A Hybrid System for Three-Dimensional Objects Reconstruction from Point-Clouds Based on Ball Pivoting Algorithm and Radial Basis Functions

Size: px
Start display at page:

Download "A Hybrid System for Three-Dimensional Objects Reconstruction from Point-Clouds Based on Ball Pivoting Algorithm and Radial Basis Functions"

Transcription

1 A Hybrd System for Three-Dmensonal Objects Reconstructon from Pont-Clouds Based on Ball Pvotng Algorthm and Radal Bass Functons ASHRAF HUSSEIN Scentfc Computng Department, Faculty of Computer and Informaton Scences, An Shams Unversty, Abbassa, 11566, Caro, Egypt. Abstract:- In ths paper, a new hybrd system for 3D objects reconstructon s presented. Ths system s based on the well-known Ball Pvotng Algorthm (BPA) of Bernardn et al. combned wth the Radal Bass Functon (RBF) model. The BPA can reconstruct surfaces from very large data sets n a lttle executon tme, usng small amount of memory, and exhbts lnear complexty but t produces holes n case of low samplng densty besdes t s not capable to fll ms-regstraton domans. So, the BPA s coupled wth the RBF n an ntegrated system to gan the edge of the RBF as a powerful nterpolator for generatng surface vertces wthn severe holes and/or ms-regstraton domans. Performance and accuracy of the system components are nvestgated through reconstructng benchmark objects. The results obtaned compare favorably wth other emnent algorthms. Fnally, the proposed system exhbted hgh robustness n reconstructng objects from real large clouds wth non-unform samplng and ms-regstratons. Key-Words: - Object modelng, computatonal geometry, surface reconstructon, and surface and object representatons. 1. Introducton There are many applcatons that rely on buldng accurate models of real-world objects such as sculptures, damaged machne parts, archaeologcal artfacts, and terran. Technques for dgtzng objects nclude laser range fndng, mechancal touch probes, and computer vson technques such as depth from stereo. Some of these technques can yeld mllons of 3D pont locatons on the object that s beng dgtzed. Once these ponts have been collected, t s a non-trval task to buld a surface representaton that s fathful to the collected data. Some of the desrable propertes of a surface reconstructon method nclude speed, low memory overhead, fathful reproducton of sharp features, and robustness n the presence of holes and low samplng densty. Fgure (1) llustrates a typcal example for object reconstructon problem. Few technques are avalable for automatcally modelng ths type of data. Mathematcally, the problem can be expressed as follows, gven a set of N ponts P={p 1,,p N }, whch are sampled from some unknown smooth surface U, fnd a mesh M that approxmates the surface U,.e. the mesh and the surface are near everywhere [15]. The output s ntally some mesh (mostly a trangulaton) that has the same general shape of the orgnal surface. If the generated mesh connects the orgnal nput ponts, then t nterpolates the surface. Alternatvely, the mesh may connect generated ponts other than those of the nput set. In the latter case, the mesh s sad to approxmate the surface. A common classfcaton crteron of surface reconstructon technques s to categorze the technque as ether a surface nterpolaton algorthm or a surface approxmaton one [17]. (a) (b) Fg. 1: Example of a typcal object reconstructon problem (a) Raw data of a statue wth natve holes and ms-regstratons. (b) Reconstructed object usng the present system.

2 One of the frst surface nterpolaton algorthms s the one presented by Edelsbrunner and Mücke n [11], whch uses the heurstc alpha-shapes. Amenta and Bern [1] used the propertes of the medal axs and Vorono dagram to ntroduce the Crust algorthm, whch was the frst guaranteed surface reconstructon algorthm. They extracted an approxmated trangulated surface from the Delaunay trangulaton of the nput set. Amenta et al. [] descrbed the co-cone term and used t n ther Cocone algorthm. The Cocone computes the Delaunay trangulaton of the nput set and then selects canddate trangles to construct the surface. Dey and Goswam [9] modfed the Cocone nto a new algorthm, the Tght Cocone, to solve problem nstances of specfc domans that requre hole-free surfaces. Kollur et al. [16] ntroduced a noseresstant algorthm for watertght surfaces. The latter algorthm uses a varant of spectral graph parttonng to select trangles from the Delaunay trangulaton to construct the surface. All the aforementoned algorthms can be descrbed as Vorono based algorthms and they generally suffer from slow performance and dffculty n handlng large data. To overcome ths problem, Dey et al. [1] parttoned the entre sample space nto smaller clusters usng octree subdvson. Then, they appled the Cocone algorthm on each of these clusters separately, and they called ths algorthm the Super Cocone. An nterestng nterpolaton non-vorono based algorthm s the Ball Pvotng algorthm, presented by Bernardn et al. [5]. Ths algorthm ncrementally reconstructs surfaces by usng a rollng ball to construct trangles one by one and append them to the surface. Typcal examples of surface approxmaton algorthms are the algorthms presented by Hoppe et al. [15] and by Curless and Levoy [8]. They used the normal orentatons to estmate the tangent planes and approxmate the local surface. Another surface approxmaton algorthm was ntroduced by Bossonat and Cazals [4] that s based on natural neghbors nterpolaton and represents the surface mplctly as a zero-set of some pseudo-dstance functon. Gop et al. [13] suggested an algorthm that uses a plane projecton technque where the neghbors of a sample pont s computed by projectng ts close ponts onto ts tangent plane. Amenta et al. [3] suggested usng the power dagram (dual of weghted Delaunay trangulaton) n ther Power Crust algorthm. Although the Power Crust algorthm s a Vorono based algorthm, but -unlke the algorthms belongng to ths famly- t s classfed as a surface approxmaton algorthm. Another famly of surface approxmaton algorthms uses the mplct surface approach, for example [6], [] and [7], whch s sutable for reconstructng pont clouds wth local problems lke holes, low samplng densty, and ms-regstraton but not sutable for processng large clouds. To adapt these algorthms for large clouds, local support or parttonng based solutons are used (for example [18], [19] and [1]). Stll, mesh nterpolaton based methods are preferable for surface reconstructon than those based on mesh approxmaton because the former gve better accuracy and have better performance from both computatonal tme and memory usage ponts of vew. On the other hand, mesh approxmaton based methods that utlze mplct surface approach have the edge of robustness n reconstructng surfaces from cloud of ponts natvely have holes, low samplng densty and msregstraton. So, ganng the advantages of the two types of reconstructon methods can be obtaned by ntegratng methods from both of them. In ths work, a hybrd system s developed by combnng two well-known methods n order to obtan a reconstructon scheme for large data sets n the presence of low samplng densty and/or ms-regstraton domans. The Ball Pvotng algorthm [5] s used as an effcent tool for surface reconstructon whle the Radal Bass Functons (RBF) [6] s used as a powerful nterpolaton tool. The paper s organzed as follows: secton provdes the theoretcal backgrounds of the technque. In secton 3, the mplementaton s presented showng the practcal choces to obtan an easy, effcent and robust reconstructon algorthm. Secton 4 presents a study for the performance and accuracy of the present system and the applcaton of reconstructng some realstc data that have natve holes and ms-regstraton domans, and n secton 5 the work s concluded.. Theoretcal backgrounds.1. Ball pvotng algorthm The approach of Bernardn, Mttleman and Rushmeer [5] mtates the ball eraser used n α- shapes. Assume that the sampled data set P s dense enough that a ρ-ball, a ball of radus ρ, can not pass through the surface wthout touchng sample ponts. The algorthm starts by placng ρ- ball n contact wth three sample ponts. Keepng contact wth two of these ntal ponts, the ball s

