Object based Pseudo-3D Conversion of 2D Videos

Size: px
Start display at page:

Download "Object based Pseudo-3D Conversion of 2D Videos"

Transcription

1 Object based Pseudo-3D Coversio of 2D Videos J. Jiag 1,2 ad G. Xiao 1 1 Southwest Uiversity; 2 Uiversity of Bradford ABSTRACT: I this paper, we describe a ew algorithm to costruct pseudo-3d videos out of covetioal 2D videos at the viewig ed, where o additioal 3D iformatio is attached at the source of 2D video productio. We ame such costructed videos as pseudo-3d o the groud that the coverted video is ot true 3D but presets perceptual 3D effect whe viewed with a pair of polarized glasses. The proposed algorithm cosists of two video processig stages: (i) sematic video object segmetatio; ad (ii) estimatio of disparities. I the first stage, we propose a costraied regio-growig ad filterig approach to improve existig segmetatio techiques based o chage detectios. Such a processig stage esures that disparities are estimated i terms of sematic video objects rather tha textured regios, ad thus improve the pseudo-3d effect i terms of huma visual perceptio. I the secod stage, we propose a VO-size (video object) based disparity estimatio to costruct additioal video frames for the proposed pseudo-3d video coversio. Experimets carried out demostrate that, the proposed algorithm effectively produces perceptually harmoious pseudo-3d videos with advatages of simplicity ad low computig cost. Idexig Terms: Pseudo-3D videos, stereo video processig, ad sematic video object segmetatio 1. Itroductio Followig the o-goig research upo high resolutio digital TV, 3D TV is believed to be the ext major revolutio i the history of TV techologies. Several compaies have exhibited their 3D glass-free displayig systems[1-2], which illustrate a promisig potetial that 1

2 customers would ot eed to wear a pair of 3D glasses i order to ejoy the visual cotet i 3Ds. However, aother key issue is the eed for 3D video cotet geeratio. It is clear that the success of ay future 3D TV system will greatly deped o the availability of sufficiet 3D video materials. While it is obvious that the demad for 3D video cotet ca be partially met by ew 3D productios with higher cost, stereo coversio of existig rich source of 2D videos would remai a excellet choice [3]. The problem with such choice is that it is extremely challegig to automatically covert 2D videos ito 3D, sice o true stereo iformatio is geerally available iside covetioal 2D videos. As a result, people ofte rely o maual 3D coversio. I the commercial sector, there exist a rage of compaies who provide 3D coversio services, which are both time cosumig ad expesive. I the research commuity, attempts have bee made to develop computig algorithms for automatic 3D coversio of 2D videos, which are maily represeted by a Europea IST project ATTEST [4]. Approaches ivestigated i this project iclude: (i) prior kowledge based approach, where research is focused o huma objects oly to exploit those pre-kow features o huma bodies [4]; (ii) camera calibratio approach, where camera motio is to be detected from 2D videos to provide cues for 3D coversio [5] ad (iii) motio based approach, where motio iformatio is primarily targeted to eable o-lie depth augmetatio to be added ito received 2D videos [6-7]. Dyamic Digital Depth Research i Australia developed series of techical solutios to the 2D/3D coversio problem [8]. Their approaches ca be characterised by two directios. Oe is o the productio side ad the other is o the viewers side. O the viewers side, the 2D-to-3D coversio is challegig. Their curret solutio is based o a semi-automatic approach, where a software called DeepSee Studio is developed to eable a operator to lasso a object ad segmet it out from the rest of the image. The operator is the required to assig a artistic depth to the object (a apparet distace from the viewer). Followig that, the software (DeepSee Studio) ca fill the object 2

3 with the appropriate shade of grey correspodig to this distace ad complete the 2D/3D coversio. I 2001, Motorola Australia Research Cetre reported their work o 2D to 3D coversio for header & shoulder photos[9]. Their approach is characterized by five steps: (i) usig sake modellig algorithm to locate the head ad shoulder boudary ad remove the existig backgroud; (ii) parametric represetatio of header cotour usig multiple ellipsoids; (iii) locatig mai facial features (eyes, ose etc.); (iv) creatig parametric depth map based o facial features; ad (v) estimatio/compositio of left ad right views based o the origial photo ad the parameterised depth map. To automatically implemet the techology, three costraits must be coformed, which iclude: (1) the origial photo must cotai a huma head ad shoulder; (2) the head orietatio is such that the subject is lookig directly towards the camera ad (3) the photo should be take with a ucluttered backgroud. Otherwise, a user itervetio is required. Give the challege ad the difficulties illustrated i existig research towards 3D coversio of 2D videos, we propose a pseudo-3d approach to provide alterative solutios, where the coverted videos are ot true 3D but achieve perceptual 3D effect. Such a specificatio provides eormous scopes ad flexibilities for developmet of video coversio ad processig algorithms. To this ed, we report our ivestigatios alog this route to produce automatic pseudo-3d coversio of 2D videos. Our algorithm desig features a sematic object segmetatio ad a simple disparity estimatio based o the size of the segmeted objects, where operatioal elemets ca be implemeted o real-time basis sice the computig cost ivolved is low ad the simplicity of the algorithm is high. The rest of the paper is orgaized ito three sectios. While sectio 2 describes the sematic object segmetatio, sectio 3 describes disparity estimatio ad pseudo-3d video 3

4 geeratio, ad sectio 4 reports experimetal results ad provide discussios o experimetal aalysis. Fial cocludig remarks are also icluded i this sectio. 2. Sematic Video Object Segmetatio Existig research o video object segmetatio is built upo detectio of chages assisted by other sidelie iformatio icludig spatial segmetatio, edge detectio ad backgroud registratio etc.[10~19]. I [10], Kim et. Al described a spatio-temporal approach for automatic segmetatio of video objects, where hypothesis test based o the estimated variaces withi a widow is proposed to exploit the temporal iformatio, ad spatial segmetatio is icluded to assist with detectio of object boudaries. The fial decisio o foregroud ad backgroud objects is made i combiig the spatially segmeted object mask with the temporally segmeted object mask, i which a two-stage process is desiged to cosider both the chage detectio ad the historical attributes. I [11], aother similar approach was described towards a robust or oise-isesitive video object segmetatio, which follows the idea of combiig spatial edge iformatio with motio-based edge detectio. The described algorithm starts with edge detectio to derive a edge map by Cay edge detector [11], i which a gradiet operatio o the Gaussia covoluted image is performed. Give the th video frame I, the Cay edge detectig operatio ca be represeted as: ( G ) Φ( I ) = θ * (1) I where G*I stads for the Gaussia covoluted image, for the gradiet operatio, ad θ for the applicatio of o-maximum suppressio ad the thresholdig operatio with hysteresis to detect ad lik the edges. 4

5 To segmet the video objects, the algorithm requires three edge maps to be extracted, which are differece edge map DE Φ( I I ) = 1, curret edge map E = Φ I ), ad backgroud edge map E b, which cotais backgroud edges to be defied by maual process or by coutig the umber of edge occurrece for each pixel through the first several frames [11]. The three edge maps are the used to produce a curretly movig edge map, ( chage ME, by selectig all edge pixels withi a small distace of DE, ad the temporarily still movig edge map, still ME by cosiderig the previous frame s movig edges. Specific represetatios of such operatios ca be highlighted as follows: ME chage = e E mi e x T chage (2) x DE ME still = e E e E b, mi e x Tstill (3) x ME 1 Where e stads for the edge poits detected from curret frame I, ad x stads for the edge pixel detected from differece frame (I -1 -I ). T chage ad T still are thresholds empirically determied to defie the small distace. Ref [11] selected T T = 1 for their experimets ad simulatios. chage = still After the idetificatio of those movig edges by (2) ad (3), the remaiig operatio for extractig video objects is combiig the two edge maps ito a fial movig edge map: ME = ME U ME, ad the select the object pixels via a logic AND operatio of those chage still pixels betwee the first ad the last edge pixel i both rows ad colums[11]. 5

6 Our assessmet of the above algorithm reveals that, while the algorithm performs well geerally, there exist a uder-segmetatio problem, where parts of the object regio are missig or there exist holes iside the sematic object beig segmeted. Whe the grey level differece betwee the object ad the backgroud is small, part of the object at its boudary will have similar itesity level to that of backgroud. I this circumstace, the edge detector fails to detect all the edges of the object, ad thus some parts iside the objects become missig. To reduce such a effect, we propose a simple liear trasformatio to ehace the cotrast of the lumiace compoet of the video frame before the Cay edge detectio is applied. Although there exist may cotrast ehacemet algorithms that may provide better performaces, our primary aim here is ot oly improvig the segmetatio accuracy, but also maitaiig the ecessary simplicity for real-time applicatios. Cosiderig the fact that icrease of cotrast will ievitably itroduce additioal oise, we also desiged a simple 2D filter to remove the oise. Give a iput video frame, ( x, y), assumig that their itesity values are limited to the I rage of [a, b], its trasformed video frame ca be geerated as: b' a' g ( x, y) = a' + ( I ( x, y) a) (4) b a The above liear trasform is applied to the whole video frame before the edge detectio ad mappigs. To remove the additioal oise itroduced by the liear trasformatio, we apply a filter based o a so-called 8-coected likig regio sig techique to both the movig edge map ME ad the extracted object sequeces. 6

