IMPROVING AND EXTENDING THE INFORMATION ON PRINCIPAL COMPONENT ANALYSIS FOR LOCAL NEIGHBORHOODS IN 3D POINT CLOUDS

Size: px
Start display at page:

Download "IMPROVING AND EXTENDING THE INFORMATION ON PRINCIPAL COMPONENT ANALYSIS FOR LOCAL NEIGHBORHOODS IN 3D POINT CLOUDS"

Transcription

1 IMPROVING AND EXTENDING THE INFORMATION ON PRINCIPAL COMPONENT ANALYSIS FOR LOCAL NEIGHBORHOODS IN 3D POINT CLOUDS Davd Belton Cooperatve Research Centre for Spatal Informaton (CRC-SI) The Insttute for Geoscence Research (TIGeR) Department of Spatal Scences, Curtn Unversty of Technology Perth WA, Australa - d.belton@curtn.edu.au Commsson V, WG 3 KEY WORDS: Laser Scannng, Pont clouds, Curvature Approxmaton, Surface Normal Estmaton, Prncpal Component Analyss ABSTRACT: Prncpal Component Analyss (PCA) s often utlsed n pont cloud processng as provdes an effcent method to approxmate local pont propertes through the examnaton of the local neghbourhoods. Ths process does sometmes suffer from the assumpton that the neghbourhood contans only a sngle surface, when t may contan multple dscrete surface enttes, as well as relatng the propertes from PCA to real world attrbutes. Ths paper wll present two methods. The frst s a correcton method to flter out the presence of multple surfaces through an teratve process. The second s to combne the PCA preformed on the neghbourhood of pont coordnates and normal approxmatons n order to estmate the radus of curvature n the maxmum and mnmum curvature drectons.. INTORDUCTION Pont cloud processng reles on the analyss and examnaton of dfferent observed attrbutes such as poston, ntensty and colour. Although these attrbutes are sampled drectly from the laser scanner, attrbutes are more often derved from the examnaton of the local neghbourhood surroundng a pont of nterest. These estmated attrbutes, whch nclude curvature, surface normal and geometrc surface propertes, are of great mportance n pont cloud-processng procedures such as for surface classfcaton and segmentaton. Often these attrbutes are estmated wth the use of Prncpal Component Analyss (PCA) performed on the local neghbourhoods of ponts, as t effcently retreves the local propertes of a neghbourhood (Gumhold et al., 00). Some common uses have ncluded approxmatng the normal drecton (Mtra et al., 004), fttng frst order planar surfaces (Wengarten et al., 003), approxmatng surface curvature (Pauly et al., 00), defnng the tensors for tensor votng (Tong et al., 004) and provdng a local pont coordnate systems (Danels et al., 007). There are two problems that can occur when usng PCA. The frst s that a neghbourhood may contan multple dscrete surface enttes. For most attrbutes, they are calculated under the assumpton that there s only one surface structured. The affect of multple surfaces can cause bases n the attrbutes e.g. the surface normal approxmaton. Whle deceasng the sze of the neghbourhood may help n reducng the probablty that a neghbourhood contans more than one sampled surface, the neghbourhood needs to be of suffcent sze n order to reduce the effect of random errors and nose. The other problem s how to relate the PCA results to the surface attrbutes. Often nformaton can be lost, such as the approxmaton of curvature through surface varaton (Pauly et al., 00) where the comparable level of curvature s ndcated, but there s no drectonal component or unt of measurement assocated wth the approxmaton. It s the am of ths paper to present an teratve method of adjustng the neghbourhood to remove effects of multple surface enttes. In addton, a formula between the egenvalues of the PCA and the surface propertes such as the radus and drecton of local curvature wll be presented. From ths formula, the maxmum and mnmum curvature drectons can be also calculated n a closed form soluton and ths nformaton provdes a novel and detaled nformaton of the neghbourhood of the pont of nterest.. PRINCIPAL COMPONENT ANALYSIS PCA s performed by frst calculatng the covarance matrx Σ by the formula: Σ = ( x c0 )( = x c ) () T 0 where x s defned as the vector form of the poston of the th pont n the neghbourhood contanng the nearest ponts and c 0 represents the centrod of the neghbourhood calculated as the mean of the neghbourhood. Snce Σ s a symmetrc and postve sem-defnte matrx, t can be decompled by egenvalue decomposton such that the real postve egenvalues, λ 0, λ and λ, along wth the correspondng 477

