HYPERSPECTRAL IMAGE CLASSIFICATION USING A SELF-ORGANIZING MAP . (2)

Size: px
Start display at page:

Download "HYPERSPECTRAL IMAGE CLASSIFICATION USING A SELF-ORGANIZING MAP . (2)"

Transcription

1 HYPERSPECTRAL IMAGE CLASSIFICATION USING A SELF-ORGANIZING MAP P. Martínez, 1 J.A. Gualter, 2 P.L. Agular, 1 R.M. Pérez, 1 M. Lnaje, 1 J.C. Precado, 1 A. Plaza 1 1. INTRODUCTION The use of hyperspectral data to determne the abundance of consttuents n a certan porton of the Earth s surface reles on the capablty of magng spectrometers to provde a large amount of nformaton at each pxel of a certan scene. Today, hyperspectral magng sensors are capable of generatng unprecedented volumes of radometrc data. The Arborne Vsble/Infrared Imagng Spectrometer (AVIRIS), for example, routnely produces mage cubes wth 224 spectral bands (Green, ). Ths undoubtedly opens a wde range of new possbltes, but the analyss of such a massve amount of nformaton s not an easy task. In fact, most of the exstng algorthms devoted to analyzng multspectral mages are not applcable n the hyperspectral doman, because of the sze and hgh dmensonalty of the mages. Hyperspectral unmxng or lnear pxel unmxng s becomng ncreasngly popular for the analyss and nterpretaton of hyperspectral mages (Martínez et al., 1999). The basc assumpton s that the sgnal receved from each pxel can be consdered as a smple lnear combnaton of the spectral contrbutons of all pxel components. The technque, therefore, decomposes the scene n such a way as to recover the fractonal contrbutons of the fundamental components or endmembers (as abundance or fracton mages). Ths provdes a means of extractng sub-pxel nformaton from the scenes, whch s partcularly advantageous when the sze of the nterestng ground elements s much smaller than the mage resoluton and there s domnance of mxed pxels. Unsupervsed clusterng s a challengng problem n many areas of data analyss (Antonlle & Gualter, 2000). It can be stated as follows: gven a set of N data ponts n a feature space of D dmensons, D { x, x,..., x } R, x = (x, x,..., x ), 1,..., N, (1) 1 2 N 1 2 D = we wsh to characterze K clusters for the data, where K s obtaned from statstcal nformaton about the data usng some dstance metrc, d j = d(x, x j). (2) The prncpal characterstc of unsupervsed clusterng s that t does not ncorporate any prevous knowledge about the data. Snce ground truth data n remote sensng s expensve and hard to obtan, the use of unsupervsed procedures has become more relevant n ths feld durng recent years. In partcular, several unsupervsed procedures to process hyperspectral data are avalable n well-known commercal software systems as Research Systems ENVI TM. The applcaton of neural networks to perform unsupervsed classfcaton of hyperspectral data has been tested by several authors (Jménez et al., 1999) and also by us n some prevous work (Agular et al., 1998; Martínez et al., 1999; Agular et al., 2000a, Agular et al., 2000b). We have also focused on analyzng the ntrnsc capablty of neural networks to parallelze the whole hyperspectral unmxng process (Pérez et al., 1999). The results shown n ths work ndcate that neural network models are able to fnd clusters of closely related hyperspectral sgnatures, and thus can be used as a powerful tool to acheve the desred classfcaton. 1 Departamento de Informátca, Unversdad de Extremadura, Avda. de la Unversdad s/n, Cáceres, SPAIN E-mal: pablomar@unex.es 2 Global Scence and Technology at Appled Informaton Scences Branch, NASA/GSFC, Greenbelt, Maryland 20771, U.S.A. E-mal: gualt@peep.gsfc.nasa.gov

