture #15 Lecture Notes

Lecture #15 Lecture Notes Lecture #15 Lecture Notes The ocean water column s very much a 3-D spatal entt and we need to represent that structure n an economcal way to deal wth t n calculatons. We wll dscuss one way to do so, emprcal

More information

A Deflected Grid-based Algorithm for Clustering Analysis

A Deflected Grid-based Algorithm for Clustering Analysis A Deflected Grd-based Algorthm for Clusterng Analyss NANCY P. LIN, CHUNG-I CHANG, HAO-EN CHUEH, HUNG-JEN CHEN, WEI-HUA HAO Department of Computer Scence and Informaton Engneerng Tamkang Unversty 5 Yng-chuan

More information

Image Interpretation Based On Similarity Measures of Visual Content Descriptors An Insight Mungamuru Nirmala

Image Interpretation Based On Similarity Measures of Visual Content Descriptors An Insight Mungamuru Nirmala Internatonal Journal of Computer Scence & Emergng Technologes (E-ISS: 2044-6004) 242 Image Interpretaton Based On Smlarty Measures of Vsual Content Descrptors An Insght Mungamuru rmala Lecturer, Department

More information

Learning-Based Top-N Selection Query Evaluation over Relational Databases

Learning-Based Top-N Selection Query Evaluation over Relational Databases Learnng-Based Top-N Selecton Query Evaluaton over Relatonal Databases Lang Zhu *, Wey Meng ** * School of Mathematcs and Computer Scence, Hebe Unversty, Baodng, Hebe 071002, Chna, zhu@mal.hbu.edu.cn **

More information

Alignment Results of SOBOM for OAEI 2010

Alignment Results of SOBOM for OAEI 2010 Algnment Results of SOBOM for OAEI 2010 Pegang Xu, Yadong Wang, Lang Cheng, Tany Zang School of Computer Scence and Technology Harbn Insttute of Technology, Harbn, Chna pegang.xu@gmal.com, ydwang@ht.edu.cn,

More information

Efficient Mean Shift Algorithm based Color Images Categorization and Searching

Efficient Mean Shift Algorithm based Color Images Categorization and Searching 152 Effcent Mean Shft Algorthm based Color Images Categorzaton and Searchng 1 Dr S K Vay, 2 Sanay Rathore, 3 Abhshek Verma and 4 Hemra Sngh Thakur 1 Professor, Head of Dept Physcs, Govt Geetanal Grl s

More information

Shape Representation Robust to the Sketching Order Using Distance Map and Direction Histogram

Shape Representation Robust to the Sketching Order Using Distance Map and Direction Histogram Shape Representaton Robust to the Sketchng Order Usng Dstance Map and Drecton Hstogram Department of Computer Scence Yonse Unversty Kwon Yun CONTENTS Revew Topc Proposed Method System Overvew Sketch Normalzaton

More information

Load Balancing for Hex-Cell Interconnection Network

Load Balancing for Hex-Cell Interconnection Network Int. J. Communcatons, Network and System Scences,,, - Publshed Onlne Aprl n ScRes. http://www.scrp.org/journal/jcns http://dx.do.org/./jcns.. Load Balancng for Hex-Cell Interconnecton Network Saher Manaseer,

More information

Module Management Tool in Software Development Organizations

Module Management Tool in Software Development Organizations Journal of Computer Scence (5): 8-, 7 ISSN 59-66 7 Scence Publcatons Management Tool n Software Development Organzatons Ahmad A. Al-Rababah and Mohammad A. Al-Rababah Faculty of IT, Al-Ahlyyah Amman Unversty,

More information

Edge Detection in Noisy Images Using the Support Vector Machines

Edge Detection in Noisy Images Using the Support Vector Machines Edge Detecton n Nosy Images Usng the Support Vector Machnes Hlaro Gómez-Moreno, Saturnno Maldonado-Bascón, Francsco López-Ferreras Sgnal Theory and Communcatons Department. Unversty of Alcalá Crta. Madrd-Barcelona

More information

Efficient Content Representation in MPEG Video Databases

Efficient Content Representation in MPEG Video Databases Effcent Content Representaton n MPEG Vdeo Databases Yanns S. Avrths, Nkolaos D. Doulams, Anastasos D. Doulams and Stefanos D. Kollas Department of Electrcal and Computer Engneerng Natonal Techncal Unversty

More information

A Knowledge Management System for Organizing MEDLINE Database

A Knowledge Management System for Organizing MEDLINE Database A Knowledge Management System for Organzng MEDLINE Database Hyunk Km, Su-Shng Chen Computer and Informaton Scence Engneerng Department, Unversty of Florda, Ganesvlle, Florda 32611, USA Wth the exploson

More information

Range images. Range image registration. Examples of sampling patterns. Range images and range surfaces

Range images. Range image registration. Examples of sampling patterns. Range images and range surfaces Range mages For many structured lght scanners, the range data forms a hghly regular pattern known as a range mage. he samplng pattern s determned by the specfc scanner. Range mage regstraton 1 Examples