7 Give a pixel at (x, y) whose value is 1 i the biary map, we examie its 3 3 eighborhood to produce a set of all coected poits N={A 1, A 2, A k }. If k T, a threshold for oise removal, the set of poits N will be regarded as oise ad thus beig removed. Whe part of the movig object is relatively still across a few frames, the edge maps i both (2) ad (3) will fail to iclude those poits, ad thus creatig large holes iside the segmeted object. I this case, existig post-processig such as morphological operatios will ot be able to recover those missig parts iside or at the boudary of the segmeted object. To provide a solutio, we propose a costraied regio-growig techique to recover those missig parts. Ulike the ormal regio growig used by those spatial segmetatio techiques for still images[20], our proposed growig is uder costraied situatio to reflect the fact that: (i) the seed selectio is fixed at those edge poits at the boudary of the fial edge map; (ii) the umber of pixels outside the first ad the last edge poits must be smaller tha a certai limit. I other words, if the majority of the pixels o ay row are outside the boudary of the edge map, the costraied regio grow will ot be applied. Give the fial edge map ME, we examie those pixels outside the first ad the last pixel i each row to see if ay further growig ca be facilitated by usig the pixel as seeds at the boudary of the edge map. Specifically, give the set of pixels outside the first ad the last edge poits i the ith row: PO i ={PO 1, PO 2, PO k }, we decide whether the regio of those edge poits should be grow ito ay of the poits iside PO i or ot by a test as give below: ei if POi e Te POi = (5) POi else where T e is a threshold idicatig that the pixel tested is very similar to e, which is the first or the last edge pixel depedig o which of these two edge poits is closer to the positio of PO i. 7

8 If the coditio is satisfied, the PO i will be grow ito the edge poits. Otherwise, they will stay as they are outside the edge map. As the fial segmeted object is produced by cosiderig both colums ad rows iside the video frame [11], such costraied regio-grow will also apply to those poits alog the colums. After both rows ad colums are processed via costraied regio growig, the fial sematic VO is extracted by logic AND operatio of both row ad colum edge pixels as described i [11]. I summary, we proposed three simple operatios o top of the movig edge-based object segmetatio algorithm [11], which ca be highlighted as: (i) liear trasform to highlight ad stregthe the edges iside the video frames; (ii) oise filterig via 8-coected pixel approach; ad (iii) costraied regio-grow. 3. Pseudo-3D Video Geeratio By aalyzig the stereo visio system [21-22], it ca be see that the disparity value is iverse proportioal to the distace betwee huma eyes ad the object poit (depth), which ca be summarized as follows: f D = B (6) Z where D stads for the disparity value, f the focus legth, Z the depth, ad B the distace betwee left ad right optical ceters. As huma visual perceptio is tolerat to certai level of pixel distortio, which is the priciple of lossy video compressio, we may ot eed true disparity values to costruct stereo versio of those 2D videos ad achieve perceptual 3D effects. As explaied i the itroductio, 8

9 we refer such coverted 3D videos as pseudo-3d videos, idicatig the fact that the proposed 3D coversio is ot true 3D. To this ed, automatic coversio of 2D videos ito their pseudo-3d versio becomes less challegig. The remaiig issue becomes how to maximize its pseudo-3d effectiveess. Although geeral 2D videos do cotai some iformatio leadig to their true disparity estimatio [3], which formulate the foudatio for those multi-view algorithms, such pixel-wise estimatio of disparities could ivolve high computig cost ad legthy processig. To speed up the process ad reduce the computatioal cost, we could just segmet the scees ito a umber of meaigful cotet layers, such as object layers ad backgroud layer, ad the estimate disparity values for those layers rather tha each idividual pixel. As a result, sematic object segmetatio becomes crucial for such pseudo- 3D coversio. Although segmetatio has bee researched for past decades, sematic object segmetatio has ot started util recet years whe image processig starts to move from low-level to sematic level. The advatage of such sematic object segmetatio lies i the fact that segmeted objects are ot textured cosistet regios aymore, but represetig idividual object, matchig huma visual cotet uderstadig. I our pseudo-3d video coversio, such sematic object based disparity estimatio ad allocatio will sigificatly ehace our 3D viewig experieces sice the disparity allocated matches huma visual cotet uderstadig. Cosequetly, such approach will also provide additioal tolerace for iaccuracy i disparity estimatios. To estimate the disparity for each segmeted object, equatio (6) ca be further exploited with the fact that the disparity is iverse proportioal to the magificatio (Z/f), ad the ature of such magificatio ca be characterized by: f O Z S (7) where O stads for the object mask area, ad S for the surface area of the object. 9

10 Thus equatio (6) ca be rearraged as: f O D = B B = λ O (8) Z S As B is fixed oce the stereoscopic geometry (such as parallax) is determied, the oly varyig factor is S. As a matter of fact, S ca be practically regarded as fixed with respect to the segmeted object. As a example, oce a car is segmeted as the foregroud object, its surface area is a costat value relative to the object area. Therefore, it ca be iferred that S is a parameter depedet o the area of the segmeted object. Such aalysis provides a practical platform for us to estimate the disparity value for each segmeted object. I other words, the disparity should be estimated as adaptive to the area of the segmeted object. To facilitate the estimatio of S o real-time basis, we replace B/S by a sigle scalig factor. Further, to take ito cosideratio the total frame size, we cosider such a sigle scalig factor i relatio to both the segmeted object area ad the total frame size rather tha the object area aloe. To this ed, we ca arrage (8) ito the followig: O D = λ (9) T where λ is the scalig factor ad T is the total frame size. By so doig, the proposed algorithm achieves pseudo-3d coversio of 2D videos via the fact that the frame backgroud T provides a referece for all objects to create the pseudo- 3D effect. I other words, each object is allocated a disparity value agaist the frame backgroud. I the case of multiple objects, as log as they stay apart, their pseudo-3d effect is maily perceived by viewers with respect to the frame backgroud rather tha idividual 10

11 object. From equatio (9), it is see that the disparity estimated for each separate object depeds o the object size ad the frame size. Therefore, large objects will have large disparity ad thus produce stroger pseudo-3d effect agaist the frame backgroud. Sice larger objects attract more attetio from viewers, visual perceptio of the proposed pseudo-3d coversio will maily be implemeted betwee those large objects ad the frame backgroud. For some small objects, eve their disparity allocatio may ot be covicig i relatio to large objects, viewers ca hardly perceive ay discomfort due to the fact that: (i) the multiple objects stay apart; (ii) small objects attract less attetio from viewers; ad (iii) all objects have their disparity estimated with referece to the frame backgroud rather tha agaist each other. Therefore, large disparity may exist betwee each object ad the frame backgroud, but the disparity amog the objects themselves remais to be small depedig o their relative sizes. Essetially, the performace of the proposed algorithm is depedet o the accuracy of object segmetatio to a large extet. Therefore, there exist a few costraits for the proposed algorithm to achieve effective pseudo-3d coversio ad esure that such coversio is perceptually covicig. These costraits iclude: (i) temporal segmetatio ca effectively divide iput videos ito sectios, where all video frames cotai cosistet scees ad movig objects are smooth ad cotiuous. As a result, segmeted objects would provide cotiuous disparity estimatio ad thus the coverted pseudo-3d video presets smooth 3D effect amog objects agaist the frame backgroud; (ii) there is o camera movemet iside the video sequeces to esure that the backgroud ca be effectively separated by the proposed segmetatio algorithm. For those video frames with camera movemet, additioal tool is eeded to detect such movemet [5, 25] ad dedicated disparity estimatio has to be desiged; (iii) i the case of multiple objects, they do ot cotiuously move together ad overlap. Otherwise, occlusio will occur ad the proposed algorithm will segmet the overlapped objects ito oe sigle object. As a result, the disparity estimated could suddely become much 11

