ONLINE KANNADA HANDWRITTEN WORDS RECOGNITION USING DEEP LEARNING

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

ONLINE KANNADA HANDWRITTEN WORDS RECOGNITION USING DEEP LEARNING

Sangeetha G R1, Dr. Lakshman Naika R2

1P G Scholar, Department of Electronics and Communication Engineering, University BDT College of Engineering, Davanagere, Karnataka, India

2Professor, Department of Electronics and Communication Engineering, University BDT College of Engineering, Davanagere, Karnataka, India ***

Abstract The main objective of this paper is to recognize online Kannada handwritten alphabet, kagunita and ottakshara. Data collection is done using the graphics tablet and stylus pen. Data collection consists of 25 samples of 49 characters of Kannada alphabet, 25 samples of kagunitaof all the consonants, and 25 samples of 34 ottakshara. In total, 15,675 samples are collected. The main techniques used here are Image Processing and Deep Learning. 80% of data is used for training and 20% for testing. The classifier used in the Deep Learning is CNN which yields 97.09% accuracy.

Key Words: Graphics Tablet, Stylus Pen, Deep Learning, CNN, Image Processing

1. INTRODUCTION

Basically, Kannada is a major language in Karnataka. Kannada is the official language in Karnataka and all the notifications in government institutions are in Kannada language.RecognizingthehandwrittenKannadawordsis theintentionofthispaper.Insteadofusingthekeyboard and typing Kannada by Nudi in laptop or computer, if graphicstabletisusedtowrite,whateverwegiveinput,that will be automatically printed. This decreases the burden. Onewhoisunawareofkeyboard,evenhecanwriteandget thewrittenwordsinthestandardtextform.

RecognitionisdonebyusingImageProcessingandDeep Learning.InDeepLearningConvolutionNeuralNetworkis the best classifier that suits proper recognition of online Kannadahandwriting.

Datacollectionof15,675samplesiscollectedtorecognize alphabet,kagunitaaswellasottakshara.Thepenmovement onthetablethelpstocollectthedataset.Eachdataisstored in different folders.Each folder consistsof 25samples of Kannadacharacter.Pythonlanguageisusedtowritecode fordatacollectionanddataprediction.

2. LITERATURE SURVEY

ResearchisgoingonregardingrecognitionofKannada handwritinginonlinemode.KeerthiPrasadG,VinayHegde [1]proposedonlineKannadahandwritingrecognition.Data iscollectedthroughthepenmovementonthegraphicstablet.

It can be writer dependent or independent. This work focused on writer independent handwriting recognition. Anirudh Ganesh[2] proposed for Kannada numbers using CNNandDBN.Thedatacollectionis74,000andtheaccuracy ofrecognitionis97%inCNNand98%inDBN.SalmaShofia Roshyda, Tito Waluyo Parbuda[3] proposed for various handwritingrecognition.AfterreviewingtheyconcludedCNN isthebestclassifierwhichyieldsmoreaccuracy.Gandhana M[4]propsedforrecognitionofonlineKannadacharacters andnumbers.BVijayKumarandRamakrishna[5]proposed for recognition of Kannada printed text. Rampalli and Krishnamurthy[6]proposedforrecognitionofcharactersin onlineandoffline.Recognitioninonlineis 11%morethan offline.RaghaLR et al., [7]therecognizedforcharacterswith recognitionresults86%forvowelsand65%forconsonants. ShashikalaParameshwarappa etal.,[8]proposedforKannada handwritten character recognition with 99.09% accuracy usingSVMClassifier.ParikshithH et l.,[9]proposedamethod in which any new style characters can be predicted. Ganapatsingh G Rajputh et al., [10], proposed a method to recognizecharacterswithK NNandSVMclassifiers.

3. DATASET

Thedatasetiscollectedthroughgraphicstabletandstylus pen.tocollectdatasetweusecommandprompt.Thestepsare asfollows.