More information

Corner-Based Image Alignment using Pyramid Structure with Gradient Vector Similarity

Corner-Based Image Alignment using Pyramid Structure with Gradient Vector Similarity Journal of Sgnal and Informaton Processng, 013, 4, 114-119 do:10.436/jsp.013.43b00 Publshed Onlne August 013 (http://www.scrp.org/journal/jsp) Corner-Based Image Algnment usng Pyramd Structure wth Gradent

More information

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour 6.854 Advanced Algorthms Petar Maymounkov Problem Set 11 (November 23, 2005) Wth: Benjamn Rossman, Oren Wemann, and Pouya Kheradpour Problem 1. We reduce vertex cover to MAX-SAT wth weghts, such that the

More information

Sequential search. Building Java Programs Chapter 13. Sequential search. Sequential search

Sequential search. Building Java Programs Chapter 13. Sequential search. Sequential search Sequental search Buldng Java Programs Chapter 13 Searchng and Sortng sequental search: Locates a target value n an array/lst by examnng each element from start to fnsh. How many elements wll t need to

More information

Analysis of Continuous Beams in General

Analysis of Continuous Beams in General Analyss of Contnuous Beams n General Contnuous beams consdered here are prsmatc, rgdly connected to each beam segment and supported at varous ponts along the beam. onts are selected at ponts of support,

More information

Optimizing Document Scoring for Query Retrieval

Optimizing Document Scoring for Query Retrieval Optmzng Document Scorng for Query Retreval Brent Ellwen baellwe@cs.stanford.edu Abstract The goal of ths project was to automate the process of tunng a document query engne. Specfcally, I used machne learnng

More information

Description of NTU Approach to NTCIR3 Multilingual Information Retrieval

Description of NTU Approach to NTCIR3 Multilingual Information Retrieval Proceedngs of the Thrd NTCIR Workshop Descrpton of NTU Approach to NTCIR3 Multlngual Informaton Retreval Wen-Cheng Ln and Hsn-Hs Chen Department of Computer Scence and Informaton Engneerng Natonal Tawan

More information

A fast algorithm for color image segmentation

A fast algorithm for color image segmentation Unersty of Wollongong Research Onlne Faculty of Informatcs - Papers (Arche) Faculty of Engneerng and Informaton Scences 006 A fast algorthm for color mage segmentaton L. Dong Unersty of Wollongong, lju@uow.edu.au

More information

PYTHON IMPLEMENTATION OF VISUAL SECRET SHARING SCHEMES

PYTHON IMPLEMENTATION OF VISUAL SECRET SHARING SCHEMES PYTHON IMPLEMENTATION OF VISUAL SECRET SHARING SCHEMES Ruxandra Olmd Faculty of Mathematcs and Computer Scence, Unversty of Bucharest Emal: ruxandra.olmd@fm.unbuc.ro Abstract Vsual secret sharng schemes

More information

Information Retrieval

Information Retrieval Anmol Bhasn abhasn[at]cedar.buffalo.edu Moht Devnan mdevnan[at]cse.buffalo.edu Sprng 2005 #$ "% &'" (! Informaton Retreval )" " * + %, ##$ + *--. / "#,0, #'",,,#$ ", # " /,,#,0 1"%,2 '",, Documents are

More information

Programming in Fortran 90 : 2017/2018

Programming in Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Exercse 1 : Evaluaton of functon dependng on nput Wrte a program who evaluate the functon f (x,y) for any two user specfed values

More information

Real-time Motion Capture System Using One Video Camera Based on Color and Edge Distribution

Real-time Motion Capture System Using One Video Camera Based on Color and Edge Distribution Real-tme Moton Capture System Usng One Vdeo Camera Based on Color and Edge Dstrbuton YOSHIAKI AKAZAWA, YOSHIHIRO OKADA, AND KOICHI NIIJIMA Graduate School of Informaton Scence and Electrcal Engneerng,

More information

Image Alignment CSC 767

Image Alignment CSC 767 Image Algnment CSC 767 Image algnment Image from http://graphcs.cs.cmu.edu/courses/15-463/2010_fall/ Image algnment: Applcatons Panorama sttchng Image algnment: Applcatons Recognton of object nstances

More information

Adaptive Knowledge-Based Visualization for Accessing Educational Examples

Adaptive Knowledge-Based Visualization for Accessing Educational Examples Adaptve Knowledge-Based Vsualzaton for Accessng Educatonal Examples Peter Bruslovsky, Jae-wook Ahn, Tbor Dumtru, Mchael Yudelson School of Informaton Scences, Unversty of Pttsburgh {peterb, jaa38, mvy3}@ptt.edu

More information

Discriminative Dictionary Learning with Pairwise Constraints

Discriminative Dictionary Learning with Pairwise Constraints Dscrmnatve Dctonary Learnng wth Parwse Constrants Humn Guo Zhuoln Jang LARRY S. DAVIS UNIVERSITY OF MARYLAND Nov. 6 th, Outlne Introducton/motvaton Dctonary Learnng Dscrmnatve Dctonary Learnng wth Parwse