2 The Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences. Vol. XXXVII. Part B5. Bejng 008 egenvectors e 0, e and e form an orthogonal bass of the neghbourhood n R 3 (Golub and Loan, 989). respectvely. An nternal relatonshp of x to x 0 wll be equvalent to an external relatonshp of x 0 to x. The covarance matrx Σ can be decomposed as follows: Σ = = 0 λ T ee () where λ 0 λ λ. Note that egenvectors e represent the prncpal components, wth the correspondng egenvalues λ denotng the sgnfcance of each component n the form of the varance n these drectons (Golub and Loan, 989). For a local neghbourhood of a pont cloud, e 0 approxmates the the local surface normal, wth e and e approxmatng the tangental plane through c 0 (Pauly et al., 00). Ths result s equvalent to a frst order least squares plane ft (Shaarj, 998). 3. IMPROVING NEIGHBOURHOOD SELECTION The ntal neghbourhood can be selected n varous ways. The problem s how to ensure that the assumpton of the neghbourhood contanng only one surface entty s vald. Technques exst that remove multple surfaces or reduce them down to a sngle domnate surface. Many are based on random samplng technques, such as RANSAC (RANdom SAmplng Consensus) and ts varants (Bolle and Vemur, 99), outler detecton (Danuser and Strer, 998), votng methods (Page et al., 00) and flterng (Tang et al., 007). Ths secton ams to present a method that wll teratvely converge to the correct soluton for a neghborhood by adjustng the weghts of ponts wthn a neghbourhood dependng on the lelhood of that pont beng sampled from the domnate surface entty sampled wthn the neghbourhood. Ths wll be done by examnng the PCA of the neghbourhoods and, usng statstcal sgnfcance testng, determne and adjust weghts teratvely for each pont untl the method converges to a stable soluton (.e. the weghts reman constant). It wll be demonstrated that the soluton wll have an unform weghtng for those ponts that are determned to belong to a domnate surface structure present n the neghbourhood, and zero for those that do not. The frst stage of the proposed method s outlnng how to determne the relatonshp between two ponts wthn a neghbourhood and whether they belong to the same surface entty. Ths s separated nto two models: an nternal relatonshp between a pont and the neghbourhood beng corrected, and an external relatonshp between the pont of nterest and the neghbourhood of another pont. (a) Fgure : (a) Example of an nternal relatonshp of x to x 0. The threshold s defned n red by Eq. 5 wth all ponts nsde consdered to have an nternal relatonshp. (b) Example of an external relatonshp of x to x 0. The threshold s defned n red by Eq. 6. The man concept of ths method s that a geometrc attrbute of a pont x 0, whch s related to ts underlyng surface, s dependent on ts nternal and external relatonshp wth other surroundng ponts. In other words, the surroundng neghbourhood N 0 for the pont x 0 should not only reflect the attrbutes wthn the neghbourhood, but x 0 should also reflect the attrbutes for the neghbourhood around x f they are to be consdered to belong to the same surface entty. To determne these relatonshps, two defntons of dstance are utlsed. Let c 0 and n 0 denote the centrod and normal approxmaton for neghbourhood N 0 around pont x 0. The equaton for the dstance for the nternal relatonshp s defned as: dst r (b) = ( c0 x ) no (3) wth c 0 beng calculated from the mean of the neghbourhood and n 0 s specfed through PCA. In a smlar manner, let c and n denote the centrod and normal approxmaton for neghbourhood N around pont x. The equaton for the dstance for the external relatonshp s defned as: dst er = ( c x0 ) n (4) Agan, c s calculated from the mean of the neghbourhood and n s specfed through PCA. In order to determne f an nternal or external relatonshp exsts, the Boolean operators can be defned respectvely as: 3. Internal and external neghbourhood relaton between ponts An nternal relatonshp s defned as the relaton of a pont x to the neghbourhood N 0 surroundng a pont of nterest x 0 gven that x N 0. Conversely, an external relatonshp s defned as the relaton of a pont of nterest x 0 to the neghbourhood N surroundng x, gven that x N 0. Illustraton of the nternal and external relatonshp s gven n Fgure (a) and Fgure (b) true r = false dst dst for the nternal relatonshp and: r r t > t α υ, s0 α υ, s0 (5) 478

3 The Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences. Vol. XXXVII. Part B5. Bejng 008 true r = false dst dst r r t > t α υ, s0 α υ, s0 for the external relatonshp. Ths Boolean operators come from a statstcally sgnfcance test to determne f the relatonshp s lely to exst. s 0 and s are the error n the approxmate normal drecton for neghbourhood N 0 and N respectvely. These values can be set as λ 0 from the PCA of the respectve neghbourhoods. The value t s the test statstc from the t- dstrbuton wth υ degrees of freedom and a sgnfcance factor of α. 3. Iteratve generaton of pont membershp weghtng (6) so that the summaton of the weghts equal unty. p s then used as the weght n the recalculaton of the covarance matrx for the next teraton. 3.3 D case example In ths secton, a D example of an ntersecton wll be presented. The example s presented n Fgure. As can be seen, the surface normals for the ponts near the edges are ntally perturbed away from the normal drecton of the surface they are sampled from as the neghbourhood s affected by more than one dscrete surface structure. As defned method s appled, the weghts are teratvely updated untl they become stable,.e. that successve teratons do not effect the soluton and p p. The results are shown n Fgure 3. PCA s performed to obtan the ntal approxmaton for the varance and normal drectons. The proposed method wll use a slghtly modfed verson of the covarance matrx formula such that: Σ = = p ( x c )( T 0 x c0 ) (7) p s the weght of pont x n the neghbourhood. Intally, the ponts wll all be equally weghted as p = /, wth beng the number of ponts n the neghbourhood. In a smlar manner, the formula for the centrod value wll be modfed to: c 0 = = p x (8) Fgure : Example of correcton appled to a D ntersecton. The blue lnes represent the ntal normal approxmaton and red lnes represent the corrected normal approxmatons. If a pont p s not related ether nternally or externally to pont p 0, t s lely that t s not sampled from the same surface and the weght s decreased. Conversely, f a pont p has an nternal and external relatonshp to pont p 0, then t s lely that they belong to the same surface, and the weght s ncreased. If there s only one relatonshp, then t s lely that at least one of the neghbourhoods (N 0 or N ) s affected by multply surfaces, and as such the weghs are left as t s untl the neghbourhoods become more refned. From ths, the rules for updatng the weghts are specfed as: p' = p p + δ δ p ( r = true) and ( ( r = false) and ( er er otherwse = = true false where p are the adjusted weghts and δ s the small change to modfy t by. If p becomes a negatve, then t s set to be zero n order to ensure all weghts are non-negatve. The new weghts are then normalsed by: = ) ) (9) p" = p' p' (0) Fgure 3: Angle of the normal orentaton. The values for the surfaces should be approxmately -45 and 45 degrees. The blue lnes represent the orentaton of the ntal normal approxmaton and red lnes represent the corrected values. If the nternal relatonshps were solely used to update the weghts, the ponts whose neghbourhoods are only mldly affected by another surface structure are corrected as t behaves smlar to an outler removal process. If only the external relatonshps are used, whle the results are smlar to when both are used, the process can become unstable wth all the ponts n the neghbourhoods beng removed. If the weghts of the ponts are examned at every teraton, as shown n Fgure 4, t can be seen that they stablsed wth ether a zero value or an unform value for those that are non-zero. It should be noted that f the pont closest to the ntersecton dd not get corrected n order to conform to one of the surfaces. Ths s because the surfaces are equally represented n the neghbourhood and therefore one cannot be consdered better than the other. The normal can be forced to algn to one of the 479