1. Createafolderandlabelit.Opencommandpromptby the instruction ‘cmd’ the command window will be displayed.Type,‘code.’thecodewillbeopened,make necessarychangesregardingcharactercountandfolder name.Saveitbypressing‘S’.

2. Go to the command prompt window, type ‘activate project’.

3. Next,type‘pythondataset_main.py’.Itprovidesacheck windowtocollectdataset.Collectdatasetof25samples of a specified character. Press ‘S’, to save. Press ‘Q’ to exit.

4. The data will be cropped and stored in the specified folder.

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072

5. Create new folder before collecting data for the specifiedcharacterandlabelit. Repeatstep3,untilthe collectionofdataforallthecharacters.

6. Close the window after the collection of data. (a) (b) (c)

Fig 1 Datasetcollection(a)alphabet,(b)kagunita (c)ottakshara Fig 2 Croppingofottakshara

three types, namely, character segmentation, word segmentation,sentencesegmentation.

Fig 3 Datasetcollectedinfolders

4. METHODOLOGY

Thissectiondescribesthemethodsthatareemployed.

Data acquisition

ThedatasetiscollectedbyusingtheGraphicstabletand styluspen.Oncethegraphicstabletisconnectedtolaptop, theconnectionissetup.Thentheedatasetiscollectedand saved in the specified folder. The size of the ottakshara shouldbelessthan50%sizeofthealphabetorkagunita.So thatitwillbeappendedtothemaincharacter.

Pre processing

Here the collected data is subjected to the resizing. Whateverthesizeofinput,theimagewillberesizedto52X 52.Datacollectionisthroughonline,sothereisnoneedto remove noise. The collected data is also subjected to binarization.Thecharactersappearingontheblackcolored checkwindowwillbewhite.Thedatawillbesegmentedinto

Fig 4 Flowdiagramofproposedmethodology

The flow diagram of the proposed methodology is shown above.Itdepictstheflowoftheprocssesthatareinvolved.

ItmainlyusesthetechniquesofImageProcessingand CNNinDeepLearningwhicharebestsuitedforthispaper. Asonlinerecognitionisimplementedthereisnooverheadof noise removal, scanning the characters etc. Online recognition is very interesting and can be widely used. Online recognition gives better results and more accurate whencomparedtothatofofflinerecognition.Therewillbe nodiscontinutiesbecausedilation,amorphologicalopration isemployed.Thecolorimagesareconvertedtograythento blackandwhiteforpreprocessing.

Segmentation

After data collection, the data need to be segmented. The segmentationisofthreetypes.

CharacterSegmentation

WordSegmentation

SentenceSegmentation

Let us take an example of sentences in which alphabet, kagunitaandottaksharaarepresent.

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072

then the remaining next processes are easily carried out. Afterdatacollectionthecharactersneedtobesegmented. The data collection contains 25 characters in each folder. Therewillbe15,675charactersneedtobesegmentedand savedintherespectivefolders.Thedatasetincludesthese characterswhichareshownbelow.

Fig 5:Segmentationofcharacter,wordandsentence

Inthecaseofottakshra,whentheimageistaken,thecolor conversionofcharactersfromcolortograyandthenfrom graytoblackandwhiteconversiontakesplace.

Bonding boxes are applied. To identify the presence of ottakshara,theheightoftheottaksharashouldbelessthan 50% height of the original character. If not, there is no ottaksharaappendedtothecharacter.

Intheabovedepictedfigure5,thecharactersegmentationis showninthinredline,wordsegmentationisstoredinthe thick red rectangular box. The sentence segmentation is showninbluecolorrectangularbox.

Intheoutputwindowwhentheusergivestheinput,the segmentedcharacters,wordsandsentencesaresavedina folder.Let’stakeanexample.Iftheimageissavedas01 01 01 0.png,itmeansthattheesegmentationisdoneforfirst sentence,firstwordandfirstcharacter.0atthendtellsthat thereisnoottakshara.If1ispresentattheenditmeansthat thereisapresenceofottakshara.