3 pvotng untl t touches another pont. The pvotng operaton s depcted n Fgure (). y rejected f the dot product of the trangle normal wth the surface normal (n the vertex) s negatve, see Fgure (3). l k s k s jo σ k c k ρ γ m σ,σ j Fg. : The ball pvotng operaton. The ball of radus ρ sttng on the trangle τ formed from the vertces σ, σ j, σ s pvotng around the edge m here perpendcular to the mage. The center of the ball c j rotates around the edge m and descrbes a crcular trajectory γ. The radus of the trajectory s c j m. Whle pvotng the ball hts the pont σ k. The crcle s k s a ntersecton of ρ-ball centered at σ k wth z =. The crcle s j s the ntersecton of the pvotng ball wth z =. (mage courtesy of Fausto Bernardn [5]). The ρ-ball s pvoted around each edge of the current mesh boundary. Trplets of ponts that the ball contacts form new trangles. The set of trangles formed whle the ρ-ball walks on the surface form the nterpolatng mesh. The ball pvotng algorthm (BPA) s closely related to α- shapes. In fact, every trangle τ computed by the ρ- ball walk obvously has an empty smallest open ball b τ whose radus s less than ρ. Thus, the BPA computes a subset of the -faces of the ρ-shape of P. These faces are also a subset of the -skeleton of the three-dmensonal Delaunay trangulaton of the pont set. The surface reconstructed by the BPA retans some of the qualtes of alpha-shapes: It has provable reconstructon guarantees under certan samplng assumptons and an ntutvely smple geometrc meanng. The nput data ponts are augmented wth approxmate surface normals computed from the range maps. The surface normals are used to dsambguate cases that occur when dealng wth mssng or nosy data. Areas of densty hgher than ρ present no problem to the algorthm, but mssng ponts create holes that cannot be flled by the pvotng ball. Any post-process hole-fllng algorthm could be employed; n partcular, BPA can be appled multple tmes wth ncreasng ball rad. To handle possble ambgutes, the normals are used as stated above. When pvotng around a boundary edge, the ball can touch an unused pont lyng close to the surface. Therefore a trangle s τ c jo n σ o x Fg. 3: The normal check (a D case). The rghtmost pont s boundary and the ball pvots around ts edge and touches a hdden pont, but the dot product of the trangle normal wth the vertex normal s negatve, so the trangle s rejected. (mage courtesy of Fausto Bernardn [5]) Ths algorthm s very effcent because t exhbts lnear tme and storage requrements. However ts dsadvantages are the selecton of the radus ρ of the ρ-ball and ncapablty to handle data that have natve problems lke low samplng densty, holes and ms-regstratons... Radal bass functons Gven the set of N dstnct ponts P= {p 1,,p N }of dmenson d: p k R d, and the set of values {h 1,,h N }, we want to fnd a functon f: R d R wth f ( p ) = h (1) In order to obtan a radal bass functon reconstructon of the pont set P, a functon f satsfyng the equaton f k ( p) = ω φ( p p ) + π ( p) =1, () has to be found. We denote here p, p j the Eucldean dstance, ω the weghts, φ : R R a bass functon, and π a polynomal of degree m π p = c π p dependng on the choce of φ. ( ) ( ) wth { π } Q α a bass n the d-dmensonal null α = 1 space contanng all real-valued polynomals n d varables and of order at most m, hence m + d Q =. d The bass functon φ has to be condtonally postve defnte [1], and some popular choces proposed n the lterature are shown below: bharmonc φ ( r ) = r wth π of degree 1 (3) 3 pseudo-cubc φ ( r ) = r wth π of degree 1 (4) 3 trharmonc φ( r ) = r wth π of degree (5)

4 The bharmonc and trharmonc splnes are known as the smoothest nterpolators n the sense that they mnmze certan energy functons and nterpolate the data [6]. So, the bharmonc splne s used n the present work. Gven a set of nodes N { p} R 3 N and a set of functon values { h } = 1 R, = 1 the bharmonc RBF f(p) satsfes the nterpolaton f p = h and mnmzes condtons ( ) f ( p) f ( p) f ( p) f f = 3 R x f + x y + ( p) f ( p) f ( p) y + x z + z + y z dx (6) s a measure of the energy n the second dervatve of f. As we have an under-determned system wth N+Q unknowns (ω and c) and N equatons, socalled natural addtonal constrants for the coeffcents ω are added, so that ω c1 = ωc =... = ωcq = (7) The equatons (1), (), and (7) determne the followng lnear system: Ax = b (8) T Φ P A = P (9) Φj = φ ( p, p j ) = N j = N P = π α p = N = α... Q x = ω ω,...,, c, c,..., ( ) [ ] T 1, ω N 1 c Q b = h1, h,..., h N,, 1443,..., (1) Q tmes The soluton vector x s composed of the weghts ω and the polynomal coeffcents c for equaton () and represents a soluton of the nterpolaton problem gven by (1). 3. Reconstructon scheme 3.1. Hybrd reconstructon To overcome the dsadvantages of the Ball Pvotng Algorthm, a hybrd system s developed by combnng BPA and RBF and works as shown T n Fgure (4). The system calls the BPA teratvely by defnng the trangulaton doman. The RBF soluton vectors (whch contan the weghts ω and the polynomal coeffcents c ) are computed once after the dentfcaton of holes and/or msregstraton domans and ther neghborhood zones from the ntal reconstructed surface obtaned by the BPA or gven as an nput to the present system. Each teraton, holes are dentfed (whch le wthn that dentfed prevously) then the RBF s employed as nterpolator to unformly generate surface ponts wthn the holes and the BPA s used to trangulate the generated ponts. The process of callng BPA and generatng sampled ponts usng RBF surface nterpolator s repeated untl reachng a perfect surface. End Yes Cloud of ponts or mesh contanng holes to be flled RBF_Sol_Vec_Comp= No Defne BPA workng doman Apply BPA Identfy holes loops Nholes> Yes RBF_Computed=1 No Compute RBF soluton vector for each hole loop RBF_Sol_Vec_Comp=1 Apply nterpolaton For each hole doman Man teraton loop Fg. 4: The hybrd system for 3D object reconstructon 3.. Hole or ms-regstraton dentfcaton The applcaton searches for every edge whch les n only one trangle and constructs a lst of these edges. Startng wth a seed vertex whch les on any edge wthn the constructed lst, the applcaton can trace a seres of edges untl t dentfes a closed loop defnng a hole. The dentfed loop of edges s removed from the lst. The process of tracng and removng edges s repeated untl dentfyng all the holes.