4 The Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences. Vol. XXXVII. Part B5. Bejng 008 surfaces by replacng the centrod c 0 wth x 0. Ths nstablty caused n the covarance matrx allows t to algn to one of the surfaces; however t can also affect the normal approxmaton detrmentally and provde based solutons. (a) (b) Fgure 6: (a) shows the Gaussan sphere of the uncorrected normal drectons and (b) the Gaussan sphere of the corrected normal drectons. The colour ndcates the densty of normal drectons from blue representng zero to red representng n excess of a hundred. Fgure 4: Trend of the weghts for ponts n a neghbourhood as the teratons progress. The neghbourhood s affected by the presence of multple surfaces wth the ponts belongng to the domnant surface tendng to a value of 0.077, and the others tendng to a value of zero. On examnng those ponts near a surface ntersecton Fgure 8(a) shows the orentaton angles for the uncorrected approxmatons and Fgure 8(b) for the corrected approxmatons. Ths llustrates an ncrease n the accuracy of normal algnment after the correcton wth approxmately 90% of the edge ponts now beng algned to wthn 5 degrees of ther correct orentaton. 3.4 Practcal Example Ths secton wll demonstrate the outlned correcton procedure as appled to a 3D pont cloud. The pont cloud, presented n Fgure 5, s scanned from a door arch wth a Leca ScanStaton wth a nomnal pont spacng of 0.0m. The correcton procedure s appled to the pont cloud wth a neghbourhood sze of 30 and the threshold for determnng nternal and external relatonshps set at a sgnfcance level of α = 0.5. (a) (b) Fgure 7: Hstograms for edge ponts of the orentaton angles for the normal drectons. (a) s the uncorrected normal approxmatons and (b) s the corrected normal approxmatons. Peas n the hstogram denote orentaton of the surface present n the pont cloud. 4. APPROXIMATING CURVATURE Fgure 5: Pont cloud sampled from a secton of a door archway wth a Leca Scanstaton. In Fgure 6(a), the ntal normal approxmatons are dsplay on a Gaussan sphere. The clusters on the sphere represent the presence of surfaces wth that normal orentaton. The strpng effect occurrng between the clusters represents normal approxmatons beng effected by more than one surface. Fgure 6(b) shows the Gaussan sphere of the corrected normal approxmatons usng the proposed method. As can be seen, the strppng effect s sgnfcantly reduced as those ponts affected by more than one surface are corrected to algn wth the domnant surface element n the neghbourhood. The neghbourhood weghts are stablsed wthn 50 teratons of the procedure. There are a varety of technques to approxmate curvature such as surface fttng (Besl and Jan, 988), varaton n surface normals (Jang et al., 005), tensor votng (Tong et al., 004), and angles between neghbourhood members (Dyn et al., 00). Two of the methods that nvolve PCA are gven n Pauly et al. (00) and Jang et al. (005). Pauly et al. (00) uses the percentage of total populaton varaton n the normal to determne the surface varaton as a measure of curvature for a local neghbourhood. However, there s no unt or drecton gven to ths curvature approxmaton. Jang et al. (005) provdes an approxmaton of curvature based on the varaton of the approxmate normal drectons n a local neghbourhood. Whle ths measure as a drectonal component, t does not have a unt of measurement. The advantages of both are ther robust nature and fast computatonal tme. A method wll now be presented to combne the PCA of the coordnates and the normal drectons for a neghbourhood to determne the prncpal drectons of maxmum and mnmum curvature, n addton to the radus of curvature approxmaton n these drectons. 480

