izpis_h1_title_alt

Učinkovitost prepoznave izbranih črkovnih vrst s pomočjo OCR programskih orodij
ID Boršič, Blaž (Author), ID Urbas, Raša (Mentor) More about this mentor... This link opens in a new window, ID Možina, Klementina (Co-mentor)

.pdfPDF - Presentation file, Download (3,28 MB)
MD5: CA0E0314EC8239EDBADC93CED2A65D26
PID: 20.500.12556/rul/ec1beb13-21e5-4a26-9912-b619f09c4048

Abstract
Danes obstajajo številne različice programskih orodij za optično prepoznavanje znakov (ang. Optical character recognition; krajše OCR), ki omogočajo optično prepoznavo in digitalizacijo nabora različnih znakov in črkovnih vrst. OCR programska orodja se razlikujejo v načinu delovanja in kakovosti prepoznave, rezultati pa so lahko zelo različni. Največkrat se težave pojavijo pri prepoznavi nabora znakov kompleksnejših in nenavadnih črkovnih vrst, pogosto težave delajo tudi posebni znaki (npr. šumniki) in besedila postavljena v preglednicah. V teoretičnem delu diplomske naloge je predstavljena zgodovina OCR programov, njihov način delovanja, opis izvedbe digitalizacije in izboljšave prepoznavanja. Poleg omenjenega, so predstavljene tudi lastnosti optičnih čitalcev in programskih orodji za optično prepoznavanje znakov. V nalogi so opredeljene bistvene razlike med prosto dostopnimi in plačljivimi OCR programi, podane so tudi smernice izbire primernosti OCR programov za lastno uporabo. V eksperimentalnem delu naloge je predstavljena opravljena študija načina prepoznave posameznih izbranih črkovnih vrst s pomočjo plačljivih in brezplačnih OCR programov. V ta namen je bila izdelana ustrezna grafična predloga, ki vsebuje različno definirane in oblikovane elemente izbranih črkovnih vrst. Za primerjavo kakovosti pretvorbe sta bili poleg posameznih znakov v grafično predlogo dodani tudi preglednica s podatki in fotografija z besedilom. Na podlagi dobljenih rezultatov je bila opravljena natančna analiza, ki je omogočila možnost podajanja oz. predlaganja najugodnejše rešitve uporabe določenega brezplačnega OCR programskega orodja.

Language:Slovenian
Keywords:optično prepoznavanje znakov, OCR, črkovna vrsta, znak, besedilo
Work type:Bachelor thesis/paper
Organization:NTF - Faculty of Natural Sciences and Engineering
Year:2017
PID:20.500.12556/RUL-95394 This link opens in a new window
Publication date in RUL:20.09.2017
Views:1909
Downloads:365
Metadata:XML RDF-CHPDL DC-XML DC-RDF
:
Copy citation
Share:Bookmark and Share

Secondary language

Language:English
Title:Recognition efficiency of selected typefaces with OCR software
Abstract:
Today, there arenumerous versions of optical character recognition software (shorter OCR), that enable the optical identification and digitization of a set of various characters and alphanumeric types. OCR software tools differ in operating mode and recognition quality, and the results can differ significantly. Most often, problems arise in the identification of the character set of more complex and unusual alphanumeric types, often with special characters (e.g. postalveolar consonants) and texts, set out in tables. In the theoretical part of the thesis, the history of OCR programs, their operating mode, the description of the implementation of digitization and recognition improvements are presented. In addition, features of optical readers and optical character recognition software are also presented. In the thesis, the essential differences between free and payable OCR programs have been defined and the guidelines for selecting the suitability of OCR programs for personal use are also given. In the experimental part, a study on how to identify individual selected letters and typefaces using free and payable OCR software tools was conducted. For this purpose, an appropriate graphic design template was created, containing variously defined and shaped elements of the selected alphanumeric typefaces. To compare the conversion quality, in addition to the individual characters, a spreadsheet with data and a photo with a text was added to the graphic design template. Based on the obtained results, a detailed analysis was performed, which enabled the possibility of submitting – suggesting the best solution for using a particular free OCR software tool.

Keywords:optical character recognition, OCR, typeface, character, text

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Back