5 3.3. RBF computatons After the dentfcaton of the closed loops of holes and/or ms-regstraton domans, the RBF computatons are performed for each of these loops. For each loop, the neghborhood zone depth should be defned prory (the zone depth s a user selected value as a rato of the hole loop largest dagonal). The vertces wthn ths zone are remarked by a smple tracng process through edges untl reachng the predefned zone depth as shown n Fgure (5). The remarked vertces are employed n the RBF computatons to get the soluton vector (whch contans the weghts and the polynomal coeffcents) usng equaton (8). The z-components of these vertces wll be consdered as the functon values n the rght hand sde of the equaton. Ths process s appled once to get the soluton vector for each hole or ms-regstraton doman. The nterpolaton process s appled to fll each hole or ms-regstraton doman by generatng surface vertces usng dfferent x and y coordnates values, wth nearly unform dstrbuton, then the correspondng z-coordnates are obtaned by the same equaton usng the computed soluton vector. Identfed hole boundary loop Identfed hole boundary loop b Largest dagonal Largest dagonal Fg. 5: Tracng of neghborhood zone vertces Selectng x and y coordnates values s the key parameter n generatng surface vertces wth nearly unform dstrbuton wthn each hole or ms-regstraton doman. If the dstrbuton of the generated surface vertces s not suffcently unform, more than one teraton wll be needed untl reachng perfect surface as mentoned n Fgure (4). In such case, the nterpolaton process wll be repeated to generate surface vertces wthn the dentfed holes usng the same soluton vector computed at the frst teraton. c a d g e f 4. Applcatons and results All results presented n ths secton were performed on Pentum IV.4 GHz PC computng platform wth 51 MB RAM runnng wndows XP. Source codes are developed usng Mcrosoft vsual C/C++ verson 6.. Also, Intel Math Kernel Lbrary was employed to mprove the performance of matrx operatons BPA performance verfcaton The performance of the BPA s verfed aganst other emnent algorthms, namely Crust, Power Crust and Tght Cocone, n reconstructng bench mark objects of [14] wth dfferent resolutons. These algorthms are developed on the same platform and usng the same programmng tools to get proper conclusons. Fgures (6.a) to (6.c) show the varaton of computatonal tme versus model sze for the three selected objects for comparson (Bunny, Dragon and Happy Buddha) whle Fgures (6.d) to (6.f) show the correspondng varaton of memory usage. All the algorthms exhbted lnear performance ndependent of the object shape but BPA bested others from both computatonal tme and memory usage ponts of vew. Fgure (7) shows the reconstructed meshes usng the dfferent algorthms for the 1889 vertces Bunny model. The Crust output (Fgure 7.a) s good but t s not a manfold. Besdes beng watertght, the Power Crust output (Fgure 7.b) may be consdered the best, except that the mesh s very dense and not trangular. The Tght Cocone output (Fgure 7.c) s also watertght, but shows a problem at the left ear end (a trangle s not connected to the rest of the ear). The Ball Pvotng output (Fgure 7.d) shows also problems at both of ears whch can be consdered as holes n the obtaned mesh. 4.. Hybrd system applcaton and accuracy assessment The system s tested va the reconstructon of the Londog object used by Ohtake et al. [18], Fgure (8.a), after makng an artfcal hole n ts head to assess the accuracy of the present system n fllng holes and ms-regstraton domans. The orgnal model contans 493 unformly sampled ponts. The mesh s reconstructed ntally by applyng the BPA usng a sngle ball (ts radus s estmated as the average samplng dstance) n 8. secs. Then, the artfcal hole s successfully dentfed by ts

6 boundary loop as shown n Fgure (8.b). Dfferent values for the hole neghborhood zone depth are consdered. The values consdered are 5%, 7% and 1% of the hole loop largest dagonal (HLD). The RBF soluton vector s computed for the three cases n 9, 16 and 7 secs usng the drect method [6]. Fnally, the teratve process between the BPA and RBF s appled untl achevng a perfect surface. Only one teraton s performed n 6 secs to get surface wth fnal qualty as shown n the same fgure. The accuracy of the present system n fllng holes and ms-regstraton domans s assessed by comparng the reconstructed surface for the artfcal hole and the orgnal surface (for each of the consdered values of the hole-neghborhood zone depth). The dfference between the two surfaces at each pont wth the same x and y coordnates s consdered as the dfference n the magntude of the z-coordnate. The contours representng the dstrbuton of the dfference between surfaces are shown n Fgures (8.c) to (8.e) for the three consdered values. It s found that ncreasng the hole-neghborhood zone depth decreasng the accuracy of fllng and ncreasng the smoothness Reconstructon of realstc data The robustness of the present system n reconstructon of 3D objects from large cloud of ponts that have natve holes and/or msregstraton s confrmed usng a realstc cloud of ponts for a statue of an ancent Egyptan man. The model contans 6,461 ponts wth many msregstraton domans n ts head, face, shoulders, hands and the base as shown n Fgure (9.a). The mesh s reconstructed ntally by applyng the BPA usng sngle ball n 91 secs. Then, holes and msregstraton domans are dentfed. The value of the neghborhood zone depth for the holes and msregstraton domans s taken as 5% of HLD of the correspondng holes or ms-regstraton domans. The RBF soluton vectors for all the dentfed holes are computed n 194 secs. Then, three teratons are performed n 89 secs to get surface wth fnal qualty. Fgures (9.b), (9.c) and (9.d) compare the reconstructed object usng the present system aganst the results of the reconstructon usng the Tght Cocone algorthm and the multlevel partton of unty mplcts of [18]. It can be shown that the Tght Cocone, Fgure (9.b), exhbted flat fllng for the ms-regstraton domans whle the reconstructed object usng the mult-level partton of unty mplcts of [18] suffers from spkes n the ms-regstraton domans besdes losses n the object detals as shown n Fgure (9.c). On the other hand, the present system, Fgure (9.d), exhbted curved fllng for the msregstraton domans wth contnuty n the normals dstrbuton at ther boundares as shown n Fgure (9. e). Hence, the present system s more sutable for reconstructon of complcated surfaces from large clouds sufferng from holes and/or msregstraton domans. 5. Conclusons A hybrd system for surface reconstructon based on the Ball Pvotng Algorthm (BPA) and Radal Bass Functons (RBF) was developed and presented. The present system employs the Ball Pvotng algorthm as a surface reconstructon tool has the capablty of processng very large data sets n a lttle executon tme and usng small amount of memory. To overcome the weakness of the BPA n the presence of holes, ms-regstratons and low sample densty, the RBF s employed by the system to unformly nterpolate ponts representng the mssng surface parts. The RBF soluton vector s computed once after dentfyng the holes and/or ms-regstraton domans from the ntal reconstructed surface obtaned by the BPA. Then, the process of the nterpolaton and callng the BPA to trangulate the dentfed holes and/or msregstraton domans are appled teratvely untl reachng perfect surface. The accuracy and robustness were verfed n reconstructng surfaces from realstc data wth non-unform samplng and ms-regstratons. References: [1] Amenta N. and Bern M., Surface reconstructon by Vorono flterng, In Proc. of ACM SIGGRAPH 98, 1998, pp [] Amenta N., Cho S., Dey T. and Leekha N., A smple algorthm for homeomorphc surface reconstructon, In Proc. of 16 th ACM Sym. on Comp. Geom.,, pp [3] Amenta N., Cho S. and Kollur R., The power crust, unon of balls, and the medal axs transform, In proc. of the 6 th ACM Sym. on sold modelng and applcatons, 1, pp [4] Bossonat J. and Cazals F., Smooth surface reconstructon va natural neghbor nterpolaton of dstance functons, In Proc. of the 16 th annual sym. on Comp. Geom.,, pp. 3-3.