12 larger i compariso with its previous oes ad thus create a abrupt effect, which may cause discomfort for viewers. However, the proposed algorithm will ot produce cotradictig pseudo-3d effect i ay circumstaces. This is because the proposed algorithm always estimates disparity for each object agaist the backgroud frame rather tha agaist each other amog objects. Therefore, whether with differet objects ad they stay apart or occluded, their pseudo-3d effect created by the proposed algorithm will ot cotradict with their origial 2D view, sice the worst case is that the occluded objects are allocated with the same disparity ad thus they become 2D withi themselves, but still pseudo-3d with respect to the backgroud.. To view such coverted pseudo-stereo videos ad prove the cocept illustrated i (9), we arraged the stereo display as such that objects i the left eye chael are moved toward right by a few pixels, ad objects i the right eye chael are moved toward left by a few pixels. The visual effect becomes that the objects are behid the display plae, ad the backgroud layers for the two chaels are geerated as a itermediate objects layer that are moved by a few more pixels. As a result, the backgroud layer is displayed behid the object layer. Such arragemet is equivalet to the sceario that a 3D real world scee is preseted behid a widow (the widow glass is the display plae). I additio, to complete the fial pseudo-3d video coversio, we take the existig 2D video as the left chael ad usig the estimated disparity to costruct a right chael by shiftig pixels as explaied. To save the cost of viewig polarized images, we simply put the left frame ito the gree compoet ad the right frame ito the red compoet of the origial coloured video frames. Hece, with a glass of red ad gree (red o the right), people ca view the coverted pseudo-stereo videos ad assess their 3D effect. 12

13 4. Experimets Desig ad Cocludig Remarks I this sectio, we desig two experimetal phases to evaluate the proposed pseudo-3d video algorithm. The first phase is desiged to evaluate the proposed object segmetatio ad the secod phase is to evaluate the 3D effect of the coverted pseudo-3d videos. Throughout the experimets, we use a set of three video clips: Hall-moitor, Mother&daughter, ad Girl, all of which are publicly available ad Hall-moitor is the same as that used i [11]. This arragemet will have advatages that (i) such ope test data set eables ay other developed algorithms to be bechmarked by our proposed algorithm without repeatig our work; (ii) comparative assessmet of our proposed algorithm ca be easily implemeted by usig these ope test data set. For the first phase evaluatio, our desig aims at eablig detailed aalysis of how each elemet of our proposed object segmetatio actually cotributes to the effect of sematic video object segmetatio. To this ed, we implemeted the VO segmetatio algorithm described i [11] as our bechmark, ad carried out experimets each time oe elemet of the proposed is added. These elemets iclude: (i) liear trasform for cotrast ehacemet; (ii) filterig for removal of oise; ad (iii) costraied regio-growig. To highlight the advatages achieved by the above elemets i sematic video object segmetatio, we select Hall-moitor ad mother&daughter to illustrate our experimetal results. Figure-1 illustrates the segmeted results by addig liear trasformatio ito the bechmark, where part-(a), (c), (e) ad (f) are the segmeted objects by the bechmark ad part-(b), (d), (f) ad (g) are the segmeted objects by the bechmark with the proposed liear trasform. As see, the proposed liear trasform itroduced additioal oise while the segmetatio accuracy is improved. 13

14 Figure-2 illustrates the segmeted results by addig both the liear trasform ad the filter ito the bechmark, which show that the oise itroduced is effectively removed. Figure-3 illustrates the segmeted results by addig the elemet of costraied regiogrowig ito the bechmark, where part-(a) ad (c) represets the results of the bechmark ad part-(b) ad (d) the results of the proposed. Although the proposed regio-growig ca ot recover all the missig parts, it is see that the recovery of the proposed techique does have improvemet compared with the bechmark. By puttig all the elemets together, the segmeted video objects by the proposed algorithm ca be illustrated i Figure-4. Note that all the figures illustrated here are much larger tha those illustrated i refereces [10-19]. If we make the pictures smaller, the segmetatio results will look better as those boudaries will look smoother. Correspodigly, a sequece of the segmeted objects is illustrated i Figure-5 to provide a cotiuous visual ispectio. For the secod phase experimet, we implemeted the proposed algorithm by addig the segmetatio ad disparity estimatio together o a PC computig platform ad ru the software simulatio i C++ to covert the 2D videos iside the test data set ito their pseudo- 3D videos. Figure-6 illustrates such coverted pseudo-3d video for Hall-Moitor ad Girl, i which part (a~c) show the segmeted foregroud objects, ad part (d~f) show the coverted Hall&moitor i pseudo-3d, ad part (g~i) show the coverted Girl i pseudo-3d. To visually ispect the 3D effect of these coverted pseudo-3d video frames, we eed to wear a paper glass with gree o the left ad red o the right. Such 3D display is the cheapest optio suitable for academic research laboratories, where facilities are ofte uder fuded. From all the illustratios, it is see that the proposed algorithm achieves pseudo-3d coversio of 2D videos with certai level of perceptual 3D effect. Apart from beig able to improve the existig sematic video object segmetatio, the proposed work explores a ew 14

15 approach i geeratig pseudo-3d video cotet rather tha true 3D out of covetioal 2D videos, where the criteria of 3D coversio is chaged to relax the whole process ad thus make the coversio less challegig. Compared with existig approaches, our proposed algorithm features a pseudo-3d criterio, where oly perceptual 3D effect is targeted istead of tryig to be close to their origial true 3D. Uder this criterio, a rage of ew possibilities ca be explored for further research ad the existig efforts ca also be further exteded. Examples of such directios could iclude: (i) 3D modelig iside videos [23]; (ii) depth estimatio [24], (iii) multi-view imagig [26, 27], ad (iv) further processig tools to overcome the costraits [5, 25]. Fially, the authors wish to ackowledge fiacial support from both HERMES project, fuded uder Europea Framework-7 programme, ad LIVE project, fuded uder Europea Framework-6 programme. Refereces [1] [2] [3] [4] [5] Marc Pollefeys et. Al Self-calibratio ad metric recostructio i spite of varyig ad ukow itrisic camera parameters, Iteratioal Joural of Computer Visio, Kluwer Academic Pulishers, 1998 [6] F. Erst, P. Wiliski, K va Overveld Dese structure from motio-a approach based o segmet matchig, Lecture Note i Computer Sciece, Spriger-Verlag GmbH, Vol 2351, 2002; [7] P. Wiliski ad K. va Overveld Depth from motio usig cofidece based block matchig, Proceedigs of Image ad Multidimesioal Sigal Processig Workshop, Alpbach, Austria, 1998, pp [8] Phil Harma Home based 3D etertaimet-a overview, ICIP 2000, Vacouver, September th, 2000; [9] C. Weerasighe, P. Oguboa ad W. Li 2D to Pseudo-3D coversio of head ad shoulder images usig feature based parametric disparity maps, IEEE /01, pp

16 [10] Kim M. et al. A VOP geeratio tool: automatic segmetatio of movig objects i image sequeces based o spatio-temporal iformatio, IEEE Tras. Circuits, Systems for Video Techology, Vol 9, NO 8, 1999, pp ; [11] Kim C. ad Hwag J.N. fast ad automatic video object segmetatio ad trackig for cotet-based applicatios, IEEE Tras. Circuits ad Systems for Video Techology, Vol 12, No. 2, 2002, pp ; [12] Salembier P. ad Pardas M. Hierarchical morphological segmetatio for image sequece codig, IEEE Tras. Image Processig, Vol 3, No 5, 1994, pp ; [13] Chie S.Y. et. Al. Efficiet movig object segmetatio algorithm usig backgroud registratio techique, IEEE Tras. Circuits ad Systems for Video Techology, Vol 12, No 7, 2002, pp ; [14] Shamim A. ad Robiso A. Object-based video codig by global-to-local motio segmetatio, IEEE Tras. Circuits, Systems for Video Techology, Vol 12, No 12, 2002, pp ; [15] Feg G.C. ad Jiag J "Image segmetatio i compressed domai" Joural of Electroic Imagig, Vol.12, No.3, SPIE, 2003, pp ; [16] Toklu C. et. Al. Semi-automatic video object segmetatio i the presece of occlusio, IEEE Tras. Circuits ad Systems for Video Techology, Vol 10, NO 4, 2000, pp ; [17] Kervra C. ad Heitz F. Statistical deformable model-based segmetatio of image motio, IEEE Tras fast ad automatic video object segmetatio ad trackig for cotet-based applicatios, IEEE Tras. Image Processig, Vol 8, No 4, 1999, pp ; [18] Meier T. ad Nga K. Automatic segmetatio of movig objects for video object plae geeratio, IEEE Tras. Circuits, Systems for Video Techology, Vol 8, No 5, 1998, pp ; [19] Xu Y. et. Al. Object-based image labelig through learig by example ad multi-level segmetatio, Patter Recogitio, Vol 36, pp , [20] Adams R. ad Bischof L., Seeded regio growig, IEEE Tras. Patter Aal. Machie Itell, vol.16, o.6, 1994, [21] J. Jiag ad E. Edirisighe A hybrid scheme for low bit rate codig of stereo images, IEEE Tras. O Image Processig, Vol 11, NO 2, 2002, pp ; [22] S.Barard, W. Tompso, Disparity aalysis of images, IEEE Tras. Patter Aal. Mach. Itell. 2 (July 1980) ; [23] D. Hoiem, A.A. Efros, ad M. Hebert, "Automatic Photo Pop-up", ACM SIGGRAPH, 2005; [24] Torralba A, Oliva A Depth estimatio from image structure IEEE Tras. PATTERN ANAL 24 (9): SEP 2002 ; [25] Sato T, Kabara M, Yokoya N 3-D modelig of a outdoor scee from multiple image sequeces by estimatig camera motio parameters LECT NOTES COMPUT SC 2749: ; [26] Sato T, Kabara M, Yokoya N, et al. Dese 3-D recostructio of a outdoor scee by hudreds-baselie stereo usig a had-held video camera INT J COMPUT VISION 47 (1-3): APR-JUN 2002; [27] Fa J. et al MultiView: multilevel video cotet represetatio ad retrieval, Joural of Electroic Imagig, Vol 10, No 4, 2001, pp