2 One of the most wdely used unsupervsed neural network algorthms s the Self-Organzng Map (SOM), proposed by Kohonen, Ths approach has been successfully appled n many dfferent felds, ncludng mage analyss and computer vson, handwrtten text recognton, analyss and recognton of human voce and telecommuncatons. Some reasons for usng SOM to perform hyperspectral analyss have been descrbed by Bruske and Mereny, 1999, who hghlght the computatonal speed provded by ths method when mplemented by hardware n the form of a massvely parallel algorthm, surpassng the performance of conventonal classfcaton algorthms. Snce the Kohonen algorthm s smple and ntutve, hghly parallelzable (whch can lead to an easy VLSI mplementaton based on systolc arrays or FPGAs) and s easly extendable to a hgh number of dmensons, we have selected t as our startng pont to deal wth hyperspectral data. The present work dscusses the possblty of usng a Self Organzng neural network to perform unsupervsed classfcaton of hyperspectral mages. In sectons 3 and 4, the topology of the proposed neural network and the tranng algorthm are respectvely descrbed. Secton 5 provdes the results we have obtaned after applyng the proposed methodology to real hyperspectral data, descrbed n secton 2. Dfferent parameters n the learnng stage have been modfed n order to obtan a detaled descrpton of ther nfluence on the fnal results. Fnally, n secton 6 we provde the conclusons at whch we have arrved. 2. DATA The hyperspectral unmxng algorthms proposed n ths work have been tested usng the publc doman Indan Pnes hyperspectral dataset, whch has been prevously used n many dfferent studes. Ths mage was obtaned from the AVIRIS magng spectrometer at Northern Indana on June 12, 1992 from a NASA ER2 flght at hgh alttude wth ground pxel resoluton of 17 meters. The dataset comprses 145x145 pxels and 220 bands of sensor radance wthout atmospherc correcton. It contans two thrds of agrculture (some of the crops are n early stages of growth wth low coverage), and one thrd of forest, two hghways, a ral lane and some houses. Ground truth determnes sxteen dfferent classes (not mutually exclusve). Water absorpton bands ( , and 220) were removed (Tadjudn and Landgrebe, 1998), obtanng a 200 band spectrum at each pxel. In order to reduce the tme of tranng and testng, we have selected a subscene of the complete Indan Pnes dataset (depcted n Fgure 1) of sze 68 samples x 86 lnes at [27-94] x [31-116] n the orgnal mage, consderng left n the full scene s at (1,1). In the selected subscene there are four known ground truth classes. Corn-notll Soybean 1 Grass Soybean 2 Fgure 1. A subset of the Indan Pnes hyperspectral dataset wth ground truth. 3. TOPOLOGY OF THE PROPOSED NEURAL NETWORK The Self-Organzng Map (SOM) s based on compettve learnng that leads to the constructon of topologc maps representng class prototypes. In order to understand the topology of the proposed neural network, we frst need to defne some basc concepts. A neuron s an nformaton-processng unt. Neurons are connected

3 by synapses or connectng lnks, each of them characterzed by a weght. Specfcally, a sgnal x j at the nput of synapse j connected to neuron k s multpled by the synaptc weght w kj. A neural network s a set of neurons organzed n the form of layers. In the smplest form, an nput layer projects onto an output layer of neurons. If the nput layer has N unts and the output layer has M unts, each unt n the output layer owns N weghts assocated to the connectons whch come from the nput layer, so that the set of neural weghts s organzed n the form of a two-dmensonal lattce (W MxN ). Our proposed network archtecture s depcted n Fgure 2 (Agular et al., 2000b). In our case, N corresponds to the number of channels of the hyperspectral mage and M s the number of classes or prototypes to be extracted by the network. M must be carefully selected accordng to some metrc (we wll nsst on ths ssue later on n the paper). There are feedforward connectons from the nput layer to the output layer and selffeedback and lateral feedback connectons n the output layer. These two types of local connectons serve two dfferent purposes: a) In the classfcaton stage, the weghted sum x W (scalar product ) of the nput sgnals x at each neuron performs feature detecton: each neuron produces a selectve response to nput sgnals. b) In the learnng stage, lateral and self-feedback connectons produce exctatory or nhbtory effects dependng on the dstance from the correspondng output layer neuron to the wnnng neuron (Agular et al., 2000c). Ther assocated weghts are used to determne the W classfcaton prototype for each neuron. 1 SOM Characterstcs R M y N W Fgure 2. SOM neural network topology ncludng weght matrx W and self-feedback and lateral feedback connectons. S y c sy 4. TRAINING ALGORITHM There are fve basc steps nvolved n the tranng algorthm. These steps are repeated untl the topologcal map s completely formed: a) Intalzaton: choose random values for the ntal weght vectors, w (0), = 1,2,... M. It s desrable to keep the magntude of the weghts small. b) Samplng: choose an nput pattern x(n) belongng to a set of learnng patterns or references, R. The selecton s done randomly. c) Smlarty Matchng: fnd the best-matchng (wnnng) neuron * at tme t, usng the mnmum-dstance crteron as shown n the followng equaton, where dst s the eucldean dstance: * [ x(n) ] mn dst{ x(n), w (t)}, j = 1,2,..., M = (3) j j