7 [5] Bernardn F., Mttleman J., Rushmeer H., Slva C. and Taubn G., The ball-pvotng algorthm for surface reconstructon, IEEE Trans. n Vs. and Comp. Graph. 5, 4, 1999, pp [6] Carr J., Beatson R., Cherre J., Mtchell T., Frght W., McCallum B., and T. Evans., Reconstructon and representaton of 3D objects wth radal bass functons, In Proc. of ACM SIGGRAPH 1, 1, pp [7] Carr J., Beatson R., McCallum B., Frght W., McLennan T., and Mtchell T., Smooth surface reconstructon from nosy range data, In Proc. of Graphte 3, 3, pp [8] Curless B. and Levoy M., A volumetrc method for buldng complex models from range mages, In Proc. of SIGGRAPH 96, 1996, pp [9] Dey T. and Goswam S., Tght Cocone: A water-tght surface reconstructor, In Proc. of 8 th ACM Sym. on sold modelng and applcatons, 3, pp [1] Dey T., Gesen J. and Hudson J., Delaunay based shape reconstructon from large data, In Proc. Of IEEE Sym. n parallel and large data vsualzaton and graphcs, 1, pp [11] Edelsbrunner H. and Mücke D., Threedmensonal alpha shapes, ACM TOG 13, 1, 1994, pp [1] Franke R., Scattered data nterpolaton: tests of some methods, Mathematcs of Computaton 38, 157, 198, pp [13] Gop M., Krshnan S. and Slva C., Surface reconstructon based on lower dmensonal localzed Delaunay trangulaton, In Proc. of Eurographcs 19, 3,. tme n secs memory n MB crust Crust power Power Crust tght Tght Cocone pvot Ball Pvot x 1 crust Crust power Power Crust tght Tght Cocone pvot Ball Pvot nput sze tme n secs crust Crust power Power Crust tght Tght Cocone pvot Ball Pvot [14] ep [15] Hoppe H., Derose T., Duchamp T., McDonald J. and Stuetzle W., Surface reconstructon from unorganzed ponts, In Proc. of SIGGRAPH 9, 199, pp [16] Kollur R., Shewchuk J., O'Bren J., Spectral surface reconstructon from nosy pont clouds., In Proc. of Eurographcs Geometry Processng Sym., SGP4, 4, pp [17] Peng L. and Shamsuddn S., 3D object reconstructon and representaton usng neural networks, In Proc. of the nd nternatonal conference on computer graphcs and nteractve technques n Australasa and Southe East Asa, 4, pp [18] Ohtake Y., Belyaev A., Alexa M., Turk G., and Sedel H., Mult-level partton of unty mplcts, In Proc. of ACM SIGGRAPH 3, 3, pp [19] Ohtake Y., Belyaev A., and Sedel H., Multscale approach to 3D scattered data nterpolaton wth compactly supported bass functons, In Proc. of Shape Modelng Internatonal, 3, pp [] Turk G. and O bren J., Modellng wth mplct surfaces that nterpolate, ACM TOG 1, 4,, pp [1] Tobor I., Reuter P., and Schlck C., Effcent reconstructon of large scattered geometrc datasets usng the partton of unty and radal bass functons, WSCG 4 1,3, x 1 nput sze 9 8 Crust crust 7 Power power Crust 6 Tght tghtcocone 5 Ball pvot Pvot x 1 nput sze (a) (b) (c) x 1 nput sze memory n MB crust Crust power Power Crust tght Tght Cocone Ball Pvot pvot x 1 nput sze tme n secs memory n MB Crust crust power Power Crust tght Tght Cocone pvot Ball Pvot nput sze x 1 (d) (e) (f) Fg. 6: Comparson between dfferent algorthms from computatonal tme and memory usage ponts of vew. (a, d) Bunny Model. (b, e) Dragon Model. (c, f) Happy Buddha Model.

8 (a) (b) (c) (d) Fg. 7: Reconstructed Stanford Bunny from 1889 vertces usng dfferent algorthms. (a) Crust. (b) Power Crust. (c) Tght Cocone. (d) BPA (a) Identfed hole after BPA reconstructon (b) Fnal qualty reconstructon (c) 5% HLD (d) 7% HLD (e) 1% HLD Fg. 8: Accuracy assessment of the present system for three consdered values of the hole-neghborhood zone depth. (a) Londog after BPA reconstructon and hole dentfcaton. (b) fnal qualty reconstructon usng neghborhood zone depth of 5% HLD (c,d,e) contours of the offset dstance between the generated surface and the orgnal surface for the three consdered values of the hole-neghborhood zone depth.