Eachcharacterextendspixelto5pixelstowardsright. Sothatonewordisrecognized.

If 50 pixels are extended then it considers it as sentence. Rectangular shape is used to recognize the different words. To recognize the characters 5 pixels are extended.

Image processing is the main technique that is employedheretodosegmentation.Croppingisdoneinthis process.Thepixelsareextendedmainlytogetthecontentof data,sothatnoinformationislost.

Aftercharactersegmentation,ifthesegmentedcharacters thatarestoredinthefoldersarenotproperlysegmented,it leadstowrongrecognitionofcharacters.Suchdataneedto bediscarded.

Ifthesegmentedcharacterscontainthepropercharacter thatarewrittenbytheusers,thensuchdataisconsideredas itgivesthebestaccuracyintheoutput.Datasegmentation playsaveryimportantrole.Ifsegmentationisdoneproperly

Fig-6 KannadaAlphabets

Fig 7 KannadaKagunita

Fig-8 Kannadaottakshara

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072

CNN

UsingtheCNNmodel,80%ofthedataisusedfortraining and20%ofthedatafortesting.Amongtheavailableneural network algorithms CNN is considered as the best. It consists of three layers, namely 1. Input Layer 2. Hidden Layers3.OutputLayer

Training and Testing

Inthetrainingphase80%ofthecollecteddata isused and20%ofthedatafortesting.Theaccuracyoftrainingand validation increases and the training and validation loss decreases.

Fig 9 CNNArchitecture

FeatureextractionandclassificationaredoneusingCNN. Hidden layers in CNN are: 1.Convolution Layer 2.Pooling Layer 3.ReLUCorrectionLayer 4.Fullyconnectedlayer5. Softmaxlayer.

Labeling the data

Each collected data is assigned with a number. The foldernamethatisassignedtothecharacterisassignedto the standard characters. So that computer recognizes easilyandefficiently.

Fig 11 TrainingandvalidationAccuracy

Fig-12 Trainingandvalidationloss

5. RESULT

Anacondapromptisusedheretogetoutputwindow.There arethreewindowswhichdisplaytheresult.First window enablesusertogiveinputandinsamewindowoutputwill bedisplayed.

Fig-10 Labelingthedata

Fig 13 Inputandoutputdisplayedinthesamewindow

Thesecondtypeofoutputwindowenablesusertogive inputinone‘check’windowandaftersavingitbypressing ‘S’,theresultwillbedisplayedinthe‘result’window.

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072

‘check’ window and displays the output in the ‘result’ window.Here,theinputcanbeclearedbypressing‘C’on thekeyboard.Iftheinputisproperlywrittentheoutputwill bedisplayedassoonastheuserpress’S’onthekeyboard

This window also does not provide editing the output, meansincreasingthefontsizeorchangingthefontstyleor copyingthecontentetc.

The input window is somewhat wider, but the output windowisverysmall.Itdoesnotdisplayiftheusergives inputas a paragraph. Butitwill recognizethecharacters perfectly.Itdoesnotneedmuchtimetoreceiveinputand displayit.Theoutputstylecanbechangedwhentheuser hassomedifferentfontstylesdownloadedinthesystem.

Inthethirdtype,theinputandoutputwindowbothare wide.

In the input window, where user gives the input, if he writesanythingwronghemayclearthescreen.Theinput canbeclearedbyusing‘C’buttononthekeyboard.Then usercangiveanewinputintheinputwindow.Thisprocess goesonforeverynewinput.

Fig.14 Inputandoutputdisplayedinseparatewindows

Thiswindownotonlydisplaysthecombinationofkagunita, itcanalsoexecutecharactersofalphabet.Aswegivesome spaceaftereachcharacter,thespacewillalsobedisplayedin theoutput‘result’window.Thisisillustratedinthefigure14.

Oncetheinputisfedtotheinputwindow,theuserhasto press‘S’.ittakessometimetoaccepttheinput.Thenthe userhastoopenthe‘data’fileinhissystemwhichissaved before.