17 (a) (b) (c) (d) Fig. 1. Illustratio of sematic object segmetatio, (a) Hall&moitor frame 71 by the bechmark; (b) frame71 by bechmark+liear-trasform; (c) Mother&daughter segmeted by bechmark; (d) bechmark+liear-trasform for mother&daughter. (a) (b) Fig. 2 Sematic object segmetatio by bechmark+liear_trasform+filterig: (a) Hall&moitor (frame-71) segmetatio by bechmark+liear_trasform+filterig, (b) mother&daughter (frame-304) segmetatio by bechmark+liear_trasform+filterig. 17

18 (a) (b) Fig. 3 (a) Hall&moitor (frame-73) segmetatio by bechmark, (b) same frame segmetatio by bechmark + costraied regio-growig. (a) (b) (c) (d) Fig. 4 Fial segmetatio results by the proposed algorithm: (a)~(b) are origials, ad (c)~(d) are the segmeted video objects. 18

19 (a) segmeted objects i sequetial frames by the bechmark (frame 45~51) (b) segmeted objects i sequetial frames by the proposed algorithm (frame 45~51) (c) segmeted objects i sequetial frames by the bechmark (frame 55~61) (d) segmeted objects i sequetial frames by the proposed algorithm (frame 55~61) (e) segmeted objects i sequetial frames by the bechmark (frame 65~71) (f) segmeted objects i sequetial frames by the proposed algorithm (frame 65~71) Fig. 5 Fial segmetatio compariso with sequetial frames for Hall&Moitor 19

20 (a) (b) (c) (d) (e) (f) (g) (h) (i) Figure-6: Illustratio of experimetal results. (a) segmeted object i frame 35 of Hall&Moitor; (b) segmeted object i frame 145 of Hall&moitor; (c) segmeted object i frame 655 of Girl. (d) pseudo-3d coverted frame 37 of Hall&moitor; (e) Pseudo-3D coverted frame 55 of Hall&moitor; (f) Pseudo-3D coverted frame 78 of Hall&moitor; (g) Pseudo-3D coverted frame 655 of Girl; (h) Pseudo-3D coverted frame 663 of Girl; (i) Pseudo-3D coverted frame 684 of Girl. 20

Image Segmentation EEE 508

Image Segmentation EEE 508 Image Segmetatio Objective: to determie (etract) object boudaries. It is a process of partitioig a image ito distict regios by groupig together eighborig piels based o some predefied similarity criterio.

More information

Accuracy Improvement in Camera Calibration

Accuracy Improvement in Camera Calibration Accuracy Improvemet i Camera Calibratio FaJie L Qi Zag ad Reihard Klette CITR, Computer Sciece Departmet The Uiversity of Aucklad Tamaki Campus, Aucklad, New Zealad fli006, qza001@ec.aucklad.ac.z r.klette@aucklad.ac.z

More information

Dynamic Programming and Curve Fitting Based Road Boundary Detection

Dynamic Programming and Curve Fitting Based Road Boundary Detection Dyamic Programmig ad Curve Fittig Based Road Boudary Detectio SHYAM PRASAD ADHIKARI, HYONGSUK KIM, Divisio of Electroics ad Iformatio Egieerig Chobuk Natioal Uiversity 664-4 Ga Deokji-Dog Jeoju-City Jeobuk

More information

Neuro Fuzzy Model for Human Face Expression Recognition

Neuro Fuzzy Model for Human Face Expression Recognition IOSR Joural of Computer Egieerig (IOSRJCE) ISSN : 2278-0661 Volume 1, Issue 2 (May-Jue 2012), PP 01-06 Neuro Fuzzy Model for Huma Face Expressio Recogitio Mr. Mayur S. Burage 1, Prof. S. V. Dhopte 2 1

More information

A New Morphological 3D Shape Decomposition: Grayscale Interframe Interpolation Method

A New Morphological 3D Shape Decomposition: Grayscale Interframe Interpolation Method A ew Morphological 3D Shape Decompositio: Grayscale Iterframe Iterpolatio Method D.. Vizireau Politehica Uiversity Bucharest, Romaia ae@comm.pub.ro R. M. Udrea Politehica Uiversity Bucharest, Romaia mihea@comm.pub.ro

More information

Improvement of the Orthogonal Code Convolution Capabilities Using FPGA Implementation

Improvement of the Orthogonal Code Convolution Capabilities Using FPGA Implementation Improvemet of the Orthogoal Code Covolutio Capabilities Usig FPGA Implemetatio Naima Kaabouch, Member, IEEE, Apara Dhirde, Member, IEEE, Saleh Faruque, Member, IEEE Departmet of Electrical Egieerig, Uiversity

More information

3D Model Retrieval Method Based on Sample Prediction

3D Model Retrieval Method Based on Sample Prediction 20 Iteratioal Coferece o Computer Commuicatio ad Maagemet Proc.of CSIT vol.5 (20) (20) IACSIT Press, Sigapore 3D Model Retrieval Method Based o Sample Predictio Qigche Zhag, Ya Tag* School of Computer

More information

RESEARCH ON AUTOMATIC INSPECTION TECHNIQUE OF REAL-TIME RADIOGRAPHY FOR TURBINE-BLADE

RESEARCH ON AUTOMATIC INSPECTION TECHNIQUE OF REAL-TIME RADIOGRAPHY FOR TURBINE-BLADE RESEARCH ON AUTOMATIC INSPECTION TECHNIQUE OF REAL-TIME RADIOGRAPHY FOR TURBINE-BLADE Z.G. Zhou, S. Zhao, ad Z.G. A School of Mechaical Egieerig ad Automatio, Beijig Uiversity of Aeroautics ad Astroautics,

More information

Stone Images Retrieval Based on Color Histogram

Stone Images Retrieval Based on Color Histogram Stoe Images Retrieval Based o Color Histogram Qiag Zhao, Jie Yag, Jigyi Yag, Hogxig Liu School of Iformatio Egieerig, Wuha Uiversity of Techology Wuha, Chia Abstract Stoe images color features are chose

More information

IMP: Superposer Integrated Morphometrics Package Superposition Tool

IMP: Superposer Integrated Morphometrics Package Superposition Tool IMP: Superposer Itegrated Morphometrics Package Superpositio Tool Programmig by: David Lieber ( 03) Caisius College 200 Mai St. Buffalo, NY 4208 Cocept by: H. David Sheets, Dept. of Physics, Caisius College

More information

Harris Corner Detection Algorithm at Sub-pixel Level and Its Application Yuanfeng Han a, Peijiang Chen b * and Tian Meng c

Harris Corner Detection Algorithm at Sub-pixel Level and Its Application Yuanfeng Han a, Peijiang Chen b * and Tian Meng c Iteratioal Coferece o Computatioal Sciece ad Egieerig (ICCSE 015) Harris Corer Detectio Algorithm at Sub-pixel Level ad Its Applicatio Yuafeg Ha a, Peijiag Che b * ad Tia Meg c School of Automobile, Liyi

More information

Pattern Recognition Systems Lab 1 Least Mean Squares

Pattern Recognition Systems Lab 1 Least Mean Squares Patter Recogitio Systems Lab 1 Least Mea Squares 1. Objectives This laboratory work itroduces the OpeCV-based framework used throughout the course. I this assigmet a lie is fitted to a set of poits usig

More information

Derivation of perspective stereo projection matrices with depth, shape and magnification consideration

Derivation of perspective stereo projection matrices with depth, shape and magnification consideration Derivatio of perspective stereo projectio matrices with depth, shape ad magificatio cosideratio Patrick Oberthür Jauary 2014 This essay will show how to costruct a pair of stereoscopic perspective projectio

More information

Improving Template Based Spike Detection

Improving Template Based Spike Detection Improvig Template Based Spike Detectio Kirk Smith, Member - IEEE Portlad State Uiversity petra@ee.pdx.edu Abstract Template matchig algorithms like SSE, Covolutio ad Maximum Likelihood are well kow for