4 d) Learnng: adjust the synaptc weght vectors of all neurons, usng the update formula (4), where η(t) s a ( ) * learnng-rate parameter, and t,, [ x(n) ] γ s a Gaussan neghborhood functon centered around the wnnng neuron. The sze of the neghborhood s determned by the reference dstance σ (t) (see equaton 5). * ( t,, [ x(n) ])( x(n) w (t)) w (t + 1) = w (t) + η(t) γ (4) From the dfferent optons to select the prevously mentoned parameters, takng nto account the studes done n Agular, 2000, we have selected the followng ones: 1 ( t) = t ( ) e γ =, σ(t) * η, t,, [ x(n) ] dst( * ) 2, σ 0 σ (t) =. (5) t e) Contnue from step b) untl no notceable changes n the weght space are observed, or untl the maxmum convergence tme s acheved. In order to analyze a hyperspectral mage usng ths algorthm, the network must be traned wth hyperspectral sgnatures obtaned drectly from the mage. The weghts ntally assocated wth each output layer neuron contan the hyperspectral sgnatures of some carefully selected pxels on the mage (accordng wth ther spatal dstrbuton). 5. RESULTS AND DISCUSSION We have appled our proposed neural network to real hyperspectral data, descrbed n secton 2. Snce there are several parameters nvolved n the tranng algorthm (descrbed n the prevous secton) n ths secton we analyze the nfluence of those parameters n the process of class prototype extracton. In partcular, the parameters that we consder n the present study are the number of teratons untl convergence of the neural network s reached, the sze of neghborhood functon γ centered around the wnnng neuron, whch s determned by σ (t), and the number of neurons n the output layer of the neural network. The experment s performed as follows. We tran the network wth all the hyperspectral sgnatures of the mage. Durng the learnng stage, we go through all the pxels of the mage startng from a random pxel whch s dfferent n each of the teratons. Once class prototypes have been extracted, each pxel s classfed and the confuson matrx (Chuveco, 2000) s obtaned. Ths matrx allows us to vsualze wnnng neuron densty for each class. The characterstcs of the confuson matrx provdes us wth a comprehensve vsualzaton of dstrbuton n N-dmensonal space, and may ndcate the accuracy of the classfcaton. Snce each column corresponds to an output neuron, f one column presents hgh values for dfferent classes, the overall accuracy of the classfcaton should be low. In order to measure the degree of accuracy of the classfcaton, we propose the followng metrc based on the topology of the confuson matrx: Xm E =, Xm Max(X j) Xs =, Xs = X j. (6) Xm s the maxmum value for a column of the confuson matrx and Xs s the sum of all the values n that column. E provdes nformaton about the capacty of each neuron to dscrmnate between the classes, and can be averaged for all the neurons n the network, provdng a general measure about the accuracy of the classfcaton. The performance of the SOM Neural Network depends on a lot of adjustng parameters: 1. Number of output neurons: the deal numbers of output neurons must be equal to the number of ground truth classes, assocatng exactly one neuron wth one class. Usually ths fact s not possble f we have a larger number of output neurons that ground truth classes. In ths way, one correct classfcaton j

5 uses several neurons for each ground truth class. In the Indan Pnes ground-truth mage there are 16 classes (plus one class of unclassfed pxels). 2. Wegth ntalzaton: the weghts w j ntally assocated wth each neuron contan random values. 3. Order to scan the mage: a random ntal pont of the mage s selected at each teraton. 4. Neghborhood functon: as mentoned before, our choce for the neghborhood functon s the Gaussan functon. 5. Reference dstance σ(t): when the algorthm starts (t=1), the neural lattce s n a random state, the neghborhood functon n these frst teratons of the algorthm must have smlar values for to nclude a large number of neurons and obtan some average values (hgh σ(0) values). When t ncreases, γ(t,, * [x(n)] ) needs to be adjusted to reduce the number of neghbor neurons. Care must be taken nto account to avod a quckly reducton of the number of neghborhood neurons, ths reducton can be accomplshed by changng the reference dstance σ (t). Our choce for the reference dstance evoluton s σ (t)=(1/t) 2 The parameters that we analyze n the present study are the number of neurons n the output layer, the number of teratons of the neural network and the neghborhood startng value σ 0. Next, some results obtaned for the hyperspectral data descrbed n secton 2 are provded. In our frst experment, we have consdered 16 neurons n the output layer, 100 teratons and σ 0 = 2. Table 1 shows the resultng confuson matrx and Fgure 3 shows the the resultng classfcaton provded by the neural network usng the prevously mentoned parameters along wth a greyscale representaton of the confuson matrx shown n Table 1. A favorable result would be obtaned f neurons actvate exclusvely for a partcular class, dscrmnatng ths class from the others. In the confuson matrx, ths can be graphcally expressed as a row for whch several columns present hgh values. In Fgure 3 we can apprecate ths stuaton at four dfferent rows (2,6,10,11 and 17). The fact that column values overlap ndcates an naccurate classfcaton. Another ndcator of the qualty of the classfcaton s the contnuty of hgh values n the rows of the confuson matrx (ths fact should produce brght contguous rows n the confuson matrx mage). In ths experment, the topology of the resultng classes s not preserved snce we can apprecate several dscontnutes n the learnt classes. The overall accuracy of the classfcaton obtaned n ths experment was 60% accordng to the measure provded n equaton 6. Nevertheless, we have to take nto account that class 17 n the confuson matrx (see Table 1) corresponds to pxels whch have not been classfed durng the process, and we are consderng these pxels when calculatng the overall accuracy (f we do not consder these pxels, accuracy ncreases to 80%). Classes Neurons n the output layer CORN-NOTILL GRASS SOYBEAN SOYBEAN UNCLASS Table 1. Resultng confuson matrx for the Indan Pnes dataset consderng 100 teratons, 16 neurons n the output layer and σ 0 = 2. Non-zero cells n the matrx are hghlghted.