Assoonasthefileisopened,outputwillbedisplayedin theenotepad.Theusercanchangethefontstyle,fontsize, savetheoutput.Thesechangescanbemadebytheuserin thisoutputwindow.

ThiswindowenablesusertogetthestandardKannada textforcharacters,words,sentences,paragraphetc.Sothis windowisconsideredasthebestamongthethreeoutput windows.

Themainprobleminthispaperishugedatacollection. Theuserhastoknowhowtoholdthestyluspen,howto holdthegraphicstablet,wheretostartandwheretostop writing.

Fig 16 Outputdisplayedinnotepad

The output window in which input and output are displayedinthesamewindow,thesizeoinputwindowis small. The input can be cleared by the user by pressing ‘clear’buttonthatispresentbelowtheoutputdisplay.

Theoutputcannotbeedited.Inotherwords,theoutput cannotbecopiedtoanyotherfiles.cut,copy,pasteoptions arenotavailable.

The next type of output window in which input and outputaredisplayedseparatelyenablestheusertowritein

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072

If any scratches are made to the words in the input windowwillnotbedisplayedintheoutputwindow.Ifany symbolsarefedbytheuserintheinputwindow,thoseare not executed in the output window, because the dataset collectioniscomposedofcharactersonly.Thedatasetisnot collectionofsymbols.

Theoutputpresentedinthethirdtypeisveryhelpfuland canbeusedtocopytheoutputinthedesiredformatinthe otherfilesorfolders.InKarnatakathisastheofficialnotices arecirculatedingovernmentinstitutionsandofficesisin Kannadalanguage,therewillbevariationinthefontsizein theheadingandcontent.Suchmodificationscanbemadein theoutputwhichisdisplayedinthestandardKannadatext format.

Fig 17 Graphicstabletconnectedtothelaptop

Itiseasyforuseronewhoknowsall thesetechniques. That person can collect data very fast. There will be no overhead of discarding the characters once they are cropped and saved in the respective folders. If any characters are not cropped properly, or if two or more charactersarecroppedinthesameimage,suchimagescan bediscarded.

Theoutputcanbeobtainedatthatinstantassoonasthe userprovidestheinputintheinputwindow.Thismakesthe timetobesaved.InsteadoftypingKannadainlaptopand get it printed, if digital padis connected and written it is easytouseandeasytogettheoutput.

ManypeopleareunawareoftypingKannadainthelaptop or in the computers. It is very useful for them. As the technologyisrapidlyincreasingfromdaytodayeveryone hastoknowandimplementandgttheworktobdonein easywaywithlesstime.Inputgiventothesystemcaneven beasongorproverb,itwillbeprinted.

In this last method, thee background is white and the inputisblackcolor.Theoutputdisplayedisalsoinblack color.ColorconversionprocessusestheImageProcessing technique.Thethicknessofinputcharactersismorewhen compared to that of the output. The thickness of output characterscanbeincreasedbyincreasingthefontsizein thenotepad.

Fig-18 Croppedimagesarestoredinthefolders

While collecting dataset for ottakshara, the size of the data should be small and many of the collected data is discarded. Only the properly cropped characters are considered.Againtheremainingsamplesarecollected.

Whilecollectingdatasetforkagunita,somecharactersare widerbuttheyshouldbecollectedinthespecifiedsize.If the supportingcharacters thatare appended to the main charactersaremissingitisalsoaoverhead.Suchdataneed tobecollectedtillthecollectionof25samplesiscollected.

Theaccuracyisnot100%.Ifmoresamplesarecollected, then the accuracy will be increased. The input for some charactersneedstobewritteninaspecifiedformat.Ifthe input is not fed in that specified format, the detection of suchcharactersiswrong.

