Character recognition in unconstrained environments

Cowell, JR, 1990. Character recognition in unconstrained environments. PhD, Nottingham Trent University.

Preview

Text 10290167.pdf - Published version Download (7MB) | Preview

Abstract

A multi-stage algorithm is devised for the recognition of alpha-numeric characters in unconstrained environments. The images used for examining the performance of the new algorithms are digitized pictures of vehicles in a wide variety of typical environments. The alpha-numeric characters read are on the vehicle licence plates.

The recognition process begins with a new algorithm for finding the licence plate region in the image. The extraction of the individual characters is by a novel region growing technique. As a precursor to the identification of the characters a new thinning algorithm is used, which is both faster than conventional methods and yields skeletons which conform more closely to the forms which are produced by humans than by traditional thinning techniques. A syntactic approach is used for the final stage of the recognition process. The selection of primitives is novel and it is based upon nodes where the character strokes intersect. Neither the size nor the orientation of the strokes affect the recognition process. The pattern grammar developed yields the same string for a wide range of patterns representing alpha-numeric characters irrespective of their position, size and orientation.

The thesis is illustrated throughout by the detection and reading of vehicle licence plates in unconstrained environments; however, the algorithms and underlying techniques are appropriate to a wide range of applications.

Item Type:	Thesis
Creators:	Cowell, J.R.
Date:	1990
ISBN:	9781369324143
Identifiers:	Number Type PQ10290167 Other
Rights:	This copy of the thesis has been supplied on condition that anyone who consults it is understood to recognise that its copyright rests with its author and that no quotation from the thesis and no information derived from it may be published without the author's prior written consent.
Divisions:	Schools > School of Science and Technology
Record created by:	Linda Sullivan
Date Added:	10 Nov 2020 16:44
Last Modified:	11 Oct 2023 09:05
URI:	https://irep.ntu.ac.uk/id/eprint/41602

Actions (login required)

Edit View

Statistics

Views

Downloads