6 In our second experment, we consder a smaller number of neurons n the output layer (5), 80 teratons and σ = 4 0. Fgure 4 shows the resultng classfcaton and the assocated confuson matrx. We can apprecate an mprovement n the topology of the confuson matrx (horzontal lnes are more contguous). In ths case, the overall accuracy s 60%, wthout consderng the unclassfed pxels. Fgure 3. Resultng classfcaton and greyscale representaton of the confuson matrx for the Indan Pnes dataset consderng 100 teratons, 16 neurons n the output layer and σ 0 = 2. The tme used for ths computaton was approxmately 60 mnutes n an AMD K MHz Processor wth 128 Mb of SDRAM memory and IDL 5.4. Fgure 4. Resultng classfcaton and assocated confuson matrx for the Indan Pnes dataset consderng 80 teratons, 5 neurons n the output layer and σ = 0 4. The tme used for ths computaton was 20 mnutes n an AMD K MHz Processor wth 128 Mb of SDRAM memory and IDL 5.4. Fnally, n our last experment we ncrease the number of neurons n the output layer (16), we decrease the number of teratons (20) and consder σ = The results are addressed n Fgure 5. A general mprovement n the topology s acheved n ths experment, and the overall accuracy n ths case s 72%, due to the fact that we compensate the ncrease n the number of neurons wth a subsequent ncrease n the number of neurons that are consdered n the compettve step ( σ = 0 12 ). Table 2 shows other experments we have performed over the Indan Pnes hyperspectral dataset. As we can apprecate n Fgure 5, the ncrease n the number of neurons produces some dscontnutes n the topology, but the overall performance ncreases due to the reducton n the overlappng percentage between brght zones (hgh values) n the confuson matrx. These facts can be reduced ncreasng σ 0. From the prevously addressed results, we can conclude: 1. The number of teratons needed to obtan an acceptable accuracy s low compared to other SOM applcatons. Ths fact s probably related to the hgh dmensonalty of hyperspectral sgnatures.

7 2. It s more effcent, n terms of accuracy, to ncrease the number of neurons and decrease the number of teratons. As we ncrease the number of neurons, the overall accuracy ncreases, but the topology of the classes s poor. 3. When the SOM neural network has a large number of neurons n the output layer, the ntal reference dstance must be ncreased n order to mantan smlar performance values n the compettve process. In ths sense, σ(t) plays an mportant role. 4. Further work s stll needed n order to acheve a reasonable compromse between topology preservaton, σ(t) functon and overall accuracy. Fgure 5. Resultng classfcaton and greyscale representaton of the confuson matrx for the Indan Pnes dataset consderng 20 teratons, 16 neurons n the output layer and σ = The tme used for ths computaton was 7 mnutes n an AMD K MHhz Processor wth 128 Mb of SDRAM memory and IDL 5.4. Number of neurons Number of teratons Sze of neghborhood ( σ 0 ) Accuracy (%) Table 2. Other experments performed over the Indan Pnes dataset.