Itenablesusertogivemoreinputwordsandsentences, and displays the output. It is the most beneficial and the bestwindowforusertogiveinputandgettheoutputand increasethefontsizetothedesiredfontsizeoftheuser. Theottaksharawillbeappendedtomaincharacter.Ifthe ottaksharaislessthan50%ofthemaincharacter.Training accuracy is 99.23%, training loss is 0.0251%, testing accuracyis97.09%andtestinglossis0.210%byTensor Flowframework.

AsCNNclassifierisused,itgivesthemoreaccuracywhen compared to that of other methods. CNN is used in the DeepLearningthatisemployedhere.Ittrainsthesystem aswetrainourmind.

CNN classifier is selected after reviewing so many reference papers. Other classifiers are also used in the referencepapersbutCNNgivesthebestresult.Sothatthe prediction loss can be minimized and accuracy is more duringtrainingaswellastestingthedata.

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072

Theottakshara arecollectedina mannerthatthemain character is not used along with it in the dataset. This enables to get the output in such a manner that any ottakshara can be appended to any consonant or any kagunita.itgivesawidevarietyofcharacters,wordsand sentences to be available in the standard Kannada text form.

Theoutputexecutioninthe thirdtypeismorepreferred andwideroutputdisplayenablestomakesomanychanges likechangingthefontsizeandotherbenefits.

Theoutputcanbesavedinthestandardtextandcanbe openedatanytime anywhere ifitissavedin the laptop. Thereisnoneedofinternet toopen thesavedfileinthe system. No need of other huge external devices to be connected, it just needs a light weight digital pad to be connectedtothelaptop.

Theaccuracywillbe100%ifdataiscollectedfrommany peoplewhichconsistsasampleof100or200.Thechance ofgettingthebestaccuracyisthereforhugedatacollection of Online handwritten Kannada characters. There is no overheadofmisplacingandlosingthedatasitmayoccurin theofflinecharacterrecognition.

Theavailabledatasetcangiveawidevarietyofwordsina standard Kannada text form. Online Kannada Words recognitionisveryhelpfulandeasytoimplementinmany institutions and universities where Kannada is used the mostforcircularswhichareinKannadalanguage.Asthe accuracy is more there is less overhead of wrong recognitionofthecharacters.Theaccuracycanbeknown by‘TensorFlow’.Theconfusionmatrixalsotellsaboutthe accuracyoftheoutputfortheproposedmethod.

Thecalculationcanbemadeforoutputdisplayintermsof centimeterandpixels.

length and width of the A4 sheet is 21 X 29.7 cm. The marginconsumes2.54cmintheleftand2.54cmintheright sideoftheA4sizepage.

There are 33 Kannada characters in the first line. 11Kannadawordscanbewrittenintheaboveconsidered exampleinA4sizepageasillustratedinthefigure19.After 11words,12th wordcannotbeexecutedinthefirstline.It hasmovetothenextlineofthesentence.

CALCULATION

SizeofA4sheet :21X29.7sq.cm

Marginleftintheleftside :2.54cm

Marginleftintherightside :2.54cm

Totalwidthleftformargin :5.08cm

Available space for executable line (A) :

A = Total width 2(margin)

A = Tw 2M

WidthofA4sheet :21cm

Availablewidthforthefirstline:21cm 2(2.54)cm

Availablewidthtoforfirstline :15.92cm 2(margin)isconsideredincalculation.Itisbecause,the marginleftintheleftsideaswellasrightside. where,TwistotalwidthofA4sheet

2Mismargininleftandright

Iftheabovecalculationiswrittenintermsofpixels, Marginleftinleftside =96pixels Marginleftintherightside =96pixels

Totalwidthformargin(2M) =192pixels

WidthofA4sheet(Tw) =793.7007874pixels

Width provided to print a line (W)

W= Tw 2M

W= 601.700784 pixels

Width available for executing a line is given by the formula:

Fig 19 OutputintheMicrosoftWord

Letusconsidertheabovefigure.19.Theoutputofnotepad isdisplayedintheMicrosoftWord.ThepagesizeisA4.The

Wc = ∑Wn + ∑Gw

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Where, Wc refers to available width to execute a line consistingofcharactersandspaceaftereachword.

