The use of synthesized images to evaluate the performance of Ocr devices and algorithms

This thesis will attempt to establish if synthesized images can be used to predict the performance of Optical Character Recognition (OCR) algorithms and devices. The value of this research lies in reducing the considerable costs associated with preparing test images for OCR research. The paper reports on a series of experiments in which synthesized images of text files in nine different fonts and sizes are input to eight commercial OCR devices. The method used to create the images is explained and a detailed analysis of the character and word confusion between the output and the true text files is presented. The synthesized images are then printed and scanned to mechanically introduce "noise". The resulting images are also input to the devices and analysis performed. A high correlation was found between the output from the printed and scanned images and the output from "real world" images.

Name thesis:degree_name

Master of Science (MS)

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Computer Science

Grantor dc:publisher

University of Nevada, Las Vegas

Year

1993

Author dc:creator

• Jenkins, Frank Robert

Statement dc:rights

• IN COPYRIGHT. For more information about this rights statement, please visit http://rightsstatements.org/vocab/InC/1.0/

Language dc:language

English

Identifier

https://oasis.library.unlv.edu/rtds/304

OAI identifier oai:identifier

oai:oasis.library.unlv.edu:rtds-1303