8 6. CONCLUSIONS We have presented a new approach to unsupervsed classfcaton of hyperspectral mages usng a Self Organzng Map. The overall performance of the method has been tested by ts applcaton to real hyperspectral data. The avalablty of ground truth allows us to ntroduce a new statstcal measure to quantfy the accuracy of the resultng classfcaton. Snce the tranng stage of the neural network ncorporates several parameters, we have studed the nfluence of some of these parameters on the fnal result. ACKNOWLEDGEMENTS Fundngs from Junta de Extremadura (PRI Program, IDUAP Grant) and European Communty (LFR Program, TEITORS Grant) are also gratefully acknowledged. REFERENCES Agular, P.L, Martínez, P., Pérez R.M., Hormgo, A., Abundance Extractons from AVIRIS Images Usng a Self Organzng Neural Network, Summares of the IX JPL Arborne Earth Scence Workshop, pp , JPL/NASA, 2000a. Agular, P.L., Cuantfcacón de Frmas Hperespectrales Usando Mapas Autoorganzatvos, PhD Thess, Escuela Poltecnca de Cáceres, Unversdad de Extremadura, 2000c. Agular, P.L., Pérez, R.M., Martínez, P., Bachller, P., Merchán, A., Spectra Evaluaton and Recognton n the Mxture Problem Usng SOFM Algorthm, Proc. Internatonal Symposum on Engneerng of Intellgent Systems, (EIS 98), 1998, Vol. 2, pp Agular, P.L., Plaza, A., Martínez, P., Pérez, R.M., Endmember Extracton by a Self-Organzng Neural Network on Hyperspectral Images, Proc. Internatonal Conference on Automaton, Robotcs and Computer Vson, Nanyang Technologcal Insttute, Sngapore, 2000b. Antonlle, S. and Gualter, J.A., Vsualzng Clusters n Hgh-Dmensonal Data wth a Kohonen Self Organzng Map, Techncal Report, Bruske, J. and Merény, E., 1999, Estmatng the Intrnsc Dmensonalty of Hyperspectral Images, Proc. European Symposum on Artfcal Neural Network, ESANN 99, Brussels, Belgum, Aprl, 1999, pp Chuveco, E., Fundamentos de Teledeteccón Espacal, Edcones Ralp, Span, Green, R.O., Edtor, AVIRIS Earth Scence Workshop Proceedngs, Avalable at Jmenez, L.O., Morales-Morell, A., Creus, A., Classfcaton of Hyperdmensonal Data Based on Feature and Decson Fuson Approaches Usng Projecton Pursut, Majorty Votng, and Neural Networks, IEEE Trans. Geoscence and Remote Sensng, Vol. 37, Issue 3, Part 1, May 1999, pp Kohonen, T., Self-Organzng Maps (2 nd ed.), Sprnger Seres n Informaton Scence Martínez, P., Pérez, R.M., Agular, P.L., Bachller, P. and Daz, P., A Neuronal Tool for AVIRIS Hyperspectral Unmxng, Summares of the IX JPL Arborne Earth Scence Workshop, pp , JPL/NASA, Pérez, R.M., Agular, P.L., Bachller, P. and Martínez, P., Neural Network Quantfer for Solvng the Mxture Problem and ts Implementaton by Systolc Array, Mcroelectroncs Journal, 30 (1), pp , Taudjn, S. and Landgrebe, D., Classfcaton of Hgh Dmensonal Data wth Lmted Tranng Samples, Doctoral Thess, School of Electrcal Engneerng and Computer Scence, Purdue Unversty, 1998.

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

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

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

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

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

Classifying Acoustic Transient Signals Using Artificial Intelligence

Classifying Acoustic Transient Signals Using Artificial Intelligence Classfyng Acoustc Transent Sgnals Usng Artfcal Intellgence Steve Sutton, Unversty of North Carolna At Wlmngton (suttons@charter.net) Greg Huff, Unversty of North Carolna At Wlmngton (jgh7476@uncwl.edu)

More information

Lecture 13: High-dimensional Images

Lecture 13: High-dimensional Images Lec : Hgh-dmensonal Images Grayscale Images Lecture : Hgh-dmensonal Images Math 90 Prof. Todd Wttman The Ctadel A grayscale mage s an nteger-valued D matrx. An 8-bt mage takes on values between 0 and 55.

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

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

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

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

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

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

KOHONEN'S SELF ORGANIZING NETWORKS WITH "CONSCIENCE"

KOHONEN'S SELF ORGANIZING NETWORKS WITH CONSCIENCE Kohonen's Self Organzng Maps and ther use n Interpretaton, Dr. M. Turhan (Tury) Taner, Rock Sold Images Page: 1 KOHONEN'S SELF ORGANIZING NETWORKS WITH "CONSCIENCE" By: Dr. M. Turhan (Tury) Taner, Rock

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

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

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

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

The Research of Support Vector Machine in Agricultural Data Classification

The Research of Support Vector Machine in Agricultural Data Classification The Research of Support Vector Machne n Agrcultural Data Classfcaton Le Sh, Qguo Duan, Xnmng Ma, Me Weng College of Informaton and Management Scence, HeNan Agrcultural Unversty, Zhengzhou 45000 Chna Zhengzhou