Wnreferstowidthoccupiedbysumof characters inthefirstline

Gwreferstogapleftaftertheexecutionofeachword inasentenceinthefirstline.

6. CONCLUSION and FUTURE SCOPE

Inthispaper,onlineKannadahandwritingisrecognizedin the online mode using the graphics tablet and stylus pen. Datacollectionisveryeasy.Preprocessing,segmentationare alsoimportantinthispaper.Trainingandtestingaredone usingCNNclassifier.Pythonisthelanguageusedtowritethe coding for data collection and data prediction. Image Processing and Deep Learning techniques are used in the proposed paper. This is very beneficial especially in KarnatakaasKannadaisanofficiallanguageofKannadigas.

In spite of wasting time in typing in the keyboard, the outputwillbedisplayedinthestandardformaswewritethe words on the graphics tablet. The output is displayed in threedifferentwindows,amongthemthelastwindowisthe best. Because the user can modify the font style and size, copytheoutput,cutitorpasteitanywhere,theuserwishes todoitinotherfolders.

There are many methods to get the characters to be printed,amongthemthisisthebestmethodtobeemployed. Itisveryhelpfulasthelossofdataisless,itdoesnotneed inktowrite,itjustusestheestyluspen.Asthetechnologyis improvingthiskindoftechniquesneedtobeemployedto gettheworktobedoneevenbyauserwhoisunawareof using‘Nudi’.

This paper recognizes the Kannada characters in online modewhichisverybeneficialanthereisnolossofimages that are collected and no overhead of noise removing. Nowadays,thereisahugescopeforhandwritingrecognition. ThispaperexplainsthehandwritingrecognitionofKannada languagewhichisveryhelpful.Itnotonlydisplaysoutputin wordsorsentenceformat,butalsoinparagraphformatalso. Itenablesusertomodifythefontstyleandsizeaccordingto userwish.

Thereisnotensionfortheuserregardinglosingthedata collected,ormisplacedorlostduetoanyreason.Thereisno needofpenandpaper.Thereisnoneedtoscanthedatafor savingandsegmentinginthelaptoporcomputers.

There is no need o battery for stylus pen. If the nibs are flattenedtheycanberemovedandothernibscanbeeasily insertedwhichtakesonlyfewsecondsoftime.Anynumber ofdatacanbecollectedwithoutpenandpaper.Thereisonly one time investment for the graphics tablet. There is no

interferenceofnoisewhichisthemainprobleminaudioto textrecognition.

The execution of Online Kannada characters in the standardtextiseasyandhelpfulformanypurposeswithless time and less resources anywhere and anytime without learning ‘Nudi’ which is popularly used by most of the KannadigastogetthestandardKannadatextformatinthe laptoporpersonalcomputers.

Online Kannada Recognition is very beneficial for many Kannadigas present across the world. It recognizes the characters,words,sentencesandparagraphsandgivesthe outputatthatinstantonly.

There is no need to depend on the people who can type ‘Nudi’verywell.OneisunawareofKannadatypingcanalso get the Kannada in the standard text format. Once the processfromdatacollectiontotestingiscompleted,theuser can get any kind of output whether in the form of single characterorintheformofparagraphofmanypages.

In future, characters or words with two or more ottakshara can be executed which gives a completion of executing any characters / words. More accuracy can be made available in future. The biggest challenge for the researchersinKannada handwriting recognition inonline modeistoget100%accuracy.Nolossinaccuracyneedtobe achieved. There are many words in Kannada with two ottakshara,suchwordsareincludedinthemanysentences. If those words are also included, then only there is recognitionofsuchwordsmakesthewordsandsentences meaningfulandappropriateinKannadalanguage.

Theexecutionof‘O’ withanyKannada charactersshould alsobemadetoexecutetheminthestandardform.Onlythe consonants attached with ‘O’ are executing, but other kagunitashouldalsobemadeexecutedwith‘O’(anuswara). SomecharacterswhichareusedinKannadabutnotincluded incharacters, kagunitaandottaksharaneedtobeexecuted.

