analternate approach in resolution enhancement for mrbrain image
S. Nithya,S. Kalyani
Published in International Journal of Advanced Research in Electronics, Communication & Instrumentation Engineering and Development
ISSN: 2347 -7210 Impact Factor:1.9 Volume:1 Issue:3 Year: 08 March,2014 Pages:125-124
Abstract
Inrecentyears,theoccurrenceofbraintumorhasbeenontherise.Alotof methodshavebeenproposedtoobtainmedicalimages(CTscan,differenttypesofX-rays,MR imagesandotherradiologicaltechniques)forfurtheranalysis.Majorproblemintheimages obtainedthroughtheabovesaidmethodsisthepresenceofblur,noise,artifacts,anddistortion. Evenasmallamountofnoisemayleadtofalsediagnosis.Hencethereisaprerequisiteforthe reductionsofnoisesforcorrectdiagnosis.ToreducethenoisepresentinMRimage,inthispaper differentfiltrationtechniquesareusedandtheirperformancesarecomparedbyevaluatingMSE andPSNR.Oncewhentheimageisfiltereditsqualitygetsdegradedandhencetoenhancethe qualityofanimage,anovelresolutionenhancementtechniquewhichgeneratesahighresolution imageisproposed.Inthiswork,DWTisappliedtodecomposealowresolutionimageinto differentsubbands.SimilarlySWTisalsoappliedtodecomposeanimageintodifferentsubbands. Then the three highfrequency sub band images ofDWT isinterpolated using bi-cubicinterpolation. Thehighfrequencysub-bandsobtainedbySWToftheinputimagearesummeduptothe interpolatedhighfrequencysubbandsinordertocorrecttheestimatedcoefficients.Inparallel,the inputimageisalsointerpolatedseparately.Finally,correctedinterpolatedhighfrequencysub bandsandinterpolatedinputimagearecombinedbyusingIDWTtoachieveahighresolutionMR image.Theperformanceofthistransformtechniqueisanalysedquantitativelywiththe conventionalDWTandSWTmethods.ThePSNRfortheproposedmethodisfoundtobe10dBto 20dBmorethantheconventionalmethods.Thustheresultsobtainedprovedthattheproposed technique givesa better qualityimage.
Kewords
MagneticResonance(MR),DiscreteWaveletTransform(DWT),Stationary WaveletTransform (SWT), Peak Signalto Noise Ratio(PSNR), Mean Square Error (MSE).
Reference
[1].AlexandraConstantin,RuzenaBajcsy,BerkeleySarah,“UnsupervisedSegmentationof BrainTissueinMultivariateMRI,”IEEEinternationalconferenceonbiomedicalimaging: from nano toMacro,pp.89-92, April 2010. [2].SubhranilKoley,AurpanMajumder,“BrainMRISegmentationforTumorDetection usingCohesionbasedSelfMergingAlgorithm”,CommunicationSoftwareandNetworks– ICCSN, pp. 781-785, May2011. [3].MonaMahmoudi,GuillermoSapiro,“FastImageandVideoDenoisingviaNon-Local Meansofsimilarneighbourhood”,IEEESignalProcessingLetters,vol.12,pp.839-842, 2005. [4].L.Sahawneh,B.Carroll,“StochasticImageDenoisingusingMinimumMeanSquared Error(Wiener)Filtering,” Electrical and Computer Engineering ,pp. 471–474, 2009. [6].Y.Rener,J.Wei,andC.Ken,“Downsample-basedmultipledescriptioncodingand postprocessingofdecoding,” inProc.27thChineseControlConf.,Jul.16–18,2008,pp. 253–256. [7].H.Demirel,G.Anbarjafari,andS.Izadpanahi,“Improvedmotion-basedlocalized superresolutiontechniqueusingdiscretewavelettransformforlowresolutionvideo enhancement,”inProc.17thEur.SignalProcess. Conf.,Glasgow,Scotland,Aug.2009,pp. 1097– 1101. [8].C.B.Atkins,C.A.Bouman,andJ.P.Allebach,“Optimalimagescalingusingpixel classification,”in Proc.Int. Conf.ImageProcess.,Oct. 7–10, 2001, vol. 3, pp. 864–867. [9].S. Mallat, A Wavelet Tourof Signal Processing, 2nd ed. New York: Academic, 1999. [10].Suresh Kumar, PapendraKumar, Manoj Gupta, AshokKumarNagawat, “Performance Comparison ofMedianandwienerfilterinimagedenosing”, InternationalJournalof Computer Applications, vol. 4, pp . 0975– 8887, November2010. [11].DipankarRay,DDuttaMajumder,AmitDas,“Noisereductionandimage enhancementofMRIusingadaptivemultiscaledatacondensation,”RecentAdvancesin Information Technology,pp. 107– 113, 2012. [12].LinaTan,LiangwuShi,“Multiwavelet-BasedEstimationforImprovingMagnetic ResonanceImages,” image and signal processing,CISP 09,pp.1– 5, 2009. [13].A.Temizel,“ImageresolutionenhancementusingwaveletdomainhiddenMarkov treeandcoefficientsignestimation,”inProc.Int.Conf.ImageProcess.,2007,vol.5,pp. V-381– 384.