5 The Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences. Vol. XXXVII. Part B5. Bejng Prncpal component analyss on pont coordnates The propertes that can be derved from PCA have been descrbed prevously. In the case of curvature, the value of λ 0 descrbes that amount of surface varaton n the normal drecton. Snce curvature can cause a varaton n the normal drecton, λ 0 can be used to measure the level of curvature. In most cases λ 0 s dvded by the total varaton to provde an approxmaton (Pauly et al., 00), otherwse the measure wll be dependant on the span of the neghbourhood (Belton and Lcht, 006). It s the span of the neghbourhood that s of nterest n ths method, and s descrbed by λ and λ. In the conc secton dsplayed n Fgure 8(a), f the dstance d and the angle θ s nown, then t becomes a smple matter of determnng the radus of curvature by the followng equaton: sn θ = d r () and d can be approxmated statstcally usng confdence nterval as: λ d = s () wth λ beng the standard devaton n ths drecton calculated from PCA of the neghbourhood, and s beng a scale factor or number of standard devatons on ether sde of the mean that covers the span of the ponts. Ths leaves the only unnown n the formula as θ. that λ 0 (n) λ (n) λ (n). In ths case, λ (n) represents the varaton of the normal drecton along the drecton of maxmum curvature (denoted by e (n) ) and λ (n) represents the varaton of the normal drecton along the drecton of mnmum curvature (denoted by e (n) ). The conc secton for the normals, dsplayed n Fgure 8(b) can be examned n a smlar manner to the conc secton for the pont coordnates. In ths case, θ s solved as: (n) snθ = d (4) wth d (n) beng approxmated as: d t λ = (5) (n) wth λ beng the standard devaton n ths drecton calculated from PCA of the normal, and t beng a the number of standard devatons that covers the span of the normal varaton. 4.3 Radus of curvature approxmaton If the value of θ s the same n Fgure 8(a) and Fgure 8(b), then t becomes a smple matter of solvng the radus of curvature. Assumng that there s only a sngle curved surface, by examnng Fgure 9, t can be seen that the neghbourhood for the normal drectons s a scaled down verson of the neghbourhood of pont coordnates by a factor of r. The reason for ths s that the scalng of the coordnate values would occur along the normal drecton for each pont. Therefore, the nformaton for the PCA on the neghbourhood of pont normals can be used to solve for the value of θ. (a) (b) Fgure 8: (a) Conc secton for the neghbourhood of pont coordnates. (b) Conc secton for the neghbourhood of pont normal drectons. 4. Prncpal component analyss on normal estmaton As llustrated n Jang et al. (005), the PCA performed on the normal drectons of a neghbourhood can also provde an approxmaton of curvature. In ths case, the covarance matrx Σ s defned as: Σ = ( )( = T n μ n μ ) (3) wth n denotng the normal drecton and μ (n) denotng the mean of the normal drectons. The results from the egenvalue decomposton wth the egenvalues λ (n) (n) 0, λ and λ (n), along wth the correspondng egenvectors e (n) (n) 0, e and e (n) such Fgure 9: Neghbourhood of pont normals overlad on the neghbourhood of pont coordnates. Depcts how neghbourhood of coordnates s scaled only along the normal drecton of each pont to get the neghbourhood of the normals. Combnng Eq. and Eq. 4 and rearrangng, the resultng equaton for the radus of curvature s: s λ r = (6) t λ 48

6 The Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences. Vol. XXXVII. Part B5. Bejng 008 If t s assumed that they follow the same dstrbuton (whch s vald snce one s a scaled verson of the other), the values for s and t wll be equal, thus cancellng out to leave the approxmate radus of curvature as follows: r = λ λ (7) For the D case, ths s smple as the prncpal components wll be nomnally algned between the PCA on the neghbourhood of coordnates and normal. In the 3D case, they are often not algned. Therefore, n order to use the approxmaton, the varance n the neghbourhood of pont coordnates must be determned for the drecton specfed by e (n) and e (n), the two prncple curvature drectons. Ths can be done by usng the ellpsod defned by the egenvector decomposton (Golub and Loan, 989) to determne the varance n these drectons. If λ and λ denote the varance of neghbourhood of pont coordnates n the drectons of e (n) and e (n) respectvely, then radus of curvature n the mnmum curvature and maxmum curvature drecton can now be defned respectvely as: r r mn ' = λ λ (8) max ' 4.4 Practcal Applcaton = λ λ (9) Ths secton wll present the approxmate radus of curvature measure appled to a practcal data set. The pont cloud used s shown n Fgure 0 and was captured wth a Leca The results are shown n Table. r s the radus of curvature obtaned through the ppe fttng routne n Cyclone, μ r s the mean value for the approxmated radus of curvature on for the ppe secton and σ r s the standard devaton of the approxmated values. Whle smulated experments perform well, the nose n a practcal data set causes errors and varatons n the results. The results may be mproved by applyng a smoothng flter beforehand. 5. CONCLUSION AND FUTURE WORK PCA s an mportant tool for dervng the local pont propertes through the examnaton of a local neghbourhood. In ths paper, two methods were presented to utlse the nformaton ganed from PCA. The frst method was an teratve correcton procedure appled to a neghbourhood to flter out ponts deemed lely to belong to a dfferent surface. The second was a method to combne the PCA performed on the pont coordnates wth the PCA performed on the pont normal approxmatons to approxmate the maxmum and mnmum curvature drectons and the radus of curvature n each drecton. Whle these results are stll prelmnary, there s an ndcaton that these methods could be used to retreve nformaton for procedures such as classfcaton and segmentaton. Future wor wll focus on examnng the error propagaton of pont nose to ncrease the accuracy of the nformaton. ACKNOWLEDGEMENTS Ths wor has been supported by Curtn Unversty of Technology, the Cooperatve Research Centre for Spatal Informaton, who actvtes are funded by the Australan Commonwealth's Cooperatve Research Centres Programme, and The Insttute for Geoscence Research (TIGeR). REFERENCES Belton, D. and Lcht, D. D., 006. Classfcaton and segmentaton of terrestral laser scanner pont clouds usng local varance nformaton. Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences XXXVI(part 5), pp Besl, P. J. and Jan, R. C., 988. Segmentaton through varable-order surface fttng. IEEE Transactons on Pattern Analyss and Machne Intellgence 0(), pp Bolle, R. M. and Vemur, B. C., 99. On three-dmensonal surface reconstructon methods. IEEE Transactons on Pattern Analyss and Machne Intellgence 3(), pp. 3. Danels, J. D., Ha, L., Ochotta, T. and Slva, C. T., 007. Robust smooth feature extracton from pont clouds. In: IEEE Internatonal Conference on Shape Modelng and Applcatons 007 (SMI 07), Lyon, France, pp Fgure 0: Pont cloud sampled from an ndustral scene contanng multple ppe sectons. Colours are done based on a crude cluster analyss to dentfy dfferent ppe rad. Ppe Label r μ r σ r 0.5m 0.40m 0.047m 0.30m 0.40m 0.098m m 0.600m 0.000m m 0.83m 0.075m m 0.646m m Table : Results for the approxmated radus of curvature. Danuser, C. and Strer, M., 998. Parametrc model fttng: From nler characterzaton to outler detecton. IEEE Transacton on Pattern Anaylss and MAchne Intellgence 0(), pp Dyn, N., Hormann, K., Km, S. J. and Levn, D., 00. Optmzng 3D trangulatons usng dscrete curvature analyss. In: Proceedngs of Mathematcal methods for curves and surfaces (Oslo 000), pp Golub, G. H. and Loan, C. F. V., 989. Matrx Computatons, nd edn, Johns Hopns Press, Baltmore, MD. Gumhold, S., Wang, X. and MacLeod, R., 00. Feature extracton from pont clouds. In: 0th Internatonal Meshng Roundtable. 48

