Page - 159 - in Document Image Processing

J. Imaging 2018,4, 39 Figure7.Samplewordimageswritten in12different Indianscripts. OurdevelopeddatabasehasbeennamedasCMATERdb8.4.1,whereCMATERstands for ‘Centre for Microprocessor Applications for Training Education and Research,’ a research laboratory at Computer Science andEngineeringDepartment of JadavpurUniversity, India,where the current database is prepared. Here, db symbolizes database, the numeric value 8 represents handwritten multi-script Indic imagedatabaseandthevalue4 indicatesword-level. In thepresentwork, thefirst versionofCMATERdb8.4hasbeenreleasedasCMATERdb8.4.1. Thedatabase ismadefreelyavailable athttps://code.google.com/p/cmaterdb/. 4.2. PerformanceAnalysis Theclassifiercombinationapproaches,describedabove,areappliedonadatasetof7200words dividedinto12classeswithequalnumberof instances ineachof them. 12classes refer to the12 Indic scripts thathavebeenstudiedbeforeandforwhich theMLPclassifier results canbeobtainedwith highaccuracy. TheclassesnumberedfromA toLareDevanagari,Bangla,Oriya,Gujarati,Gurumukhi, Tamil,Telugu,Kannada,Malayalam,Manipuri,UrduandRoman in thatparticularorder. 159