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

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

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

Hyperspectral Image Classification Based on Local Binary Patterns and PCANet

Hyperspectral Image Classification Based on Local Binary Patterns and PCANet Hyperspectral Image Classfcaton Based on Local Bnary Patterns and PCANet Huzhen Yang a, Feng Gao a, Junyu Dong a, Yang Yang b a Ocean Unversty of Chna, Department of Computer Scence and Technology b Ocean

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

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 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

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

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

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

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

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

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

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching

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

More information

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 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

Simulation Based Analysis of FAST TCP using OMNET++

Simulation Based Analysis of FAST TCP using OMNET++ Smulaton Based Analyss of FAST TCP usng OMNET++ Umar ul Hassan 04030038@lums.edu.pk Md Term Report CS678 Topcs n Internet Research Sprng, 2006 Introducton Internet traffc s doublng roughly every 3 months

More information

Face Recognition Based on SVM and 2DPCA

Face Recognition Based on SVM and 2DPCA Vol. 4, o. 3, September, 2011 Face Recognton Based on SVM and 2DPCA Tha Hoang Le, Len Bu Faculty of Informaton Technology, HCMC Unversty of Scence Faculty of Informaton Scences and Engneerng, Unversty

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

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

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

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

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

A fast algorithm for color image segmentation

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

More information

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

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

IMAGE FUSION TECHNIQUES

IMAGE FUSION TECHNIQUES Int. J. Chem. Sc.: 14(S3), 2016, 812-816 ISSN 0972-768X www.sadgurupublcatons.com IMAGE FUSION TECHNIQUES A Short Note P. SUBRAMANIAN *, M. SOWNDARIYA, S. SWATHI and SAINTA MONICA ECE Department, Aarupada

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

Problem Definitions and Evaluation Criteria for Computational Expensive Optimization

Problem Definitions and Evaluation Criteria for Computational Expensive Optimization Problem efntons and Evaluaton Crtera for Computatonal Expensve Optmzaton B. Lu 1, Q. Chen and Q. Zhang 3, J. J. Lang 4, P. N. Suganthan, B. Y. Qu 6 1 epartment of Computng, Glyndwr Unversty, UK Faclty

More information

BAYESIAN MULTI-SOURCE DOMAIN ADAPTATION

BAYESIAN MULTI-SOURCE DOMAIN ADAPTATION BAYESIAN MULTI-SOURCE DOMAIN ADAPTATION SHI-LIANG SUN, HONG-LEI SHI Department of Computer Scence and Technology, East Chna Normal Unversty 500 Dongchuan Road, Shangha 200241, P. R. Chna E-MAIL: slsun@cs.ecnu.edu.cn,

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

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

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps

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

More information

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

An Improved Neural Network Algorithm for Classifying the Transmission Line Faults

An Improved Neural Network Algorithm for Classifying the Transmission Line Faults 1 An Improved Neural Network Algorthm for Classfyng the Transmsson Lne Faults S. Vaslc, Student Member, IEEE, M. Kezunovc, Fellow, IEEE Abstract--Ths study ntroduces a new concept of artfcal ntellgence

More information

Object-Based Techniques for Image Retrieval

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

More information

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

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

Optimal connection strategies in one- and two-dimensional associative memory models

Optimal connection strategies in one- and two-dimensional associative memory models Optmal connecton strateges n one- and two-dmensonal assocatve memory models Lee Calcraft, Rod Adams, and Nel Davey School of Computer Scence, Unversty of Hertfordshre College lane, Hatfeld, Hertfordshre

More information

PERFORMANCE EVALUATION FOR SCENE MATCHING ALGORITHMS BY SVM

PERFORMANCE EVALUATION FOR SCENE MATCHING ALGORITHMS BY SVM PERFORMACE EVALUAIO FOR SCEE MACHIG ALGORIHMS BY SVM Zhaohu Yang a, b, *, Yngyng Chen a, Shaomng Zhang a a he Research Center of Remote Sensng and Geomatc, ongj Unversty, Shangha 200092, Chna - yzhac@63.com

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

International Conference on Applied Science and Engineering Innovation (ASEI 2015)

International Conference on Applied Science and Engineering Innovation (ASEI 2015) Internatonal Conference on Appled Scence and Engneerng Innovaton (ASEI 205) Desgn and Implementaton of Novel Agrcultural Remote Sensng Image Classfcaton Framework through Deep Neural Network and Mult-

More information

REMOTE SENSING REQUIREMENTS DEVELOPMENT: A SIMULATION-BASED APPROACH

REMOTE SENSING REQUIREMENTS DEVELOPMENT: A SIMULATION-BASED APPROACH REMOTE SENSING REQUIREMENTS DEVEOPMENT: A SIMUATION-BASED APPROAC V. Zanon a, B. Davs a, R. Ryan b, G. Gasser c, S. Blonsk b a Earth Scence Applcatons Drectorate, Natonal Aeronautcs and Space Admnstraton,