7 The Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences. Vol. XXXVII. Part B5. Bejng 008 Jang, J., Zhang, Z. and Mng, Y., 005. Data segmentaton for geometrc feautre extracton from ldar pont clouds. Geoscence and Remote Sensng Symposum, 005. IGARSS 05, pp Mtra, M. J., Nguyen, A. and Gubas, L., 004. Estmatng surface normals n nosy pont cloud data. Internatonal Journal of Computatonal Geometry and Applcatons 4(4,5), pp Page, D. L., Sun, Y., Koschan, A. F., Pa, J. and Abd, M. A., 00. Normal vector votng: Crease detecton and curvature estmaton on large, nosy meshes. Journal of Graphcal Models 64(3/4), pp Pauly, M., Gross, M. and Kobbelt, L. P., 00. Effcent smplfcaton of pont-sampled surfaces. In: VIS 0: Proceedngs of the conference on Vsualzaton 0, IEEE Computer Socety, Boston, Massachusetts, pp Shaarj, C. M., 998. Least-squares fttng algorthms of the NIST algorthm testng system. Journal of Research of the Natonal Insttute of Standards and Technology 03(6), pp Tang, Q., Sang, N. and Zhang, T., 007. Extracton of salent contours from cluttered scenes. Pattern Recognton 40(), pp Tong,W.-S., Tang, C.-K., Mordoha, P. and Medon, G., 004. Frst order augmentaton to tensor votng for boundary nference and multscale analyss n 3D. IEEE Transactons on Pattern Analyss and Machne Intellgence 6(5), pp Wengarten, J., Gruener, G. and Segwart, R., 003. A fast and robust 3d feature extracton algorthm for structured envronment reconstructon. In: Proceedngs of th Internatonal Conference on Advanced Robotcs (ICAR), Portugal. 483

8 The Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences. Vol. XXXVII. Part B5. Bejng

AUTOMATING POST-PROCESSING OF TERRESTRIAL LASER SCANNING POINT CLOUDS FOR ROAD FEATURE SURVEYS

AUTOMATING POST-PROCESSING OF TERRESTRIAL LASER SCANNING POINT CLOUDS FOR ROAD FEATURE SURVEYS Internatonal Archves of Photogrammetry, Remote Sensng and Spatal Informaton Scences, Vol. XXXVIII, Part 5 Commsson V Symposum, Newcastle upon Tyne, UK. 2010 AUTOMATING POST-PROCESSING OF TERRESTRIAL LASER

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

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

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

More information

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

AUTOMATED REGISTRATION OF UNORGANISED POINT CLOUDS FROM TERRESTRIAL LASER SCANNERS

AUTOMATED REGISTRATION OF UNORGANISED POINT CLOUDS FROM TERRESTRIAL LASER SCANNERS AUTOMATED REGISTRATION OF UNORGANISED POINT CLOUDS FROM TERRESTRIAL LASER SCANNERS Kwang-Ho Bae and Derek D. Lcht Department of Spatal Scences, Curtn Unversty of Technology, GPO Box U1987, Perth, WA, 6845,

More information

Lecture 4: Principal components

Lecture 4: Principal components /3/6 Lecture 4: Prncpal components 3..6 Multvarate lnear regresson MLR s optmal for the estmaton data...but poor for handlng collnear data Covarance matrx s not nvertble (large condton number) Robustness

More information

Multi-stable Perception. Necker Cube

Multi-stable Perception. Necker Cube Mult-stable Percepton Necker Cube Spnnng dancer lluson, Nobuuk Kaahara Fttng and Algnment Computer Vson Szelsk 6.1 James Has Acknowledgment: Man sldes from Derek Hoem, Lana Lazebnk, and Grauman&Lebe 2008

More information

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

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

More information

Fitting & Matching. Lecture 4 Prof. Bregler. Slides from: S. Lazebnik, S. Seitz, M. Pollefeys, A. Effros.

Fitting & Matching. Lecture 4 Prof. Bregler. Slides from: S. Lazebnik, S. Seitz, M. Pollefeys, A. Effros. Fttng & Matchng Lecture 4 Prof. Bregler Sldes from: S. Lazebnk, S. Setz, M. Pollefeys, A. Effros. How do we buld panorama? We need to match (algn) mages Matchng wth Features Detect feature ponts n both

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

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

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

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

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

Some Advanced SPC Tools 1. Cumulative Sum Control (Cusum) Chart For the data shown in Table 9-1, the x chart can be generated.

