Optilist käekirjatuvastust mõjutavad tegurid
Date
2018
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Optiline kirjatuvastus viitab tehnikale, mis konverteerib trükitud, kirjutatud või prinditud teksi masinkodeeritud tekstiks, võimaldades sellega paberdokumentide nagu passide, arvete, meditsiiniliste vormide või tšekkide automaatset töötlemist. Mustrituvastus, tehisintellekt ja arvuti nägemine on kõik teadusharud, mis võimaldavad optilist kirjatuvastust. Optilise kirjatuvastuse kasutus võimaldaks paljudel kasvavatel informatsiooni süsteemidel mugavat üleminekut paberformaadilt digitaalsele. Tänapäeval on optilisest kirjatuvastusest väljaskasvanud mitme sammuline protsess: segmenteerimine, andmete eeltöötlus, iseloomulike tunnuste tuletamine, klassifitseerimine, andmete järeltöötlus ja rakenduse spetsiifiline optimiseerimine. See lõputöö pakub välja tehnikaid, millega üleüldiselt tõsta optiliste kirjatuvastussüsteemide täpsust, näidates eeltöötluse, iseloomulike tunnuste tuletamise ja morfoloogilise töötluse mõju. Lisaks võrreldakse erinevate enimkasutatud klassifitseerijate tulemusi. Kasutades selles töös mainitud meetodeid saavutati täpsus üle 98% ja koguti märkimisväärselt suur andmebaas käsitsi kirjutatud jaapani keele hiragana tähestiku tähti.
Optical character recognition (OCR) refers to a technique that converts images of typed, handwritten or printed text into machine-encoded text enabling automatic processing paper records such as passports, invoices, medical forms, receipts, etc. Pattern recognition, artificial intelligence and computer vision are all research fields that enable OCR. Using OCR on handwritten text could greatly benefit many of the emerging information systems by ensuring smooth transition from paper format to digital world. Nowadays, OCR has evolved into a multi-step process: segmentation, pre-processing, feature extraction, classification, post-processing and application-specific optimization. This thesis proposes techniques to improve the overall accuracy of the OCR systems by showing the affects of pre-processing, feature extraction and morphological processing. It also compares accuracies of different well-known and commonly used classifiers in the field. Using the proposed techniques an accuracy of over 98% was achieved. Also a dataset of handwritten Japanese Hiragana characters with a considerable variability was collected as a part of this thesis.
Optical character recognition (OCR) refers to a technique that converts images of typed, handwritten or printed text into machine-encoded text enabling automatic processing paper records such as passports, invoices, medical forms, receipts, etc. Pattern recognition, artificial intelligence and computer vision are all research fields that enable OCR. Using OCR on handwritten text could greatly benefit many of the emerging information systems by ensuring smooth transition from paper format to digital world. Nowadays, OCR has evolved into a multi-step process: segmentation, pre-processing, feature extraction, classification, post-processing and application-specific optimization. This thesis proposes techniques to improve the overall accuracy of the OCR systems by showing the affects of pre-processing, feature extraction and morphological processing. It also compares accuracies of different well-known and commonly used classifiers in the field. Using the proposed techniques an accuracy of over 98% was achieved. Also a dataset of handwritten Japanese Hiragana characters with a considerable variability was collected as a part of this thesis.