More information

Fast Fourier Transform (FFT) Algorithms

Fast Fourier Transform (FFT) Algorithms Fast Fourier Trasform FFT Algorithms Relatio to the z-trasform elsewhere, ozero, z x z X x [ ] 2 ~ elsewhere,, ~ e j x X x x π j e z z X X π 2 ~ The DFS X represets evely spaced samples of the z- trasform

More information

are two specific neighboring points, F( x, y)

are two specific neighboring points, F( x, y) $33/,&$7,212)7+(6(/)$92,',1* 5$1'20:$/.12,6(5('8&7,21$/*25,7+0,17+(&2/285,0$*(6(*0(17$7,21 %RJGDQ602/.$+HQU\N3$/86'DPLDQ%(5(6.$ 6LOHVLDQ7HFKQLFDO8QLYHUVLW\'HSDUWPHQWRI&RPSXWHU6FLHQFH $NDGHPLFND*OLZLFH32/$1'

More information

Introduction. Nature-Inspired Computing. Terminology. Problem Types. Constraint Satisfaction Problems - CSP. Free Optimization Problem - FOP

Introduction. Nature-Inspired Computing. Terminology. Problem Types. Constraint Satisfaction Problems - CSP. Free Optimization Problem - FOP Nature-Ispired Computig Hadlig Costraits Dr. Şima Uyar September 2006 Itroductio may practical problems are costraied ot all combiatios of variable values represet valid solutios feasible solutios ifeasible

More information

Copyright 2016 Ramez Elmasri and Shamkant B. Navathe

Copyright 2016 Ramez Elmasri and Shamkant B. Navathe Copyright 2016 Ramez Elmasri ad Shamkat B. Navathe CHAPTER 19 Query Optimizatio Copyright 2016 Ramez Elmasri ad Shamkat B. Navathe Itroductio Query optimizatio Coducted by a query optimizer i a DBMS Goal:

More information

Evaluation scheme for Tracking in AMI

Evaluation scheme for Tracking in AMI A M I C o m m u i c a t i o A U G M E N T E D M U L T I - P A R T Y I N T E R A C T I O N http://www.amiproject.org/ Evaluatio scheme for Trackig i AMI S. Schreiber a D. Gatica-Perez b AMI WP4 Trackig:

More information

On-line cursive letter recognition using sequences of local minima/maxima. Robert Powalka

On-line cursive letter recognition using sequences of local minima/maxima. Robert Powalka O-lie cursive letter recogitio usig sequeces of local miima/maxima Summary Robert Powalka 19 th August 1993 This report presets the desig ad implemetatio of a o-lie cursive letter recogizer usig sequeces

More information

Bayesian approach to reliability modelling for a probability of failure on demand parameter

Bayesian approach to reliability modelling for a probability of failure on demand parameter Bayesia approach to reliability modellig for a probability of failure o demad parameter BÖRCSÖK J., SCHAEFER S. Departmet of Computer Architecture ad System Programmig Uiversity Kassel, Wilhelmshöher Allee

More information

An Improved Shuffled Frog-Leaping Algorithm for Knapsack Problem

An Improved Shuffled Frog-Leaping Algorithm for Knapsack Problem A Improved Shuffled Frog-Leapig Algorithm for Kapsack Problem Zhoufag Li, Ya Zhou, ad Peg Cheg School of Iformatio Sciece ad Egieerig Hea Uiversity of Techology ZhegZhou, Chia lzhf1978@126.com Abstract.

More information

1. Introduction o Microscopic property responsible for MRI Show and discuss graphics that go from macro to H nucleus with N-S pole

1. Introduction o Microscopic property responsible for MRI Show and discuss graphics that go from macro to H nucleus with N-S pole Page 1 Very Quick Itroductio to MRI The poit of this itroductio is to give the studet a sufficietly accurate metal picture of MRI to help uderstad its impact o image registratio. The two major aspects

More information

Ones Assignment Method for Solving Traveling Salesman Problem

Ones Assignment Method for Solving Traveling Salesman Problem Joural of mathematics ad computer sciece 0 (0), 58-65 Oes Assigmet Method for Solvig Travelig Salesma Problem Hadi Basirzadeh Departmet of Mathematics, Shahid Chamra Uiversity, Ahvaz, Ira Article history:

More information

Elementary Educational Computer

Elementary Educational Computer Chapter 5 Elemetary Educatioal Computer. Geeral structure of the Elemetary Educatioal Computer (EEC) The EEC coforms to the 5 uits structure defied by vo Neuma's model (.) All uits are preseted i a simplified

More information

A Comparative Study of Color Edge Detection Techniques

A Comparative Study of Color Edge Detection Techniques CS31A WINTER-1314 PROJECT REPORT 1 A Comparative Study of Color Edge Detectio Techiques Masood Shaikh, Departmet of Electrical Egieerig, Staford Uiversity Abstract Edge detectio has attracted the attetio

More information

The Magma Database file formats

The Magma Database file formats The Magma Database file formats Adrew Gaylard, Bret Pikey, ad Mart-Mari Breedt Johaesburg, South Africa 15th May 2006 1 Summary Magma is a ope-source object database created by Chris Muller, of Kasas City,

More information

Fire Recognition in Video. Walter Phillips III Mubarak Shah Niels da Vitoria Lobo.

Fire Recognition in Video. Walter Phillips III Mubarak Shah Niels da Vitoria Lobo. Fire Recogitio i Video Walter Phillips III Mubarak Shah Niels da Vitoria Lobo {wrp65547,shah,iels}@cs.ucf.edu Computer Visio Laboratory Departmet of Computer Sciece Uiversity of Cetral Florida Orlado,

More information

Image based Cats and Possums Identification for Intelligent Trapping Systems

Image based Cats and Possums Identification for Intelligent Trapping Systems Volume 159 No, February 017 Image based Cats ad Possums Idetificatio for Itelliget Trappig Systems T. A. S. Achala Perera School of Egieerig Aucklad Uiversity of Techology New Zealad Joh Collis School

More information

Lower Bounds for Sorting

Lower Bounds for Sorting Liear Sortig Topics Covered: Lower Bouds for Sortig Coutig Sort Radix Sort Bucket Sort Lower Bouds for Sortig Compariso vs. o-compariso sortig Decisio tree model Worst case lower boud Compariso Sortig

More information

Mobile terminal 3D image reconstruction program development based on Android Lin Qinhua

Mobile terminal 3D image reconstruction program development based on Android Lin Qinhua Iteratioal Coferece o Automatio, Mechaical Cotrol ad Computatioal Egieerig (AMCCE 05) Mobile termial 3D image recostructio program developmet based o Adroid Li Qihua Sichua Iformatio Techology College

More information

Data diverse software fault tolerance techniques

Data diverse software fault tolerance techniques Data diverse software fault tolerace techiques Complemets desig diversity by compesatig for desig diversity s s limitatios Ivolves obtaiig a related set of poits i the program data space, executig the

More information

New HSL Distance Based Colour Clustering Algorithm

New HSL Distance Based Colour Clustering Algorithm The 4th Midwest Artificial Itelligece ad Cogitive Scieces Coferece (MAICS 03 pp 85-9 New Albay Idiaa USA April 3-4 03 New HSL Distace Based Colour Clusterig Algorithm Vasile Patrascu Departemet of Iformatics

More information

A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON

A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON Roberto Lopez ad Eugeio Oñate Iteratioal Ceter for Numerical Methods i Egieerig (CIMNE) Edificio C1, Gra Capitá s/, 08034 Barceloa, Spai ABSTRACT I this work

More information

A Novel Feature Extraction Algorithm for Haar Local Binary Pattern Texture Based on Human Vision System

A Novel Feature Extraction Algorithm for Haar Local Binary Pattern Texture Based on Human Vision System A Novel Feature Extractio Algorithm for Haar Local Biary Patter Texture Based o Huma Visio System Liu Tao 1,* 1 Departmet of Electroic Egieerig Shaaxi Eergy Istitute Xiayag, Shaaxi, Chia Abstract The locality

More information

Road Boundary Detection in Complex Urban Environment based on Low- Resolution Vision

Road Boundary Detection in Complex Urban Environment based on Low- Resolution Vision Road Boudary Detectio i Complex Urba Eviromet based o Low- Resolutio Visio Qighua We, Zehog Yag, Yixu Sog, Peifa Jia State Key Laboratory o Itelliget Techology ad Systems, Tsighua Natioal Laboratory for

More information

Performance Plus Software Parameter Definitions

Performance Plus Software Parameter Definitions Performace Plus+ Software Parameter Defiitios/ Performace Plus Software Parameter Defiitios Chapma Techical Note-TG-5 paramete.doc ev-0-03 Performace Plus+ Software Parameter Defiitios/2 Backgroud ad Defiitios