Some Advanced SPC Tools 1. Cumulative Sum Control (Cusum) Chart For the data shown in Table 9-1, the x chart can be generated. Some Advanced SP Tools 1. umulatve Sum ontrol (usum) hart For the data shown n Table 9-1, the x chart can be generated. However, the shft taken place at sample #21 s not apparent. 92 For ths set samples,

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

Fuzzy Filtering Algorithms for Image Processing: Performance Evaluation of Various Approaches

Fuzzy Filtering Algorithms for Image Processing: Performance Evaluation of Various Approaches Proceedngs of the Internatonal Conference on Cognton and Recognton Fuzzy Flterng Algorthms for Image Processng: Performance Evaluaton of Varous Approaches Rajoo Pandey and Umesh Ghanekar Department of

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

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

Machine Learning: Algorithms and Applications

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

More information

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

Detection of an Object by using Principal Component Analysis

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

More information

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

Feature-Preserving Denoising of Point-Sampled Surfaces

Feature-Preserving Denoising of Point-Sampled Surfaces Feature-Preservng Denosng of Pont-Sampled Surfaces Jfang L College of Computer Scence and Informaton Technology Zhejang Wanl Unversty Nngbo 315100 Chna Abstract: Based on samplng lkelhood and feature ntensty,

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

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

Reducing Frame Rate for Object Tracking

Reducing Frame Rate for Object Tracking Reducng Frame Rate for Object Trackng Pavel Korshunov 1 and We Tsang Oo 2 1 Natonal Unversty of Sngapore, Sngapore 11977, pavelkor@comp.nus.edu.sg 2 Natonal Unversty of Sngapore, Sngapore 11977, oowt@comp.nus.edu.sg

More information

Wishing you all a Total Quality New Year!

Wishing you all a Total Quality New Year! Total Qualty Management and Sx Sgma Post Graduate Program 214-15 Sesson 4 Vnay Kumar Kalakband Assstant Professor Operatons & Systems Area 1 Wshng you all a Total Qualty New Year! Hope you acheve Sx sgma

More information

TN348: Openlab Module - Colocalization

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

More information

ALTERNATIVE METHODOLOGIES FOR THE ESTIMATION OF LOCAL POINT DENSITY INDEX: MOVING TOWARDS ADAPTIVE LIDAR DATA PROCESSING

ALTERNATIVE METHODOLOGIES FOR THE ESTIMATION OF LOCAL POINT DENSITY INDEX: MOVING TOWARDS ADAPTIVE LIDAR DATA PROCESSING Internatonal Archves of the Photogrammetry, Remote Sensng and Spatal Informaton Scences, Volume XXXIX-B3, 2012 XXII ISPRS Congress, 25 August 01 September 2012, Melbourne, Australa ALTERNATIVE METHODOLOGIES

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

1. Introduction. 2. Related Work

1. Introduction. 2. Related Work ASIAGRAPH 008 PROCEEDINGS Curvature Estmaton of 3D Pont Cloud Surfaces Through the Fttng of Normal Secton Curvatures Xaopeng Zhang* /LIAMA-NLPR, Insttute of Automaton, CAS; Hongjun L /LIAMA-NLPR, Insttute

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

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

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

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

Image Alignment CSC 767

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

More information

Robust Curvature Estimation and Geometry Analysis of 3D point Cloud Surfaces

Robust Curvature Estimation and Geometry Analysis of 3D point Cloud Surfaces Robust Curvature Estmaton and Geometry Analyss of 3D pont Cloud Surfaces Xaopeng ZHANG, Hongjun LI, Zhangln CHENG, Ykuan ZHANG Sno-French Laboratory LIAMA, Insttute of Automaton, CAS, Bejng 100190, Chna

More information

An Image Fusion Approach Based on Segmentation Region

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

More information

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

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

LEAST SQUARES. RANSAC. HOUGH TRANSFORM.

LEAST SQUARES. RANSAC. HOUGH TRANSFORM. LEAS SQUARES. RANSAC. HOUGH RANSFORM. he sldes are from several sources through James Has (Brown); Srnvasa Narasmhan (CMU); Slvo Savarese (U. of Mchgan); Bll Freeman and Antono orralba (MI), ncludng ther

More information

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices Internatonal Mathematcal Forum, Vol 7, 2012, no 52, 2549-2554 An Applcaton of the Dulmage-Mendelsohn Decomposton to Sparse Null Space Bases of Full Row Rank Matrces Mostafa Khorramzadeh Department of Mathematcal

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

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

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

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

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

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

More information

Development of an Active Shape Model. Using the Discrete Cosine Transform

Development of an Active Shape Model. Using the Discrete Cosine Transform Development of an Actve Shape Model Usng the Dscrete Cosne Transform Kotaro Yasuda A Thess n The Department of Electrcal and Computer Engneerng Presented n Partal Fulfllment of the Requrements for the

More information

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

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

More information

Brushlet Features for Texture Image Retrieval

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

More information

Unsupervised Learning and Clustering

Unsupervised Learning and Clustering Unsupervsed Learnng and Clusterng Why consder unlabeled samples?. Collectng and labelng large set of samples s costly Gettng recorded speech s free, labelng s tme consumng 2. Classfer could be desgned

More information

A NEW APPROACH FOR SUBWAY TUNNEL DEFORMATION MONITORING: HIGH-RESOLUTION TERRESTRIAL LASER SCANNING

A NEW APPROACH FOR SUBWAY TUNNEL DEFORMATION MONITORING: HIGH-RESOLUTION TERRESTRIAL LASER SCANNING A NEW APPROACH FOR SUBWAY TUNNEL DEFORMATION MONITORING: HIGH-RESOLUTION TERRESTRIAL LASER SCANNING L Jan a, Wan Youchuan a,, Gao Xanjun a a School of Remote Sensng and Informaton Engneerng, Wuhan Unversty,129

More information

User Authentication Based On Behavioral Mouse Dynamics Biometrics

User Authentication Based On Behavioral Mouse Dynamics Biometrics User Authentcaton Based On Behavoral Mouse Dynamcs Bometrcs Chee-Hyung Yoon Danel Donghyun Km Department of Computer Scence Department of Computer Scence Stanford Unversty Stanford Unversty Stanford, CA

