Design of an off line recognition system based on 2d discrete wavelet transform for Latin hand written text

Dissertant

Khalil, Yazan Adnan

University

University of Technology

Faculty

-

Department

Department of Electrical Engineering

University Country

Iraq

Degree

Ph.D.

Degree Date

2006

English Abstract

Off-line handwriting recognition is the task of determining what letters or words are present in handwritten text.

It is of significant benefit to man-machine communication and can assist in the automatic processing of handwritten documents.

It is a subtask of the Optical Character Recognition (OCR), whose domain can be machine-print only.

This thesis presents a design to an Off-line Handwriting Recognition System.

The introduced system is a character-based recognition and it is a writer independent system.

It is designed to recognize the discrete and mixed handwritten styles of Latin script.

The recognition responsibility of the proposed system is for 52 character classes [uppercases (A-Z) and the lowercases (a-z)].

The suggested system includes the essential stages needed tor most to tree pattern recognition systems.

These stages are the preprocessing stage, the features extraction stage, the pattern matching and classification stage and the post processing stage.

The design work is for new approaches of the steps of the preprocessing stage like the noise removal step of the handwritten document, lines (text sentences) skew estimation and correction step, character 'slant removal and the character segmentation steps.

The tests appeared very fast, satisfactory and provided accurate results.

The design also includes a wavelet-based feature extractor.

A Two Dimensional Discrete Wavelet Transform is used to extract the features vectors of the handwritten character patterns that are required for the classification stage.

The classification stage is designed by using a minimum distance classifier depending on Euclidean Distance which has a high speed performance.

The design also includes a modest post processing stage that makes a consistency between the recognized characters within the same word in relation to their upper and lower cases.

The design task is in two phases.

The first one is the training phase which includes the handwriting collection of the training samples, some necessary preprocessing steps and the feature extraction for all characters classes.

At the end of this phase the feature vectors are ready to the second phase that is the recognition phase.

The performance test shows an improvement in the recognition task with recognition rate of about 92.6 % for uppercase characters recognition and about 89.9 % for the lowercases.

Main Subjects

Electronic engineering

Topics

• Writing
• Data processing

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-306312