More information

Collaboratively Regularized Nearest Points for Set Based Recognition

Collaboratively Regularized Nearest Points for Set Based Recognition Academc Center for Computng and Meda Studes, Kyoto Unversty Collaboratvely Regularzed Nearest Ponts for Set Based Recognton Yang Wu, Mchhko Mnoh, Masayuk Mukunok Kyoto Unversty 9/1/013 BMVC 013 @ Brstol,

More information

Local Adaptive Receptive Field Self-organizing Map for Image Segmentation

Local Adaptive Receptive Field Self-organizing Map for Image Segmentation Local Adaptve Receptve Feld Self-organzng Map for Image Segmentaton Aluzo R. F. Araúo, Dogo C. Costa Unversdade Federal de Pernambuco Centro de Informátca - CIn Departamento de Sstemas da Computação Av.

More information

Sum of Linear and Fractional Multiobjective Programming Problem under Fuzzy Rules Constraints

Sum of Linear and Fractional Multiobjective Programming Problem under Fuzzy Rules Constraints Australan Journal of Basc and Appled Scences, 2(4): 1204-1208, 2008 ISSN 1991-8178 Sum of Lnear and Fractonal Multobjectve Programmng Problem under Fuzzy Rules Constrants 1 2 Sanjay Jan and Kalash Lachhwan

More information

Contextual Classification of Hyperspectral Remote Sensing Images Using SVM-PLR

Contextual Classification of Hyperspectral Remote Sensing Images Using SVM-PLR Australan Journal of Basc and Appled Scences, 5(8): 374-382, 2011 ISSN 1991-8178 Contextual Classfcaton of Hyperspectral Remote Sensng Images Usng SVM-PLR 1 Mahd hodadadzadeh, 2 Hassan Ghasseman 1,2 Faculty

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

FINDING IMPORTANT NODES IN SOCIAL NETWORKS BASED ON MODIFIED PAGERANK

FINDING IMPORTANT NODES IN SOCIAL NETWORKS BASED ON MODIFIED PAGERANK FINDING IMPORTANT NODES IN SOCIAL NETWORKS BASED ON MODIFIED PAGERANK L-qng Qu, Yong-quan Lang 2, Jng-Chen 3, 2 College of Informaton Scence and Technology, Shandong Unversty of Scence and Technology,

More information

Application of Learning Machine Methods to 3 D Object Modeling

Application of Learning Machine Methods to 3 D Object Modeling Applcaton of Learnng Machne Methods to 3 D Object Modelng Crstna Garca, and José Al Moreno Laboratoro de Computacón Emergente, Facultades de Cencas e Ingenería, Unversdad Central de Venezuela. Caracas,

More information

USING LINEAR REGRESSION FOR THE AUTOMATION OF SUPERVISED CLASSIFICATION IN MULTITEMPORAL IMAGES

USING LINEAR REGRESSION FOR THE AUTOMATION OF SUPERVISED CLASSIFICATION IN MULTITEMPORAL IMAGES USING LINEAR REGRESSION FOR THE AUTOMATION OF SUPERVISED CLASSIFICATION IN MULTITEMPORAL IMAGES 1 Fetosa, R.Q., 2 Merelles, M.S.P., 3 Blos, P. A. 1,3 Dept. of Electrcal Engneerng ; Catholc Unversty of

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

Novel Pattern-based Fingerprint Recognition Technique Using 2D Wavelet Decomposition

Novel Pattern-based Fingerprint Recognition Technique Using 2D Wavelet Decomposition Mathematcal Methods for Informaton Scence and Economcs Novel Pattern-based Fngerprnt Recognton Technque Usng D Wavelet Decomposton TUDOR BARBU Insttute of Computer Scence of the Romanan Academy T. Codrescu,,

More information

Feature Selection for Target Detection in SAR Images

Feature Selection for Target Detection in SAR Images Feature Selecton for Detecton n SAR Images Br Bhanu, Yngqang Ln and Shqn Wang Center for Research n Intellgent Systems Unversty of Calforna, Rversde, CA 95, USA Abstract A genetc algorthm (GA) approach

More information

A Clustering Algorithm for Key Frame Extraction Based on Density Peak

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

More information

Recognition of Handwritten Numerals Using a Combined Classifier with Hybrid Features