More information

GEOMETRIC REVERSE ENGINEERING USING A LASER PROFILE SCANNER MOUNTED ON AN INDUSTRIAL ROBOT

GEOMETRIC REVERSE ENGINEERING USING A LASER PROFILE SCANNER MOUNTED ON AN INDUSTRIAL ROBOT 6th Iteratioal DAAAM Baltic Coferece INDUSTRIAL ENGINEERING 24-26 April 2008, Talli, Estoia GEOMETRIC REVERSE ENGINEERING USING A LASER PROFILE SCANNER MOUNTED ON AN INDUSTRIAL ROBOT Rahayem, M.; Kjellader,

More information

ECE4050 Data Structures and Algorithms. Lecture 6: Searching

ECE4050 Data Structures and Algorithms. Lecture 6: Searching ECE4050 Data Structures ad Algorithms Lecture 6: Searchig 1 Search Give: Distict keys k 1, k 2,, k ad collectio L of records of the form (k 1, I 1 ), (k 2, I 2 ),, (k, I ) where I j is the iformatio associated

More information

Copyright 2016 Ramez Elmasri and Shamkant B. Navathe

Copyright 2016 Ramez Elmasri and Shamkant B. Navathe Copyright 2016 Ramez Elmasri ad Shamkat B. Navathe CHAPTER 26 Ehaced Data Models: Itroductio to Active, Temporal, Spatial, Multimedia, ad Deductive Databases Copyright 2016 Ramez Elmasri ad Shamkat B.

More information

. Written in factored form it is easy to see that the roots are 2, 2, i,

. Written in factored form it is easy to see that the roots are 2, 2, i, CMPS A Itroductio to Programmig Programmig Assigmet 4 I this assigmet you will write a java program that determies the real roots of a polyomial that lie withi a specified rage. Recall that the roots (or

More information

BASED ON ITERATIVE ERROR-CORRECTION

BASED ON ITERATIVE ERROR-CORRECTION A COHPARISO OF CRYPTAALYTIC PRICIPLES BASED O ITERATIVE ERROR-CORRECTIO Miodrag J. MihaljeviC ad Jova Dj. GoliC Istitute of Applied Mathematics ad Electroics. Belgrade School of Electrical Egieerig. Uiversity

More information

n n B. How many subsets of C are there of cardinality n. We are selecting elements for such a

n n B. How many subsets of C are there of cardinality n. We are selecting elements for such a 4. [10] Usig a combiatorial argumet, prove that for 1: = 0 = Let A ad B be disjoit sets of cardiality each ad C = A B. How may subsets of C are there of cardiality. We are selectig elemets for such a subset

More information

An abandoned object detection system based on dual background segmentation

An abandoned object detection system based on dual background segmentation 2009 Advaced Video ad Sigal Based Surveillace A abadoed object detectio system based o dual backgroud segmetatio A. Sigh, S. Sawa, M. Hamadlu Departmet of Electrical Egieerig I.I.T. Delhi Delhi, Idia abhiavkumar.sigh@mail2.iitd.ac.i

More information

Fundamentals of Media Processing. Shin'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dinh Le

Fundamentals of Media Processing. Shin'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dinh Le Fudametals of Media Processig Shi'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dih Le Today's topics Noparametric Methods Parze Widow k-nearest Neighbor Estimatio Clusterig Techiques k-meas Agglomerative Hierarchical

More information

GE FUNDAMENTALS OF COMPUTING AND PROGRAMMING UNIT III

GE FUNDAMENTALS OF COMPUTING AND PROGRAMMING UNIT III GE2112 - FUNDAMENTALS OF COMPUTING AND PROGRAMMING UNIT III PROBLEM SOLVING AND OFFICE APPLICATION SOFTWARE Plaig the Computer Program Purpose Algorithm Flow Charts Pseudocode -Applicatio Software Packages-

More information

Bezier curves. Figure 2 shows cubic Bezier curves for various control points. In a Bezier curve, only

Bezier curves. Figure 2 shows cubic Bezier curves for various control points. In a Bezier curve, only Edited: Yeh-Liag Hsu (998--; recommeded: Yeh-Liag Hsu (--9; last updated: Yeh-Liag Hsu (9--7. Note: This is the course material for ME55 Geometric modelig ad computer graphics, Yua Ze Uiversity. art of

More information

Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5

Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5 Morga Kaufma Publishers 26 February, 28 COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Iterface 5 th Editio Chapter 5 Set-Associative Cache Architecture Performace Summary Whe CPU performace icreases:

More information

New Fuzzy Color Clustering Algorithm Based on hsl Similarity

New Fuzzy Color Clustering Algorithm Based on hsl Similarity IFSA-EUSFLAT 009 New Fuzzy Color Clusterig Algorithm Based o hsl Similarity Vasile Ptracu Departmet of Iformatics Techology Tarom Compay Bucharest Romaia Email: patrascu.v@gmail.com Abstract I this paper

More information

Chapter 3 Classification of FFT Processor Algorithms

Chapter 3 Classification of FFT Processor Algorithms Chapter Classificatio of FFT Processor Algorithms The computatioal complexity of the Discrete Fourier trasform (DFT) is very high. It requires () 2 complex multiplicatios ad () complex additios [5]. As

More information

Python Programming: An Introduction to Computer Science

Python Programming: An Introduction to Computer Science Pytho Programmig: A Itroductio to Computer Sciece Chapter 1 Computers ad Programs 1 Objectives To uderstad the respective roles of hardware ad software i a computig system. To lear what computer scietists

More information

Improving Information Retrieval System Security via an Optimal Maximal Coding Scheme

Improving Information Retrieval System Security via an Optimal Maximal Coding Scheme Improvig Iformatio Retrieval System Security via a Optimal Maximal Codig Scheme Dogyag Log Departmet of Computer Sciece, City Uiversity of Hog Kog, 8 Tat Chee Aveue Kowloo, Hog Kog SAR, PRC dylog@cs.cityu.edu.hk

More information

Outline. Research Definition. Motivation. Foundation of Reverse Engineering. Dynamic Analysis and Design Pattern Detection in Java Programs

Outline. Research Definition. Motivation. Foundation of Reverse Engineering. Dynamic Analysis and Design Pattern Detection in Java Programs Dyamic Aalysis ad Desig Patter Detectio i Java Programs Outlie Lei Hu Kamra Sartipi {hul4, sartipi}@mcmasterca Departmet of Computig ad Software McMaster Uiversity Caada Motivatio Research Problem Defiitio

More information

On the Accuracy of Vector Metrics for Quality Assessment in Image Filtering

On the Accuracy of Vector Metrics for Quality Assessment in Image Filtering 0th IMEKO TC4 Iteratioal Symposium ad 8th Iteratioal Workshop o ADC Modellig ad Testig Research o Electric ad Electroic Measuremet for the Ecoomic Uptur Beeveto, Italy, September 5-7, 04 O the Accuracy

More information

VALIDATING DIRECTIONAL EDGE-BASED IMAGE FEATURE REPRESENTATIONS IN FACE RECOGNITION BY SPATIAL CORRELATION-BASED CLUSTERING

VALIDATING DIRECTIONAL EDGE-BASED IMAGE FEATURE REPRESENTATIONS IN FACE RECOGNITION BY SPATIAL CORRELATION-BASED CLUSTERING VALIDATING DIRECTIONAL EDGE-BASED IMAGE FEATURE REPRESENTATIONS IN FACE RECOGNITION BY SPATIAL CORRELATION-BASED CLUSTERING Yasufumi Suzuki ad Tadashi Shibata Departmet of Frotier Iformatics, School of

More information

Counting Regions in the Plane and More 1

Counting Regions in the Plane and More 1 Coutig Regios i the Plae ad More 1 by Zvezdelia Stakova Berkeley Math Circle Itermediate I Group September 016 1. Overarchig Problem Problem 1 Regios i a Circle. The vertices of a polygos are arraged o

More information

Lecture Notes 6 Introduction to algorithm analysis CSS 501 Data Structures and Object-Oriented Programming

Lecture Notes 6 Introduction to algorithm analysis CSS 501 Data Structures and Object-Oriented Programming Lecture Notes 6 Itroductio to algorithm aalysis CSS 501 Data Structures ad Object-Orieted Programmig Readig for this lecture: Carrao, Chapter 10 To be covered i this lecture: Itroductio to algorithm aalysis

More information

Handwriting Stroke Extraction Using a New XYTC Transform

Handwriting Stroke Extraction Using a New XYTC Transform Hadwritig Stroke Etractio Usig a New XYTC Trasform Gilles F. Houle 1, Kateria Bliova 1 ad M. Shridhar 1 Computer Scieces Corporatio Uiversity Michiga-Dearbor Abstract: The fudametal represetatio of hadwritig

More information

Title: Robust Registration of Multimodal Remote Sensing Images Based on Structural Similarity.