Theabovementionedpointscanbeimplementedinoffline aswellasonlinecharacterrecognition.Thereisahugescope infutureforimplementingtheabovementionedpoints to make a blended recognition in a single system which recognizes the Online Kannada characters as well as the offline Kannada characters so that whichever method is convenientforuser,usercanuseitaccordinghischoice.

Accuracy need to be increased by trying some other different methods that are widely available. There is a chancetogetthe100%accuracy fortheinputprovidedby theuserinthefuturewithnodelay.

Insteadofwriting in the inputwindowall thecharacters can be made available in the portion of a window, in one clickonthatcharacterthatcharacterneedtobeprintedin thefuture.

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal |

International Research Journal of Engineering and Technology (IRJET) e ISSN: 2395 0056

Evenwiththelessdatacollectiontheaccuracyoftheoutput has to be made 100% in the future.So that the writer independentsystemisimplementedandcanbeusedinthe future.Sothatthereisnonecessityofwritingthewordsin theinputwindowasitisinthedatacollection.Itrecognizes for any user handwritten characters in Kannada language whichiswidelyusedinKarnatakabyKannadigas.

REFERENCES

1. KeerthiPrasadG,VinayHegde,“RecognitionofOnline Handwritten Isolated Kannada Character using PCA andDTW”,InterntionalJournalofRecentTechnolofy andEngineering(IJRTE),2019.

2. Anirudh Ganesh, Ashwin R Jadhav, K A Cibi Pragadeesh,“DeepLearningApproachforRecognition of Handwritten Kannada Numerals”, Springer InternationlPublishing,2018.

3. Salma Shofia Rosyda and Tito Waluyo Purboyo, “A ReviewofVariousHandwritingRecognitionMethods”, InternationlJournalofAppliedEngineeringResearch, Vol.13,2018,pp1155 1164.

4. Gandhana M H, “ Online Kannada Handwriting Characters and Numerical Recognition using CNN Classifier,IJERT,July2021.

5. BVijayKumarndAGRamakrishnan,2004.RadialBais FunctionandSubspaceapproachforPrintedKannada PrintRecognition.Int.Conf.Acoustics,SpeechSignal Proc.2004

6. Rampalli R Ramkrishnan, Angarai G, “ Fusio of Complementary Online and Offline Strategies for recognition of Handwritten Kannada Characters”, JournalofUniversalComputerScience,17(1).pp81 93.2011.

7. Ragha L R, Sasikumar M, “Feature Analysis for Handwritten Kannada Kagunita Recognition”, International Journal of Computer Theory and Engineering,IAC SIT3(1),pp.1793 8201,2011.

8. Shashikala Parameshwarappa, B V Dhandra, “Basic KannadaHandwritteenCharacterRecognitionSystem usingShapeBasedandTransformDomainFetures”, International Journal of Advanced Research in Computer and Communication Engineering, Vol 4, Issue7,July2015.

9. H Parikshith et al., “Handwritten Character Recognition of Kannada Languag using Convolution ConvolutionNeuralNetworksandTransferLearning”, IOPConf.Series,2021.

10. Ganapatsingh G Rajput, Rajeshwari Horakeri, “HandwrittenKannadaVowelCharacterRecognition usingCrackCodesandFourierDescriptors”,December 2011.

BIOGRAPHIES

Sangeetha G R is pursuing her MTech in Digital Communication andNetworking,ECEDepartment from UBDTCE, Davanagere, VTU, India. Her major interest are in ImageProcessing&DeepLearning.

Dr.LakshmanNaikaR,completed B.E (E&C) in Malnad College of Engg, Mysore University, MTech (ElectronicsEngg)inB.M.S.College of Engg, VTU, PhD in Image Processing,JainUniversity.

Volume: 09 Issue: 06 | Jun 2022 www.irjet.net p ISSN: 2395 0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page3364