9 (a) (b) (c) (d) (e) Fg. 9: Reconstructon of real scanned data wth natve holes and ms-regstraton domans of an ancent Egyptan man statue. (a) the scanned cloud of 6,461 ponts wth ms-regstraton domans ndcated by the red arrows. (b) reconstructed surface usng the Tght Cocone algorthm. (c) reconstructed surface usng mult-level partton of unty mplcts of [18] (d) reconstructed surface usng the present system.(e ) surface-normal dstrbuton over the reconstructed object usng the present system.

Radial Basis Functions

Radial Basis Functions Radal Bass Functons Mesh Reconstructon Input: pont cloud Output: water-tght manfold mesh Explct Connectvty estmaton Implct Sgned dstance functon estmaton Image from: Reconstructon and Representaton of

More information

Simplification of 3D Meshes

Simplification of 3D Meshes Smplfcaton of 3D Meshes Addy Ngan /4/00 Outlne Motvaton Taxonomy of smplfcaton methods Hoppe et al, Mesh optmzaton Hoppe, Progressve meshes Smplfcaton of 3D Meshes 1 Motvaton Hgh detaled meshes becomng

More information

Wavefront Reconstructor

Wavefront Reconstructor A Dstrbuted Smplex B-Splne Based Wavefront Reconstructor Coen de Vsser and Mchel Verhaegen 14-12-201212 2012 Delft Unversty of Technology Contents Introducton Wavefront reconstructon usng Smplex B-Splnes

More information

High-Boost Mesh Filtering for 3-D Shape Enhancement

High-Boost Mesh Filtering for 3-D Shape Enhancement Hgh-Boost Mesh Flterng for 3-D Shape Enhancement Hrokazu Yagou Λ Alexander Belyaev y Damng We z Λ y z ; ; Shape Modelng Laboratory, Unversty of Azu, Azu-Wakamatsu 965-8580 Japan y Computer Graphcs Group,

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

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

S1 Note. Basis functions.

S1 Note. Basis functions. S1 Note. Bass functons. Contents Types of bass functons...1 The Fourer bass...2 B-splne bass...3 Power and type I error rates wth dfferent numbers of bass functons...4 Table S1. Smulaton results of type

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

3D vector computer graphics

3D vector computer graphics 3D vector computer graphcs Paolo Varagnolo: freelance engneer Padova Aprl 2016 Prvate Practce ----------------------------------- 1. Introducton Vector 3D model representaton n computer graphcs requres

More information

Surface reconstruction from sparse data by a multiscale volumetric approach

Surface reconstruction from sparse data by a multiscale volumetric approach Proceedngs of the 5th WSEAS Int. Conf. on Sgnal Processng, Computatonal Geometry & Artfcal Vson, Malta, September 5-7, 2005 (pp35-40) Surface reconstructon from sparse data by a multscale volumetrc approach

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

Feature Reduction and Selection

Feature Reduction and Selection Feature Reducton and Selecton Dr. Shuang LIANG School of Software Engneerng TongJ Unversty Fall, 2012 Today s Topcs Introducton Problems of Dmensonalty Feature Reducton Statstc methods Prncpal Components

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

Type-2 Fuzzy Non-uniform Rational B-spline Model with Type-2 Fuzzy Data

Type-2 Fuzzy Non-uniform Rational B-spline Model with Type-2 Fuzzy Data Malaysan Journal of Mathematcal Scences 11(S) Aprl : 35 46 (2017) Specal Issue: The 2nd Internatonal Conference and Workshop on Mathematcal Analyss (ICWOMA 2016) MALAYSIAN JOURNAL OF MATHEMATICAL SCIENCES

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

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

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

Interpolation of the Irregular Curve Network of Ship Hull Form Using Subdivision Surfaces

Interpolation of the Irregular Curve Network of Ship Hull Form Using Subdivision Surfaces 7 Interpolaton of the Irregular Curve Network of Shp Hull Form Usng Subdvson Surfaces Kyu-Yeul Lee, Doo-Yeoun Cho and Tae-Wan Km Seoul Natonal Unversty, kylee@snu.ac.kr,whendus@snu.ac.kr,taewan}@snu.ac.kr

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

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

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

NAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics

NAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics Introducton G10 NAG Fortran Lbrary Chapter Introducton G10 Smoothng n Statstcs Contents 1 Scope of the Chapter... 2 2 Background to the Problems... 2 2.1 Smoothng Methods... 2 2.2 Smoothng Splnes and Regresson

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

Chapter 6 Programmng the fnte element method Inow turn to the man subject of ths book: The mplementaton of the fnte element algorthm n computer programs. In order to make my dscusson as straghtforward

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

CS 534: Computer Vision Model Fitting

CS 534: Computer Vision Model Fitting CS 534: Computer Vson Model Fttng Sprng 004 Ahmed Elgammal Dept of Computer Scence CS 534 Model Fttng - 1 Outlnes Model fttng s mportant Least-squares fttng Maxmum lkelhood estmaton MAP estmaton Robust

More information

Hermite Splines in Lie Groups as Products of Geodesics

Hermite Splines in Lie Groups as Products of Geodesics Hermte Splnes n Le Groups as Products of Geodescs Ethan Eade Updated May 28, 2017 1 Introducton 1.1 Goal Ths document defnes a curve n the Le group G parametrzed by tme and by structural parameters n the

More information

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009.

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009. Farrukh Jabeen Algorthms 51 Assgnment #2 Due Date: June 15, 29. Assgnment # 2 Chapter 3 Dscrete Fourer Transforms Implement the FFT for the DFT. Descrbed n sectons 3.1 and 3.2. Delverables: 1. Concse descrpton

More information

Multi-Resolution Geometric Fusion

Multi-Resolution Geometric Fusion Internatonal Conference on Recent Advances n 3-D Dgtal Imagng and Modellng, Ottawa, Canada May 12 15, 1997 Mult-Resoluton Geometrc Fuson Adran Hlton and John Illngworth Centre for Vson, Speech and Sgnal

More information

Smoothing Spline ANOVA for variable screening

Smoothing Spline ANOVA for variable screening Smoothng Splne ANOVA for varable screenng a useful tool for metamodels tranng and mult-objectve optmzaton L. Rcco, E. Rgon, A. Turco Outlne RSM Introducton Possble couplng Test case MOO MOO wth Game Theory

More information

Accounting for the Use of Different Length Scale Factors in x, y and z Directions

Accounting for the Use of Different Length Scale Factors in x, y and z Directions 1 Accountng for the Use of Dfferent Length Scale Factors n x, y and z Drectons Taha Soch (taha.soch@kcl.ac.uk) Imagng Scences & Bomedcal Engneerng, Kng s College London, The Rayne Insttute, St Thomas Hosptal,

More information

Scan Conversion & Shading

Scan Conversion & Shading Scan Converson & Shadng Thomas Funkhouser Prnceton Unversty C0S 426, Fall 1999 3D Renderng Ppelne (for drect llumnaton) 3D Prmtves 3D Modelng Coordnates Modelng Transformaton 3D World Coordnates Lghtng

More information

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements Module 3: Element Propertes Lecture : Lagrange and Serendpty Elements 5 In last lecture note, the nterpolaton functons are derved on the bass of assumed polynomal from Pascal s trangle for the fled varable.

