Mahidol University Logo
Faculty of ICT, Mahidol University
 

Admissions

Printable Version

 

PERSONAL INCOME TAX FORM READING SYSTEM

 

TITLE PERSONAL INCOME TAX FORM READING SYSTEM
AUTHOR SANGWAN INTASUK
DEGREE MASTER OF SCIENCE PROGRAMME IN COMPUTER SCIENCE
FACULTY FACULTY OF SCIENCE
ADVISOR SUKANYA PHONGSUPHAP
CO-ADVISOR RAWESAK TANAWONGSUWAN
 
ABSTRACT
This paper proposes a prototype system for reading numerical data on tax form images. The proposed system was used to read both handwriting and machine printed numerical data. The processes were divided into six steps: 1) Form Identification by modeling P.N.D.91 form image and matching reference components compared with the reference components on the form model using the Correlation Coefficient, 2) Component Block Decomposition by referring to the location of the information fields on the form model, 3) Numerical Field Extraction, 4) Digit Segmentation by using the technique of connecting components to separate isolated digit blocks and using the valleys of the upper and lower profiles of the touching digit to segment the touching digits, 5) Digit Feature Extraction using crossing sequence and directional codes histogram to extract features of a digit image, and 6) Digit Recognition using Neural Network. Experiments were performed by using 100 tax form images, which were composed of 50 Personal Income Tax form images (P.N.D.91) and 50 other tax form images. Results showed that the system could identify all of the forms correctly and could extract all component blocks on P.N.D.91 forms correctly. For digit segmentation and recognition, the accuracy rates were 98.06% (8,192 characters from 8,354 characters) and 95.08% (7,943 characters from 8,354 characters), respectively.
KEYWORD TAX FORM IMAGE ANALYSIS/ DOCUMENT IMAGE ANALYSIS/ FORM MODELING/ NUMERICAL RECOGNITION/ NEURAL NETWORK

 

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