More information

3D Metric Reconstruction with Auto Calibration Method CS 283 Final Project Tarik Adnan Moon

3D Metric Reconstruction with Auto Calibration Method CS 283 Final Project Tarik Adnan Moon 3D Metrc Reconstructon wth Auto Calbraton Method CS 283 Fnal Project Tark Adnan Moon tmoon@collge.harvard.edu Abstract In ths paper, dfferent methods for auto camera calbraton have been studed for metrc

More information

SAO: A Stream Index for Answering Linear Optimization Queries

SAO: A Stream Index for Answering Linear Optimization Queries SAO: A Stream Index for Answerng near Optmzaton Queres Gang uo Kun-ung Wu Phlp S. Yu IBM T.J. Watson Research Center {luog, klwu, psyu}@us.bm.com Abstract near optmzaton queres retreve the top-k tuples

More information

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration Improvement of Spatal Resoluton Usng BlockMatchng Based Moton Estmaton and Frame Integraton Danya Suga and Takayuk Hamamoto Graduate School of Engneerng, Tokyo Unversty of Scence, 6-3-1, Nuku, Katsuska-ku,

More information

Querying by sketch geographical databases. Yu Han 1, a *

Querying by sketch geographical databases. Yu Han 1, a * 4th Internatonal Conference on Sensors, Measurement and Intellgent Materals (ICSMIM 2015) Queryng by sketch geographcal databases Yu Han 1, a * 1 Department of Basc Courses, Shenyang Insttute of Artllery,

More information

Kiran Joy, International Journal of Advanced Engineering Technology E-ISSN

Kiran Joy, International Journal of Advanced Engineering Technology E-ISSN Kran oy, nternatonal ournal of Advanced Engneerng Technology E-SS 0976-3945 nt Adv Engg Tech/Vol. V/ssue /Aprl-une,04/9-95 Research Paper DETERMATO O RADATVE VEW ACTOR WTOUT COSDERG TE SADOWG EECT Kran

More information

SRBIR: Semantic Region Based Image Retrieval by Extracting the Dominant Region and Semantic Learning

SRBIR: Semantic Region Based Image Retrieval by Extracting the Dominant Region and Semantic Learning Journal of Computer Scence 7 (3): 400-408, 2011 ISSN 1549-3636 2011 Scence Publcatons SRBIR: Semantc Regon Based Image Retreval by Extractng the Domnant Regon and Semantc Learnng 1 I. Felc Raam and 2 S.

More information

Brushlet Features for Texture Image Retrieval

Brushlet Features for Texture Image Retrieval DICTA00: Dgtal Image Computng Technques and Applcatons, 1 January 00, Melbourne, Australa 1 Brushlet Features for Texture Image Retreval Chbao Chen and Kap Luk Chan Informaton System Research Lab, School

More information

Using Fuzzy Logic to Enhance the Large Size Remote Sensing Images

Using Fuzzy Logic to Enhance the Large Size Remote Sensing Images Internatonal Journal of Informaton and Electroncs Engneerng Vol. 5 No. 6 November 015 Usng Fuzzy Logc to Enhance the Large Sze Remote Sensng Images Trung Nguyen Tu Huy Ngo Hoang and Thoa Vu Van Abstract

More information

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique //00 :0 AM Outlne and Readng The Greedy Method The Greedy Method Technque (secton.) Fractonal Knapsack Problem (secton..) Task Schedulng (secton..) Mnmum Spannng Trees (secton.) Change Money Problem Greedy

More information

SCALABLE AND VISUALIZATION-ORIENTED CLUSTERING FOR EXPLORATORY SPATIAL ANALYSIS

SCALABLE AND VISUALIZATION-ORIENTED CLUSTERING FOR EXPLORATORY SPATIAL ANALYSIS SCALABLE AND VISUALIZATION-ORIENTED CLUSTERING FOR EXPLORATORY SPATIAL ANALYSIS J.H.Guan, F.B.Zhu, F.L.Ban a School of Computer, Spatal Informaton & Dgtal Engneerng Center, Wuhan Unversty, Wuhan, 430079,

More information

USING GRAPHING SKILLS

USING GRAPHING SKILLS Name: BOLOGY: Date: _ Class: USNG GRAPHNG SKLLS NTRODUCTON: Recorded data can be plotted on a graph. A graph s a pctoral representaton of nformaton recorded n a data table. t s used to show a relatonshp

More information

Course Introduction. Algorithm 8/31/2017. COSC 320 Advanced Data Structures and Algorithms. COSC 320 Advanced Data Structures and Algorithms

Course Introduction. Algorithm 8/31/2017. COSC 320 Advanced Data Structures and Algorithms. COSC 320 Advanced Data Structures and Algorithms Course Introducton Course Topcs Exams, abs, Proects A quc loo at a few algorthms 1 Advanced Data Structures and Algorthms Descrpton: We are gong to dscuss algorthm complexty analyss, algorthm desgn technques

More information