Advanced

Template Based Recognition of On-Line Handwriting

Sternby, Jakob LU (2008)
Abstract
Software for recognition of handwriting has been available for

several decades now and research on the subject have produced

several different strategies for producing competitive recognition

accuracies, especially in the case of isolated single characters.

The problem of recognizing samples of handwriting with arbitrary

connections between constituent characters (emph{unconstrained

handwriting}) adds considerable complexity in form of the

segmentation problem. In other words a recognition system, not

constrained to the isolated single character case, needs to be able

to recognize where in the sample one letter ends and another begins.

In the research... (More)
Software for recognition of handwriting has been available for

several decades now and research on the subject have produced

several different strategies for producing competitive recognition

accuracies, especially in the case of isolated single characters.

The problem of recognizing samples of handwriting with arbitrary

connections between constituent characters (emph{unconstrained

handwriting}) adds considerable complexity in form of the

segmentation problem. In other words a recognition system, not

constrained to the isolated single character case, needs to be able

to recognize where in the sample one letter ends and another begins.

In the research community and probably also in commercial systems

the most common technique for recognizing unconstrained handwriting

compromise Neural Networks for partial character matching along with

Hidden Markov Modeling for combining partial results to string

hypothesis. Neural Networks are often favored by the research

community since the recognition functions are more or less

automatically inferred from a training set of handwritten samples.

From a commercial perspective a downside to this property is the

lack of control, since there is no explicit information on the types

of samples that can be correctly recognized by the system. In a

template based system, each style of writing a particular character

is explicitly modeled, and thus provides some intuition regarding

the types of errors (confusions) that the system is prone to make.

Most template based recognition methods today only work for the

isolated single character recognition problem and extensions to

unconstrained recognition is usually not straightforward. This

thesis presents a step-by-step recipe for producing a template based

recognition system which extends naturally to unconstrained

handwriting recognition through simple graph techniques. A system

based on this construction has been implemented and tested for the

difficult case of unconstrained online Arabic handwriting

recognition with good results. (Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • Prof Srihari, Sargur, CEDAR, NY, USA
organization
publishing date
type
Thesis
publication status
published
subject
keywords
on-line handwriting recognition segmentation dynamic programming clustering discriminative clustering
pages
164 pages
defense location
Lecture Hall MH:C, Centre for Mathematical Sciences, Sölvegatan 18, Lund university, Faculty of Engineering
defense date
2008-05-30 13:15
ISSN
1404-0034
language
English
LU publication?
no
id
31a7a409-a754-43f8-ae21-35f383468618 (old id 1148383)
date added to LUP
2008-05-07 08:53:54
date last changed
2016-09-19 08:45:00
@misc{31a7a409-a754-43f8-ae21-35f383468618,
  abstract     = {Software for recognition of handwriting has been available for<br/><br>
several decades now and research on the subject have produced<br/><br>
several different strategies for producing competitive recognition<br/><br>
accuracies, especially in the case of isolated single characters.<br/><br>
The problem of recognizing samples of handwriting with arbitrary<br/><br>
connections between constituent characters (emph{unconstrained<br/><br>
handwriting}) adds considerable complexity in form of the<br/><br>
segmentation problem. In other words a recognition system, not<br/><br>
constrained to the isolated single character case, needs to be able<br/><br>
to recognize where in the sample one letter ends and another begins.<br/><br>
In the research community and probably also in commercial systems<br/><br>
the most common technique for recognizing unconstrained handwriting<br/><br>
compromise Neural Networks for partial character matching along with<br/><br>
Hidden Markov Modeling for combining partial results to string<br/><br>
hypothesis. Neural Networks are often favored by the research<br/><br>
community since the recognition functions are more or less<br/><br>
automatically inferred from a training set of handwritten samples.<br/><br>
From a commercial perspective a downside to this property is the<br/><br>
lack of control, since there is no explicit information on the types<br/><br>
of samples that can be correctly recognized by the system. In a<br/><br>
template based system, each style of writing a particular character<br/><br>
is explicitly modeled, and thus provides some intuition regarding<br/><br>
the types of errors (confusions) that the system is prone to make.<br/><br>
Most template based recognition methods today only work for the<br/><br>
isolated single character recognition problem and extensions to<br/><br>
unconstrained recognition is usually not straightforward. This<br/><br>
thesis presents a step-by-step recipe for producing a template based<br/><br>
recognition system which extends naturally to unconstrained<br/><br>
handwriting recognition through simple graph techniques. A system<br/><br>
based on this construction has been implemented and tested for the<br/><br>
difficult case of unconstrained online Arabic handwriting<br/><br>
recognition with good results.},
  author       = {Sternby, Jakob},
  issn         = {1404-0034},
  keyword      = {on-line handwriting recognition segmentation dynamic programming clustering discriminative clustering},
  language     = {eng},
  pages        = {164},
  title        = {Template Based Recognition of On-Line Handwriting},
  year         = {2008},
}