Recognition of Handwritten Numerals Using a Combined Classifier with Hybrid Features Recognton of Handwrtten Numerals Usng a Combned Classfer wth Hybrd Features Kyoung Mn Km 1,4, Joong Jo Park 2, Young G Song 3, In Cheol Km 1, and Chng Y. Suen 1 1 Centre for Pattern Recognton and Machne

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

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

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

More information

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

Classification of Face Images Based on Gender using Dimensionality Reduction Techniques and SVM

Classification of Face Images Based on Gender using Dimensionality Reduction Techniques and SVM Classfcaton of Face Images Based on Gender usng Dmensonalty Reducton Technques and SVM Fahm Mannan 260 266 294 School of Computer Scence McGll Unversty Abstract Ths report presents gender classfcaton based

More information

A Background Subtraction for a Vision-based User Interface *

A Background Subtraction for a Vision-based User Interface * A Background Subtracton for a Vson-based User Interface * Dongpyo Hong and Woontack Woo KJIST U-VR Lab. {dhon wwoo}@kjst.ac.kr Abstract In ths paper, we propose a robust and effcent background subtracton

More information

An Improved Image Segmentation Algorithm Based on the Otsu Method

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

More information

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

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

Constructing Minimum Connected Dominating Set: Algorithmic approach

Constructing Minimum Connected Dominating Set: Algorithmic approach Constructng Mnmum Connected Domnatng Set: Algorthmc approach G.N. Puroht and Usha Sharma Centre for Mathematcal Scences, Banasthal Unversty, Rajasthan 304022 usha.sharma94@yahoo.com Abstract: Connected

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

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

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

Detecting Compounded Anomalous SNMP Situations Using Cooperative Unsupervised Pattern Recognition

Detecting Compounded Anomalous SNMP Situations Using Cooperative Unsupervised Pattern Recognition Detectng Compounded Anomalous SNMP Stuatons Usng Cooperatve Unsupervsed Pattern Recognton Emlo Corchado, Álvaro Herrero, José Manuel Sáz Department of Cvl Engneerng, Unversty of Burgos, Span escorchado@ubu.es

More information

A Statistical Model Selection Strategy Applied to Neural Networks

A Statistical Model Selection Strategy Applied to Neural Networks A Statstcal Model Selecton Strategy Appled to Neural Networks Joaquín Pzarro Elsa Guerrero Pedro L. Galndo joaqun.pzarro@uca.es elsa.guerrero@uca.es pedro.galndo@uca.es Dpto Lenguajes y Sstemas Informátcos

More information

An Evaluation of Divide-and-Combine Strategies for Image Categorization by Multi-Class Support Vector Machines

An Evaluation of Divide-and-Combine Strategies for Image Categorization by Multi-Class Support Vector Machines An Evaluaton of Dvde-and-Combne Strateges for Image Categorzaton by Mult-Class Support Vector Machnes C. Demrkesen¹ and H. Cherf¹, ² 1: Insttue of Scence and Engneerng 2: Faculté des Scences Mrande Galatasaray

More information

Backpropagation: In Search of Performance Parameters

Backpropagation: In Search of Performance Parameters Bacpropagaton: In Search of Performance Parameters ANIL KUMAR ENUMULAPALLY, LINGGUO BU, and KHOSROW KAIKHAH, Ph.D. Computer Scence Department Texas State Unversty-San Marcos San Marcos, TX-78666 USA ae049@txstate.edu,

More information

Analysis of Continuous Beams in General

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

More information

Mining Image Features in an Automatic Two- Dimensional Shape Recognition System

Mining Image Features in an Automatic Two- Dimensional Shape Recognition System Internatonal Journal of Appled Mathematcs and Computer Scences Volume 2 Number 1 Mnng Image Features n an Automatc Two- Dmensonal Shape Recognton System R. A. Salam, M.A. Rodrgues Abstract The number of

More information

Editorial Manager(tm) for International Journal of Pattern Recognition and

Editorial Manager(tm) for International Journal of Pattern Recognition and Artfcal Intellgence Edtoral Manager(tm) for Internatonal Journal of Pattern Recognton and Manuscrpt Draft Manuscrpt Number: Ttle: TEXT LOCALIZATION IN COMPLEX COLOR DOCUMENTS Artcle Type: Research Paper

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

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

PYTHON IMPLEMENTATION OF VISUAL SECRET SHARING SCHEMES

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

More information

Fast Feature Value Searching for Face Detection

Fast Feature Value Searching for Face Detection Vol., No. 2 Computer and Informaton Scence Fast Feature Value Searchng for Face Detecton Yunyang Yan Department of Computer Engneerng Huayn Insttute of Technology Hua an 22300, Chna E-mal: areyyyke@63.com

More information

Local Quaternary Patterns and Feature Local Quaternary Patterns

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

More information