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

Mesh Editing in ROI with Dual Laplacian

Mesh Editing in ROI with Dual Laplacian Mesh Edtng n ROI wth Dual Laplacan Luo Qong, Lu Bo, Ma Zhan-guo, Zhang Hong-bn College of Computer Scence, Beng Unversty of Technology, Chna lqngng@sohu.com, lubo@but.edu.cn,mzgsy@63.com,zhb@publc.bta.net.cn

More information

LECTURE : MANIFOLD LEARNING

LECTURE : MANIFOLD LEARNING LECTURE : MANIFOLD LEARNING Rta Osadchy Some sldes are due to L.Saul, V. C. Raykar, N. Verma Topcs PCA MDS IsoMap LLE EgenMaps Done! Dmensonalty Reducton Data representaton Inputs are real-valued vectors

More information

Scan Conversion & Shading

Scan Conversion & Shading 1 3D Renderng Ppelne (for drect llumnaton) 2 Scan Converson & Shadng Adam Fnkelsten Prnceton Unversty C0S 426, Fall 2001 3DPrmtves 3D Modelng Coordnates Modelng Transformaton 3D World Coordnates Lghtng

More information

Smooth Approximation to Surface Meshes of Arbitrary Topology with Locally Blended Radial Basis Functions

Smooth Approximation to Surface Meshes of Arbitrary Topology with Locally Blended Radial Basis Functions 587 Smooth Approxmaton to Surface eshes of Arbtrary Topology wth Locally Blended Radal Bass Functons ngyong Pang 1,, Weyn a 1, Zhgeng Pan and Fuyan Zhang 1 Cty Unversty of Hong Kong, mewma@ctyu.edu.hk

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

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

Electrical analysis of light-weight, triangular weave reflector antennas

Electrical analysis of light-weight, triangular weave reflector antennas Electrcal analyss of lght-weght, trangular weave reflector antennas Knud Pontoppdan TICRA Laederstraede 34 DK-121 Copenhagen K Denmark Emal: kp@tcra.com INTRODUCTION The new lght-weght reflector antenna

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

Computer Animation and Visualisation. Lecture 4. Rigging / Skinning

Computer Animation and Visualisation. Lecture 4. Rigging / Skinning Computer Anmaton and Vsualsaton Lecture 4. Rggng / Sknnng Taku Komura Overvew Sknnng / Rggng Background knowledge Lnear Blendng How to decde weghts? Example-based Method Anatomcal models Sknnng Assume

More information

Reading. 14. Subdivision curves. Recommended:

Reading. 14. Subdivision curves. Recommended: eadng ecommended: Stollntz, Deose, and Salesn. Wavelets for Computer Graphcs: heory and Applcatons, 996, secton 6.-6., A.5. 4. Subdvson curves Note: there s an error n Stollntz, et al., secton A.5. Equaton

More information

Support Vector Machines

Support Vector Machines /9/207 MIST.6060 Busness Intellgence and Data Mnng What are Support Vector Machnes? Support Vector Machnes Support Vector Machnes (SVMs) are supervsed learnng technques that analyze data and recognze patterns.

More information

A Five-Point Subdivision Scheme with Two Parameters and a Four-Point Shape-Preserving Scheme

A Five-Point Subdivision Scheme with Two Parameters and a Four-Point Shape-Preserving Scheme Mathematcal and Computatonal Applcatons Artcle A Fve-Pont Subdvson Scheme wth Two Parameters and a Four-Pont Shape-Preservng Scheme Jeqng Tan,2, Bo Wang, * and Jun Sh School of Mathematcs, Hefe Unversty

More information

S.P.H. : A SOLUTION TO AVOID USING EROSION CRITERION?

S.P.H. : A SOLUTION TO AVOID USING EROSION CRITERION? S.P.H. : A SOLUTION TO AVOID USING EROSION CRITERION? Célne GALLET ENSICA 1 place Emle Bloun 31056 TOULOUSE CEDEX e-mal :cgallet@ensca.fr Jean Luc LACOME DYNALIS Immeuble AEROPOLE - Bat 1 5, Avenue Albert

More information

THE PULL-PUSH ALGORITHM REVISITED

THE PULL-PUSH ALGORITHM REVISITED THE PULL-PUSH ALGORITHM REVISITED Improvements, Computaton of Pont Denstes, and GPU Implementaton Martn Kraus Computer Graphcs & Vsualzaton Group, Technsche Unverstät München, Boltzmannstraße 3, 85748

More information

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters Proper Choce of Data Used for the Estmaton of Datum Transformaton Parameters Hakan S. KUTOGLU, Turkey Key words: Coordnate systems; transformaton; estmaton, relablty. SUMMARY Advances n technologes and

More information

Support Vector Machines

Support Vector Machines Support Vector Machnes Decson surface s a hyperplane (lne n 2D) n feature space (smlar to the Perceptron) Arguably, the most mportant recent dscovery n machne learnng In a nutshell: map the data to a predetermned

More information

An Entropy-Based Approach to Integrated Information Needs Assessment

An Entropy-Based Approach to Integrated Information Needs Assessment Dstrbuton Statement A: Approved for publc release; dstrbuton s unlmted. An Entropy-Based Approach to ntegrated nformaton Needs Assessment June 8, 2004 Wllam J. Farrell Lockheed Martn Advanced Technology

More information

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Proceedngs of the Wnter Smulaton Conference M E Kuhl, N M Steger, F B Armstrong, and J A Jones, eds A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Mark W Brantley Chun-Hung

More information

Lobachevsky State University of Nizhni Novgorod. Polyhedron. Quick Start Guide

Lobachevsky State University of Nizhni Novgorod. Polyhedron. Quick Start Guide Lobachevsky State Unversty of Nzhn Novgorod Polyhedron Quck Start Gude Nzhn Novgorod 2016 Contents Specfcaton of Polyhedron software... 3 Theoretcal background... 4 1. Interface of Polyhedron... 6 1.1.

