Mahidol University Logo
Faculty of ICT, Mahidol University
 

Admissions

Printable Version

 

SEGMENTATION OF CATALOG DOCUMENTS USING CROSSING COUNT AND PROJECTION TECHNIQUES

 

TITLE SEGMENTATION OF CATALOG DOCUMENTS USING CROSSING COUNT AND PROJECTION TECHNIQUES
AUTHOR SMITH MANGMEETAKUN
DEGREE MASTER OF SCIENCE PROGRAMME IN COMPUTER SCIENCE
FACULTY FACULTY OF SCIENCE
ADVISOR SUKANYA PHONGSUPHAP
CO-ADVISOR CHOMTIP PORNPANOMCHAI,
 
ABSTRACT
This research is aimed at using geometric layout analysis as an approach to identifying the location of both picture and text in printed documents. A combined technique of Crossing Count and Projection was used to divide an original image into smaller segments. The result of this process was categorized into two groups: the block containing text object only, called Text-Line and the block containing picture and text, called Picture-Text block. The picture and text in Picture-Text block was separated by using Run-Length Smoothing Algorithm (RLSA) and Recursive X-Y tree Cutting (RXYC). Subsequently, the text block in Picture-Text was merged with a set of Text-Line block and individual character objects were extracted for optical character recognition, (OCR). Finally, the size estimation was used to determine block size; small and large. The small block was expected to contain a single character and was ready to perform OCR. The large block, containing two or more characters needed to be separated before performing OCR. In this research picture and text object were extracted from 51 catalog chromosome images. Results showed that 10,020 relevant objects were extracted into two main groups: picture and text. The accuracy derived from the picture blocks detection was 99.27% and from the text block detection it was 95.54%. In conclusion, the study results confirmed the research objectives.
KEYWORD DOCUMENT IMAGE ANALYSIS / PROJECTION PROFILE / CROSSING COUNT / RECURSIVE X-Y CUTTING / RUN-LENGTH SMOOTHING

 

Go to Top

 

ICT Building, Mahidol University, 999 Phuttamonthon 4 Road, Salaya, Nakhonpathom 73170 Tel. +66 02 441-0909 Fax. +66 02 849-6099
Mahidol University Computing Center, The Faculty of ICT, Mahidol University , Rama 6 Road, Rajathevi, Bangkok 10400 Tel. +66 02 354-4333 Fax. +66 02 354-7333