Threshold selection and identification of character regions from complex scene images

In this research work, a comprehensive algorithm for alphanumeric character detection in general complex scene images is proposed, and specifically applied to car licence-number plate. Recently a number of papers dealing with this problem have been published. They can be broadly divided into two maj...

Full description

Saved in:
Bibliographic Details
Main Author: Li, Li
Other Authors: Chutatape, Opas
Format: Theses and Dissertations
Published: 2010
Subjects:
Online Access:http://hdl.handle.net/10356/38995
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
id sg-ntu-dr.10356-38995
record_format dspace
spelling sg-ntu-dr.10356-389952023-07-04T16:36:25Z Threshold selection and identification of character regions from complex scene images Li, Li Chutatape, Opas School of Electrical and Electronic Engineering DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing In this research work, a comprehensive algorithm for alphanumeric character detection in general complex scene images is proposed, and specifically applied to car licence-number plate. Recently a number of papers dealing with this problem have been published. They can be broadly divided into two major approaches, i.e., edge-based approach in which the number-plate location is first detected, and grey-level-feature-based approach in which the character candidates are directly detected. These approaches usually need to make some assumptions or to put some constraints on the images to limit their searching work and to improve the success rate. In this thesis, we study the generalisations of the second approach by directly detect the alphanumeric characters at any location in the complex scene image with minimum image constraints. Master of Engineering 2010-05-21T03:39:15Z 2010-05-21T03:39:15Z 1997 1997 Thesis http://hdl.handle.net/10356/38995 NANYANG TECHNOLOGICAL UNIVERSITY 146 p. application/pdf
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
topic DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing
spellingShingle DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing
Li, Li
Threshold selection and identification of character regions from complex scene images
description In this research work, a comprehensive algorithm for alphanumeric character detection in general complex scene images is proposed, and specifically applied to car licence-number plate. Recently a number of papers dealing with this problem have been published. They can be broadly divided into two major approaches, i.e., edge-based approach in which the number-plate location is first detected, and grey-level-feature-based approach in which the character candidates are directly detected. These approaches usually need to make some assumptions or to put some constraints on the images to limit their searching work and to improve the success rate. In this thesis, we study the generalisations of the second approach by directly detect the alphanumeric characters at any location in the complex scene image with minimum image constraints.
author2 Chutatape, Opas
author_facet Chutatape, Opas
Li, Li
format Theses and Dissertations
author Li, Li
author_sort Li, Li
title Threshold selection and identification of character regions from complex scene images
title_short Threshold selection and identification of character regions from complex scene images
title_full Threshold selection and identification of character regions from complex scene images
title_fullStr Threshold selection and identification of character regions from complex scene images
title_full_unstemmed Threshold selection and identification of character regions from complex scene images
title_sort threshold selection and identification of character regions from complex scene images
publishDate 2010
url http://hdl.handle.net/10356/38995
_version_ 1772827426239283200