Title: Robust Registration of Multimodal Remote Sensing Images Based on Structural Similarity. 7 IEEE. Persoal use of this material is permitted. Permissio from IEEE must be obtaied for all other uses, i ay curret or future media, icludig repritig/republishig this material for advertisig or promotioal

More information

Parallel Polygon Approximation Algorithm Targeted at Reconfigurable Multi-Ring Hardware

Parallel Polygon Approximation Algorithm Targeted at Reconfigurable Multi-Ring Hardware Parallel Polygo Approximatio Algorithm Targeted at Recofigurable Multi-Rig Hardware M. Arif Wai* ad Hamid R. Arabia** *Califoria State Uiversity Bakersfield, Califoria, USA **Uiversity of Georgia, Georgia,

More information

EE 584 MACHINE VISION

EE 584 MACHINE VISION METU EE 584 Lecture Notes by A.Aydi ALATAN 0 EE 584 MACHINE VISION Itroductio elatio with other areas Image Formatio & Sesig Projectios Brightess Leses Image Sesig METU EE 584 Lecture Notes by A.Aydi ALATAN

More information

DATA MINING II - 1DL460

DATA MINING II - 1DL460 DATA MINING II - 1DL460 Sprig 2017 A secod course i data miig http://www.it.uu.se/edu/course/homepage/ifoutv2/vt17/ Kjell Orsbor Uppsala Database Laboratory Departmet of Iformatio Techology, Uppsala Uiversity,

More information

Lecture 5. Counting Sort / Radix Sort

Lecture 5. Counting Sort / Radix Sort Lecture 5. Coutig Sort / Radix Sort T. H. Corme, C. E. Leiserso ad R. L. Rivest Itroductio to Algorithms, 3rd Editio, MIT Press, 2009 Sugkyukwa Uiversity Hyuseug Choo choo@skku.edu Copyright 2000-2018

More information

Searching a Russian Document Collection Using English, Chinese and Japanese Queries

Searching a Russian Document Collection Using English, Chinese and Japanese Queries Searchig a Russia Documet Collectio Usig Eglish, Chiese ad Japaese Queries Fredric C. Gey (gey@ucdata.berkeley.edu) UC Data Archive & Techical Assistace Uiversity of Califoria, Berkeley, CA 94720 USA ABSTRACT.

More information

An Efficient Algorithm for Graph Bisection of Triangularizations

An Efficient Algorithm for Graph Bisection of Triangularizations A Efficiet Algorithm for Graph Bisectio of Triagularizatios Gerold Jäger Departmet of Computer Sciece Washigto Uiversity Campus Box 1045 Oe Brookigs Drive St. Louis, Missouri 63130-4899, USA jaegerg@cse.wustl.edu

More information

EE 505. CMOS and BiCMOS Data Conversion Circuits

EE 505. CMOS and BiCMOS Data Conversion Circuits EE 505 CMOS ad BiCMOS Data Coversio Circuits Course Iformatio: Lecture Istructor: Rady Geiger 2133 Coover Voice: 294-7745 e-mail: rlgeiger@iastate.edu WEB: www.radygeiger.org Laboratory Istructor: TBD

More information

A Note on Least-norm Solution of Global WireWarping

A Note on Least-norm Solution of Global WireWarping A Note o Least-orm Solutio of Global WireWarpig Charlie C. L. Wag Departmet of Mechaical ad Automatio Egieerig The Chiese Uiversity of Hog Kog Shati, N.T., Hog Kog E-mail: cwag@mae.cuhk.edu.hk Abstract

More information

Algorithms for Disk Covering Problems with the Most Points

Algorithms for Disk Covering Problems with the Most Points Algorithms for Disk Coverig Problems with the Most Poits Bi Xiao Departmet of Computig Hog Kog Polytechic Uiversity Hug Hom, Kowloo, Hog Kog csbxiao@comp.polyu.edu.hk Qigfeg Zhuge, Yi He, Zili Shao, Edwi

More information

x x 2 x Iput layer = quatity of classificatio mode X T = traspositio matrix The core of such coditioal probability estimatig method is calculatig the

x x 2 x Iput layer = quatity of classificatio mode X T = traspositio matrix The core of such coditioal probability estimatig method is calculatig the COMPARATIVE RESEARCHES ON PROBABILISTIC NEURAL NETWORKS AND MULTI-LAYER PERCEPTRON NETWORKS FOR REMOTE SENSING IMAGE SEGMENTATION Liu Gag a, b, * a School of Electroic Iformatio, Wuha Uiversity, 430079,

More information

Coherent effects of flow- and pressure hull of a generic submarine on target scattering in an active sonar performance model

Coherent effects of flow- and pressure hull of a generic submarine on target scattering in an active sonar performance model Coheret effects of flow- ad pressure hull of a geeric submarie o target scatterig i a active soar performace model P. Schippers TNO-D&V-Uderwater Techology, Oude Waalsdorperweg 63, Post Box 96864, 2509

More information

DETECTION OF LANDSLIDE BLOCK BOUNDARIES BY MEANS OF AN AFFINE COORDINATE TRANSFORMATION

DETECTION OF LANDSLIDE BLOCK BOUNDARIES BY MEANS OF AN AFFINE COORDINATE TRANSFORMATION Proceedigs, 11 th FIG Symposium o Deformatio Measuremets, Satorii, Greece, 2003. DETECTION OF LANDSLIDE BLOCK BOUNDARIES BY MEANS OF AN AFFINE COORDINATE TRANSFORMATION Michaela Haberler, Heribert Kahme

More information

A Trinocular Stereo System for Highway Obstacle Detection

A Trinocular Stereo System for Highway Obstacle Detection A Triocular Stereo System for Highway Obstacle Detectio Todd Williamso ad Charles Thorpe Robotics Istitute Caregie Mello Uiversity Pittsburgh, PA 15213 {Todd.Williamso,Charles.Thorpe}@ri.cmu.edu Abstract

More information

arxiv: v2 [cs.ds] 24 Mar 2018

arxiv: v2 [cs.ds] 24 Mar 2018 Similar Elemets ad Metric Labelig o Complete Graphs arxiv:1803.08037v [cs.ds] 4 Mar 018 Pedro F. Felzeszwalb Brow Uiversity Providece, RI, USA pff@brow.edu March 8, 018 We cosider a problem that ivolves

More information

Polymorph: Morphing Among Multiple Images

Polymorph: Morphing Among Multiple Images Feature Article Polymorph: Morphig Amog Multiple Images Image metamorphosis has prove to be a powerful visual effects tool. May breathtakig examples ow appear i film ad televisio, depictig the fluid trasformatio

More information

The impact of GOP pattern and packet loss on the video quality. of H.264/AVC compression standard

The impact of GOP pattern and packet loss on the video quality. of H.264/AVC compression standard The impact of GOP patter ad packet loss o the video quality of H.264/AVC compressio stadard MIROSLAV UHRINA, JAROSLAV FRNDA, LUKÁŠ ŠEVČÍK, MARTIN VACULÍK Departmet of Telecommuicatios ad Multimedia Uiversity

More information

Empirical Validate C&K Suite for Predict Fault-Proneness of Object-Oriented Classes Developed Using Fuzzy Logic.

Empirical Validate C&K Suite for Predict Fault-Proneness of Object-Oriented Classes Developed Using Fuzzy Logic. Empirical Validate C&K Suite for Predict Fault-Proeess of Object-Orieted Classes Developed Usig Fuzzy Logic. Mohammad Amro 1, Moataz Ahmed 1, Kaaa Faisal 2 1 Iformatio ad Computer Sciece Departmet, Kig

More information

Neural Networks A Model of Boolean Functions

Neural Networks A Model of Boolean Functions Neural Networks A Model of Boolea Fuctios Berd Steibach, Roma Kohut Freiberg Uiversity of Miig ad Techology Istitute of Computer Sciece D-09596 Freiberg, Germay e-mails: steib@iformatik.tu-freiberg.de

More information

Vision & Perception. Simple model: simple reflectance/illumination model. image: x(n 1,n 2 )=i(n 1,n 2 )r(n 1,n 2 ) 0 < r(n 1,n 2 ) < 1

Vision & Perception. Simple model: simple reflectance/illumination model. image: x(n 1,n 2 )=i(n 1,n 2 )r(n 1,n 2 ) 0 < r(n 1,n 2 ) < 1 Visio & Perceptio Simple model: simple reflectace/illumiatio model Eye illumiatio source i( 1, 2 ) image: x( 1, 2 )=i( 1, 2 )r( 1, 2 ) reflectace term r( 1, 2 ) where 0 < i( 1, 2 ) < 0 < r( 1, 2 ) < 1

More information

Structuring Redundancy for Fault Tolerance. CSE 598D: Fault Tolerant Software

Structuring Redundancy for Fault Tolerance. CSE 598D: Fault Tolerant Software Structurig Redudacy for Fault Tolerace CSE 598D: Fault Tolerat Software What do we wat to achieve? Versios Damage Assessmet Versio 1 Error Detectio Iputs Versio 2 Voter Outputs State Restoratio Cotiued

More information

APPLICATION NOTE. Automated Gain Flattening. 1. Experimental Setup. Scope and Overview

APPLICATION NOTE. Automated Gain Flattening. 1. Experimental Setup. Scope and Overview APPLICATION NOTE Automated Gai Flatteig Scope ad Overview A flat optical power spectrum is essetial for optical telecommuicatio sigals. This stems from a eed to balace the chael powers across large distaces.