More information

CSCI 104 Sorting Algorithms. Mark Redekopp David Kempe

CSCI 104 Sorting Algorithms. Mark Redekopp David Kempe CSCI 104 Sortng Algorthms Mark Redekopp Davd Kempe Algorthm Effcency SORTING 2 Sortng If we have an unordered lst, sequental search becomes our only choce If we wll perform a lot of searches t may be benefcal

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

Determining the Optimal Bandwidth Based on Multi-criterion Fusion

Determining the Optimal Bandwidth Based on Multi-criterion Fusion Proceedngs of 01 4th Internatonal Conference on Machne Learnng and Computng IPCSIT vol. 5 (01) (01) IACSIT Press, Sngapore Determnng the Optmal Bandwdth Based on Mult-crteron Fuson Ha-L Lang 1+, Xan-Mn

More information

Octree Subdivision Using Coplanar Criterion for Hierarchical Point Simplification

Octree Subdivision Using Coplanar Criterion for Hierarchical Point Simplification Octree Subdvson Usng Coplanar Crteron for Herarchcal Pont Smplfcaton Pa-Feng Lee 1, Chen-Hsng Chang 1, Jun-Lng Tseng 2, Bn-Shyan Jong 3, and Tsong-Wuu Ln 4 1 Dept. of Electronc Engneerng, Chung Yuan Chrstan

More information

Image Representation & Visualization Basic Imaging Algorithms Shape Representation and Analysis. outline

Image Representation & Visualization Basic Imaging Algorithms Shape Representation and Analysis. outline mage Vsualzaton mage Vsualzaton mage Representaton & Vsualzaton Basc magng Algorthms Shape Representaton and Analyss outlne mage Representaton & Vsualzaton Basc magng Algorthms Shape Representaton and

More information

A Newton-Type Method for Constrained Least-Squares Data-Fitting with Easy-to-Control Rational Curves

A Newton-Type Method for Constrained Least-Squares Data-Fitting with Easy-to-Control Rational Curves A Newton-Type Method for Constraned Least-Squares Data-Fttng wth Easy-to-Control Ratonal Curves G. Cascola a, L. Roman b, a Department of Mathematcs, Unversty of Bologna, P.zza d Porta San Donato 5, 4017

More information

A Robust Method for Estimating the Fundamental Matrix

A Robust Method for Estimating the Fundamental Matrix Proc. VIIth Dgtal Image Computng: Technques and Applcatons, Sun C., Talbot H., Ourseln S. and Adraansen T. (Eds.), 0- Dec. 003, Sydney A Robust Method for Estmatng the Fundamental Matrx C.L. Feng and Y.S.

More information

Feature-Preserving Mesh Denoising via Bilateral Normal Filtering

Feature-Preserving Mesh Denoising via Bilateral Normal Filtering Feature-Preservng Mesh Denosng va Blateral Normal Flterng Ka-Wah Lee, Wen-Png Wang Computer Graphcs Group Department of Computer Scence, The Unversty of Hong Kong kwlee@cs.hku.hk, wenpng@cs.hku.hk Abstract

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

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

Face Recognition University at Buffalo CSE666 Lecture Slides Resources:

Face Recognition University at Buffalo CSE666 Lecture Slides Resources: Face Recognton Unversty at Buffalo CSE666 Lecture Sldes Resources: http://www.face-rec.org/algorthms/ Overvew of face recognton algorthms Correlaton - Pxel based correspondence between two face mages Structural

More information

Angle-Independent 3D Reconstruction. Ji Zhang Mireille Boutin Daniel Aliaga

Angle-Independent 3D Reconstruction. Ji Zhang Mireille Boutin Daniel Aliaga Angle-Independent 3D Reconstructon J Zhang Mrelle Boutn Danel Alaga Goal: Structure from Moton To reconstruct the 3D geometry of a scene from a set of pctures (e.g. a move of the scene pont reconstructon

More information

For instance, ; the five basic number-sets are increasingly more n A B & B A A = B (1)

For instance, ; the five basic number-sets are increasingly more n A B & B A A = B (1) Secton 1.2 Subsets and the Boolean operatons on sets If every element of the set A s an element of the set B, we say that A s a subset of B, or that A s contaned n B, or that B contans A, and we wrte A

More information

Active Contours/Snakes

Active Contours/Snakes Actve Contours/Snakes Erkut Erdem Acknowledgement: The sldes are adapted from the sldes prepared by K. Grauman of Unversty of Texas at Austn Fttng: Edges vs. boundares Edges useful sgnal to ndcate occludng

More information

Exercises (Part 4) Introduction to R UCLA/CCPR. John Fox, February 2005

Exercises (Part 4) Introduction to R UCLA/CCPR. John Fox, February 2005 Exercses (Part 4) Introducton to R UCLA/CCPR John Fox, February 2005 1. A challengng problem: Iterated weghted least squares (IWLS) s a standard method of fttng generalzed lnear models to data. As descrbed

More information

Mathematics 256 a course in differential equations for engineering students

Mathematics 256 a course in differential equations for engineering students Mathematcs 56 a course n dfferental equatons for engneerng students Chapter 5. More effcent methods of numercal soluton Euler s method s qute neffcent. Because the error s essentally proportonal to the

More information

The Theory and Application of an Adaptive Moving Least. Squares for Non-uniform Samples

The Theory and Application of an Adaptive Moving Least. Squares for Non-uniform Samples Xanpng Huang, Qng Tan, Janfe Mao, L Jang, Ronghua Lang The Theory and Applcaton of an Adaptve Movng Least Squares for Non-unform Samples Xanpng Huang, Qng Tan, Janfe Mao*, L Jang, Ronghua Lang College

More information

Meta-heuristics for Multidimensional Knapsack Problems

Meta-heuristics for Multidimensional Knapsack Problems 2012 4th Internatonal Conference on Computer Research and Development IPCSIT vol.39 (2012) (2012) IACSIT Press, Sngapore Meta-heurstcs for Multdmensonal Knapsack Problems Zhbao Man + Computer Scence Department,

More information

An Accurate Evaluation of Integrals in Convex and Non convex Polygonal Domain by Twelve Node Quadrilateral Finite Element Method

An Accurate Evaluation of Integrals in Convex and Non convex Polygonal Domain by Twelve Node Quadrilateral Finite Element Method Internatonal Journal of Computatonal and Appled Mathematcs. ISSN 89-4966 Volume, Number (07), pp. 33-4 Research Inda Publcatons http://www.rpublcaton.com An Accurate Evaluaton of Integrals n Convex and