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

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

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

CORRELATION ICP ALGORITHM FOR POSE ESTIMATION BASED ON LOCAL AND GLOBAL FEATURES

CORRELATION ICP ALGORITHM FOR POSE ESTIMATION BASED ON LOCAL AND GLOBAL FEATURES CORRELATION ICP ALGORITHM FOR POSE ESTIMATION BASED ON LOCAL AND GLOBAL FEATURES Marco A. Chavarra, Gerald Sommer Cogntve Systems Group. Chrstan-Albrechts-Unversty of Kel, D-2498 Kel, Germany {mc,gs}@ks.nformatk.un-kel.de

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

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

Helsinki University Of Technology, Systems Analysis Laboratory Mat Independent research projects in applied mathematics (3 cr)

Helsinki University Of Technology, Systems Analysis Laboratory Mat Independent research projects in applied mathematics (3 cr) Helsnk Unversty Of Technology, Systems Analyss Laboratory Mat-2.08 Independent research projects n appled mathematcs (3 cr) "! #$&% Antt Laukkanen 506 R ajlaukka@cc.hut.f 2 Introducton...3 2 Multattrbute

More information

Unsupervised Learning

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

More information

X- Chart Using ANOM Approach

X- Chart Using ANOM Approach ISSN 1684-8403 Journal of Statstcs Volume 17, 010, pp. 3-3 Abstract X- Chart Usng ANOM Approach Gullapall Chakravarth 1 and Chaluvad Venkateswara Rao Control lmts for ndvdual measurements (X) chart are

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

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

A New Approach For the Ranking of Fuzzy Sets With Different Heights

A New Approach For the Ranking of Fuzzy Sets With Different Heights New pproach For the ankng of Fuzzy Sets Wth Dfferent Heghts Pushpnder Sngh School of Mathematcs Computer pplcatons Thapar Unversty, Patala-7 00 Inda pushpndersnl@gmalcom STCT ankng of fuzzy sets plays

More information

A B-Snake Model Using Statistical and Geometric Information - Applications to Medical Images

A B-Snake Model Using Statistical and Geometric Information - Applications to Medical Images A B-Snake Model Usng Statstcal and Geometrc Informaton - Applcatons to Medcal Images Yue Wang, Eam Khwang Teoh and Dnggang Shen 2 School of Electrcal and Electronc Engneerng, Nanyang Technologcal Unversty

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

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

3D Face Recognition Fusing Spherical Depth Map and Spherical Texture Map

3D Face Recognition Fusing Spherical Depth Map and Spherical Texture Map Journal of Computer and Communcatons, 14, *, ** Publshed Onlne **** 14 n ScRes. http://www.scrp.org/journal/jcc http://dx.do.org/1.436/jcc.14.***** 3D Face Recognton Fusng Sphercal Depth Map and Sphercal

More information

The Research of Ellipse Parameter Fitting Algorithm of Ultrasonic Imaging Logging in the Casing Hole

The Research of Ellipse Parameter Fitting Algorithm of Ultrasonic Imaging Logging in the Casing Hole Appled Mathematcs, 04, 5, 37-3 Publshed Onlne May 04 n ScRes. http://www.scrp.org/journal/am http://dx.do.org/0.436/am.04.584 The Research of Ellpse Parameter Fttng Algorthm of Ultrasonc Imagng Loggng

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

Multi-View Face Alignment Using 3D Shape Model for View Estimation

Multi-View Face Alignment Using 3D Shape Model for View Estimation Mult-Vew Face Algnment Usng 3D Shape Model for Vew Estmaton Yanchao Su 1, Hazhou A 1, Shhong Lao 1 Computer Scence and Technology Department, Tsnghua Unversty Core Technology Center, Omron Corporaton ahz@mal.tsnghua.edu.cn

More information

A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION

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

More information

Multi-view 3D Position Estimation of Sports Players

Multi-view 3D Position Estimation of Sports Players Mult-vew 3D Poston Estmaton of Sports Players Robbe Vos and Wlle Brnk Appled Mathematcs Department of Mathematcal Scences Unversty of Stellenbosch, South Afrca Emal: vosrobbe@gmal.com Abstract The problem

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

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

Histogram-Enhanced Principal Component Analysis for Face Recognition

Histogram-Enhanced Principal Component Analysis for Face Recognition Hstogram-Enhanced Prncpal Component Analyss for Face ecognton Ana-ara Sevcenco and Wu-Sheng Lu Dept. of Electrcal and Computer Engneerng Unversty of Vctora sevcenco@engr.uvc.ca, wslu@ece.uvc.ca Abstract

More information

Positive Semi-definite Programming Localization in Wireless Sensor Networks

Positive Semi-definite Programming Localization in Wireless Sensor Networks Postve Sem-defnte Programmng Localzaton n Wreless Sensor etworks Shengdong Xe 1,, Jn Wang, Aqun Hu 1, Yunl Gu, Jang Xu, 1 School of Informaton Scence and Engneerng, Southeast Unversty, 10096, anjng Computer

More information

Structure from Motion

Structure from Motion Structure from Moton Structure from Moton For now, statc scene and movng camera Equvalentl, rgdl movng scene and statc camera Lmtng case of stereo wth man cameras Lmtng case of multvew camera calbraton

More information

Vanishing Hull. Jinhui Hu, Suya You, Ulrich Neumann University of Southern California {jinhuihu,suyay,

Vanishing Hull. Jinhui Hu, Suya You, Ulrich Neumann University of Southern California {jinhuihu,suyay, Vanshng Hull Jnhu Hu Suya You Ulrch Neumann Unversty of Southern Calforna {jnhuhusuyay uneumann}@graphcs.usc.edu Abstract Vanshng ponts are valuable n many vson tasks such as orentaton estmaton pose recovery