More information

An Efficient Algorithm for Graph Bisection of Triangularizations

An Efficient Algorithm for Graph Bisection of Triangularizations Applied Mathematical Scieces, Vol. 1, 2007, o. 25, 1203-1215 A Efficiet Algorithm for Graph Bisectio of Triagularizatios Gerold Jäger Departmet of Computer Sciece Washigto Uiversity Campus Box 1045, Oe

More information

Analysis of Server Resource Consumption of Meteorological Satellite Application System Based on Contour Curve

Analysis of Server Resource Consumption of Meteorological Satellite Application System Based on Contour Curve Advaces i Computer, Sigals ad Systems (2018) 2: 19-25 Clausius Scietific Press, Caada Aalysis of Server Resource Cosumptio of Meteorological Satellite Applicatio System Based o Cotour Curve Xiagag Zhao

More information

Texture Image Segmentation Using Without Re-initialization Geodesic Active Contour Model

Texture Image Segmentation Using Without Re-initialization Geodesic Active Contour Model Texture Image Segmetatio Usig Without Re-iitializatio Geodesic Active Cotour Model Kaibi Wag Biazhag Yu Departmet of Electroic ad Iformatio Egieerig, Northwester Polytechical Uiversity, Xi a 71007, Shaaxi

More information

FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS

FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS Prosejit Bose Evagelos Kraakis Pat Mori Yihui Tag School of Computer Sciece, Carleto Uiversity {jit,kraakis,mori,y

More information

FEATURE BASED RECOGNITION OF TRAFFIC VIDEO STREAMS FOR ONLINE ROUTE TRACING

FEATURE BASED RECOGNITION OF TRAFFIC VIDEO STREAMS FOR ONLINE ROUTE TRACING FEATURE BASED RECOGNITION OF TRAFFIC VIDEO STREAMS FOR ONLINE ROUTE TRACING Christoph Busch, Ralf Dörer, Christia Freytag, Heike Ziegler Frauhofer Istitute for Computer Graphics, Computer Graphics Ceter

More information

Load balanced Parallel Prime Number Generator with Sieve of Eratosthenes on Cluster Computers *

Load balanced Parallel Prime Number Generator with Sieve of Eratosthenes on Cluster Computers * Load balaced Parallel Prime umber Geerator with Sieve of Eratosthees o luster omputers * Soowook Hwag*, Kyusik hug**, ad Dogseug Kim* *Departmet of Electrical Egieerig Korea Uiversity Seoul, -, Rep. of

More information

Human Activity Tracking for Wide-Area Surveillance

Human Activity Tracking for Wide-Area Surveillance Huma Activity Trackig for Wide-Area Surveillace Patrick D. O Malley Michael C. Nechyba A. Atoio Arroyo pomalley@mil.ufl.edu echyba@mil.ufl.edu arroyo@mil.ufl.edu Machie Itelligece Laboratory Departmet

More information

Position and Velocity Estimation by Ultrasonic Sensor

Position and Velocity Estimation by Ultrasonic Sensor Positio ad Velocity Estimatio by Ultrasoic Sesor N Ramarao 1, A R Subramayam 2, J Chara Raj 2, Lalith B V 2, Varu K R 2 1 (Faculty of EEE, BMSIT & M, INDIA) 2 (Studets of EEE, BMSIT & M, INDIA) Abstract:

More information

ISSN (Print) Research Article. *Corresponding author Nengfa Hu

ISSN (Print) Research Article. *Corresponding author Nengfa Hu Scholars Joural of Egieerig ad Techology (SJET) Sch. J. Eg. Tech., 2016; 4(5):249-253 Scholars Academic ad Scietific Publisher (A Iteratioal Publisher for Academic ad Scietific Resources) www.saspublisher.com

More information

APPLICATION NOTE PACE1750AE BUILT-IN FUNCTIONS

APPLICATION NOTE PACE1750AE BUILT-IN FUNCTIONS APPLICATION NOTE PACE175AE BUILT-IN UNCTIONS About This Note This applicatio brief is iteded to explai ad demostrate the use of the special fuctios that are built ito the PACE175AE processor. These powerful

More information

Lip Contour Extraction Based on Support Vector Machine

Lip Contour Extraction Based on Support Vector Machine Lip Cotour Extractio Based o Support Vector Machie Author Pa, Xiaosheg, Kog, Jiagpig, Liew, Ala Wee-Chug Published 008 Coferece Title CISP 008 : Proceedigs, First Iteratioal Cogress o Image ad Sigal Processig

More information

How do we evaluate algorithms?

How do we evaluate algorithms? F2 Readig referece: chapter 2 + slides Algorithm complexity Big O ad big Ω To calculate ruig time Aalysis of recursive Algorithms Next time: Litterature: slides mostly The first Algorithm desig methods:

More information

ALU Augmentation for MPEG-4 Repetitive Padding

ALU Augmentation for MPEG-4 Repetitive Padding ALU Augmetatio for MPEG-4 Repetitive Paddig Georgi Kuzmaov Stamatis Vassiliadis Computer Egieerig Lab, Electrical Egieerig Departmet, Faculty of formatio Techology ad Systems, Delft Uiversity of Techology,

More information

ON THE QUALITY OF AUTOMATIC RELATIVE ORIENTATION PROCEDURES

ON THE QUALITY OF AUTOMATIC RELATIVE ORIENTATION PROCEDURES ON THE QUALITY OF AUTOMATIC RELATIVE ORIENTATION PROCEDURES Thomas Läbe, Timo Dickscheid ad Wolfgag Förster Istitute of Geodesy ad Geoiformatio, Departmet of Photogrammetry, Uiversity of Bo laebe@ipb.ui-bo.de,

More information

Homework 1 Solutions MA 522 Fall 2017

Homework 1 Solutions MA 522 Fall 2017 Homework 1 Solutios MA 5 Fall 017 1. Cosider the searchig problem: Iput A sequece of umbers A = [a 1,..., a ] ad a value v. Output A idex i such that v = A[i] or the special value NIL if v does ot appear

More information

CSC 220: Computer Organization Unit 11 Basic Computer Organization and Design

CSC 220: Computer Organization Unit 11 Basic Computer Organization and Design College of Computer ad Iformatio Scieces Departmet of Computer Sciece CSC 220: Computer Orgaizatio Uit 11 Basic Computer Orgaizatio ad Desig 1 For the rest of the semester, we ll focus o computer architecture:

More information

The Closest Line to a Data Set in the Plane. David Gurney Southeastern Louisiana University Hammond, Louisiana

The Closest Line to a Data Set in the Plane. David Gurney Southeastern Louisiana University Hammond, Louisiana The Closest Lie to a Data Set i the Plae David Gurey Southeaster Louisiaa Uiversity Hammod, Louisiaa ABSTRACT This paper looks at three differet measures of distace betwee a lie ad a data set i the plae:

More information

Lecture 6. Lecturer: Ronitt Rubinfeld Scribes: Chen Ziv, Eliav Buchnik, Ophir Arie, Jonathan Gradstein

Lecture 6. Lecturer: Ronitt Rubinfeld Scribes: Chen Ziv, Eliav Buchnik, Ophir Arie, Jonathan Gradstein 068.670 Subliear Time Algorithms November, 0 Lecture 6 Lecturer: Roitt Rubifeld Scribes: Che Ziv, Eliav Buchik, Ophir Arie, Joatha Gradstei Lesso overview. Usig the oracle reductio framework for approximatig

More information

Cubic Polynomial Curves with a Shape Parameter

Cubic Polynomial Curves with a Shape Parameter roceedigs of the th WSEAS Iteratioal Coferece o Robotics Cotrol ad Maufacturig Techology Hagzhou Chia April -8 00 (pp5-70) Cubic olyomial Curves with a Shape arameter MO GUOLIANG ZHAO YANAN Iformatio ad

More information

1.2 Binomial Coefficients and Subsets

1.2 Binomial Coefficients and Subsets 1.2. BINOMIAL COEFFICIENTS AND SUBSETS 13 1.2 Biomial Coefficiets ad Subsets 1.2-1 The loop below is part of a program to determie the umber of triagles formed by poits i the plae. for i =1 to for j =

More information