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

Some Tutorial about the Project. Computer Graphics

Some Tutorial about the Project. Computer Graphics Some Tutoral about the Project Lecture 6 Rastersaton, Antalasng, Texture Mappng, I have already covered all the topcs needed to fnsh the 1 st practcal Today, I wll brefly explan how to start workng on

More information

Modeling, Manipulating, and Visualizing Continuous Volumetric Data: A Novel Spline-based Approach

Modeling, Manipulating, and Visualizing Continuous Volumetric Data: A Novel Spline-based Approach Modelng, Manpulatng, and Vsualzng Contnuous Volumetrc Data: A Novel Splne-based Approach Jng Hua Center for Vsual Computng, Department of Computer Scence SUNY at Stony Brook Talk Outlne Introducton and

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

Skew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach

Skew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach Angle Estmaton and Correcton of Hand Wrtten, Textual and Large areas of Non-Textual Document Images: A Novel Approach D.R.Ramesh Babu Pyush M Kumat Mahesh D Dhannawat PES Insttute of Technology Research

More information

A Volumetric Approach for Interactive 3D Modeling

A Volumetric Approach for Interactive 3D Modeling A Volumetrc Approach for Interactve 3D Modelng Dragan Tubć Patrck Hébert Computer Vson and Systems Laboratory Laval Unversty, Ste-Foy, Québec, Canada, G1K 7P4 Dens Laurendeau E-mal: (tdragan, hebert, laurendeau)@gel.ulaval.ca

More information

Vectorization of Image Outlines Using Rational Spline and Genetic Algorithm

Vectorization of Image Outlines Using Rational Spline and Genetic Algorithm 01 Internatonal Conference on Image, Vson and Computng (ICIVC 01) IPCSIT vol. 50 (01) (01) IACSIT Press, Sngapore DOI: 10.776/IPCSIT.01.V50.4 Vectorzaton of Image Outlnes Usng Ratonal Splne and Genetc

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

Interactive Rendering of Translucent Objects

Interactive Rendering of Translucent Objects Interactve Renderng of Translucent Objects Hendrk Lensch Mchael Goesele Phlppe Bekaert Jan Kautz Marcus Magnor Jochen Lang Hans-Peter Sedel 2003 Presented By: Mark Rubelmann Outlne Motvaton Background

More information

Adaptive Selection of Rendering Primitives for Point Clouds of Large Scale Environments

Adaptive Selection of Rendering Primitives for Point Clouds of Large Scale Environments Adaptve Selecton of Renderng Prmtves for Pont Clouds of Large Scale Envronments Taash Maeno, Hroa Date and Satosh Kana 3 Graduate School of Informaton Scence and Technology, Hoado Unversty, Japan t_maeno@sdm.ss.st.houda.ac.jp,

More information

Sorting Review. Sorting. Comparison Sorting. CSE 680 Prof. Roger Crawfis. Assumptions

Sorting Review. Sorting. Comparison Sorting. CSE 680 Prof. Roger Crawfis. Assumptions Sortng Revew Introducton to Algorthms Qucksort CSE 680 Prof. Roger Crawfs Inserton Sort T(n) = Θ(n 2 ) In-place Merge Sort T(n) = Θ(n lg(n)) Not n-place Selecton Sort (from homework) T(n) = Θ(n 2 ) In-place

More information

Solutions to Programming Assignment Five Interpolation and Numerical Differentiation

Solutions to Programming Assignment Five Interpolation and Numerical Differentiation College of Engneerng and Coputer Scence Mechancal Engneerng Departent Mechancal Engneerng 309 Nuercal Analyss of Engneerng Systes Sprng 04 Nuber: 537 Instructor: Larry Caretto Solutons to Prograng Assgnent

More information

y and the total sum of

y and the total sum of Lnear regresson Testng for non-lnearty In analytcal chemstry, lnear regresson s commonly used n the constructon of calbraton functons requred for analytcal technques such as gas chromatography, atomc absorpton

More information

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

Quality Improvement Algorithm for Tetrahedral Mesh Based on Optimal Delaunay Triangulation

Quality Improvement Algorithm for Tetrahedral Mesh Based on Optimal Delaunay Triangulation Intellgent Informaton Management, 013, 5, 191-195 Publshed Onlne November 013 (http://www.scrp.org/journal/m) http://dx.do.org/10.36/m.013.5601 Qualty Improvement Algorthm for Tetrahedral Mesh Based on

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

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

Machine Learning 9. week

Machine Learning 9. week Machne Learnng 9. week Mappng Concept Radal Bass Functons (RBF) RBF Networks 1 Mappng It s probably the best scenaro for the classfcaton of two dataset s to separate them lnearly. As you see n the below

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

High resolution 3D Tau-p transform by matching pursuit Weiping Cao* and Warren S. Ross, Shearwater GeoServices

High resolution 3D Tau-p transform by matching pursuit Weiping Cao* and Warren S. Ross, Shearwater GeoServices Hgh resoluton 3D Tau-p transform by matchng pursut Wepng Cao* and Warren S. Ross, Shearwater GeoServces Summary The 3D Tau-p transform s of vtal sgnfcance for processng sesmc data acqured wth modern wde

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

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

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

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS ARPN Journal of Engneerng and Appled Scences 006-017 Asan Research Publshng Network (ARPN). All rghts reserved. NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS Igor Grgoryev, Svetlana

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

Topology Design using LS-TaSC Version 2 and LS-DYNA

Topology Design using LS-TaSC Version 2 and LS-DYNA Topology Desgn usng LS-TaSC Verson 2 and LS-DYNA Wllem Roux Lvermore Software Technology Corporaton, Lvermore, CA, USA Abstract Ths paper gves an overvew of LS-TaSC verson 2, a topology optmzaton tool

More information

SVM-based Learning for Multiple Model Estimation

SVM-based Learning for Multiple Model Estimation SVM-based Learnng for Multple Model Estmaton Vladmr Cherkassky and Yunqan Ma Department of Electrcal and Computer Engneerng Unversty of Mnnesota Mnneapols, MN 55455 {cherkass,myq}@ece.umn.edu Abstract:

More information

Ecient Computation of the Most Probable Motion from Fuzzy. Moshe Ben-Ezra Shmuel Peleg Michael Werman. The Hebrew University of Jerusalem

Ecient Computation of the Most Probable Motion from Fuzzy. Moshe Ben-Ezra Shmuel Peleg Michael Werman. The Hebrew University of Jerusalem Ecent Computaton of the Most Probable Moton from Fuzzy Correspondences Moshe Ben-Ezra Shmuel Peleg Mchael Werman Insttute of Computer Scence The Hebrew Unversty of Jerusalem 91904 Jerusalem, Israel Emal:

More information