More information

The Shortest Path of Touring Lines given in the Plane

The Shortest Path of Touring Lines given in the Plane Send Orders for Reprnts to reprnts@benthamscence.ae 262 The Open Cybernetcs & Systemcs Journal, 2015, 9, 262-267 The Shortest Path of Tourng Lnes gven n the Plane Open Access Ljuan Wang 1,2, Dandan He

More information

Machine Learning. Topic 6: Clustering

Machine Learning. Topic 6: Clustering Machne Learnng Topc 6: lusterng lusterng Groupng data nto (hopefully useful) sets. Thngs on the left Thngs on the rght Applcatons of lusterng Hypothess Generaton lusters mght suggest natural groups. Hypothess

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

MOTION PANORAMA CONSTRUCTION FROM STREAMING VIDEO FOR POWER- CONSTRAINED MOBILE MULTIMEDIA ENVIRONMENTS XUNYU PAN

MOTION PANORAMA CONSTRUCTION FROM STREAMING VIDEO FOR POWER- CONSTRAINED MOBILE MULTIMEDIA ENVIRONMENTS XUNYU PAN MOTION PANORAMA CONSTRUCTION FROM STREAMING VIDEO FOR POWER- CONSTRAINED MOBILE MULTIMEDIA ENVIRONMENTS by XUNYU PAN (Under the Drecton of Suchendra M. Bhandarkar) ABSTRACT In modern tmes, more and more

More information

Fitting and Alignment

Fitting and Alignment Fttng and Algnment Computer Vson Ja-Bn Huang, Vrgna Tech Many sldes from S. Lazebnk and D. Hoem Admnstratve Stuffs HW 1 Competton: Edge Detecton Submsson lnk HW 2 wll be posted tonght Due Oct 09 (Mon)

More information

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation 17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 An Iteratve Soluton Approach to Process Plant Layout usng Mxed

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

IMPLEMENTATION OF 3D POINT CLOUDS SEGMENTATION BASED ON PLANE GROWING

IMPLEMENTATION OF 3D POINT CLOUDS SEGMENTATION BASED ON PLANE GROWING IMPLEMENTATION OF 3D POINT CLOUDS SEGMENTATION BASED ON PLANE GROWING METHOD Małgorzata Jarząbek-Rychard 1 Abstract In the recent years we can observe growng demand for the extracton of physcal objects

More information

MOTION BLUR ESTIMATION AT CORNERS

MOTION BLUR ESTIMATION AT CORNERS Gacomo Boracch and Vncenzo Caglot Dpartmento d Elettronca e Informazone, Poltecnco d Mlano, Va Ponzo, 34/5-20133 MILANO boracch@elet.polm.t, caglot@elet.polm.t Keywords: Abstract: Pont Spread Functon Parameter

More information

Comparison Study of Textural Descriptors for Training Neural Network Classifiers

Comparison Study of Textural Descriptors for Training Neural Network Classifiers Comparson Study of Textural Descrptors for Tranng Neural Network Classfers G.D. MAGOULAS (1) S.A. KARKANIS (1) D.A. KARRAS () and M.N. VRAHATIS (3) (1) Department of Informatcs Unversty of Athens GR-157.84

More information

A Comparison and Evaluation of Three Different Pose Estimation Algorithms In Detecting Low Texture Manufactured Objects

A Comparison and Evaluation of Three Different Pose Estimation Algorithms In Detecting Low Texture Manufactured Objects Clemson Unversty TgerPrnts All Theses Theses 12-2011 A Comparson and Evaluaton of Three Dfferent Pose Estmaton Algorthms In Detectng Low Texture Manufactured Objects Robert Krener Clemson Unversty, rkrene@clemson.edu

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

Help for Time-Resolved Analysis TRI2 version 2.4 P Barber,

Help for Time-Resolved Analysis TRI2 version 2.4 P Barber, Help for Tme-Resolved Analyss TRI2 verson 2.4 P Barber, 22.01.10 Introducton Tme-resolved Analyss (TRA) becomes avalable under the processng menu once you have loaded and selected an mage that contans

More information

Intra-Parametric Analysis of a Fuzzy MOLP

Intra-Parametric Analysis of a Fuzzy MOLP Intra-Parametrc Analyss of a Fuzzy MOLP a MIAO-LING WANG a Department of Industral Engneerng and Management a Mnghsn Insttute of Technology and Hsnchu Tawan, ROC b HSIAO-FAN WANG b Insttute of Industral

More information

BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET

BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET 1 BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET TZU-CHENG CHUANG School of Electrcal and Computer Engneerng, Purdue Unversty, West Lafayette, Indana 47907 SAUL B. GELFAND School

More information

SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE

SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE Dorna Purcaru Faculty of Automaton, Computers and Electroncs Unersty of Craoa 13 Al. I. Cuza Street, Craoa RO-1100 ROMANIA E-mal: dpurcaru@electroncs.uc.ro

More information

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task Proceedngs of NTCIR-6 Workshop Meetng, May 15-18, 2007, Tokyo, Japan Term Weghtng Classfcaton System Usng the Ch-square Statstc for the Classfcaton Subtask at NTCIR-6 Patent Retreval Task Kotaro Hashmoto

More information

Using Fuzzy Logic to Enhance the Large Size Remote Sensing Images

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

More information

Fast Computation of Shortest Path for Visiting Segments in the Plane

Fast Computation of Shortest Path for Visiting Segments in the Plane Send Orders for Reprnts to reprnts@benthamscence.ae 4 The Open Cybernetcs & Systemcs Journal, 04, 8, 4-9 Open Access Fast Computaton of Shortest Path for Vstng Segments n the Plane Ljuan Wang,, Bo Jang

More information