Classification of Lichen Species using Artificial Neural Networks

Mak, Yoke Lai (2006) Classification of Lichen Species using Artificial Neural Networks. Masters thesis, University of Malaya.

[img] PDF
Mak Yoke Lai.pdf

Download (0B)

Abstract

Artificial Neural Networks (ANN) is widely used as a classification tool in biology and medical sciences. In particular, it can be applied to x-ray image segmentation, protein structure prediction and genetics sequencing. In biological and medical sciences, there are a large number of images available to researchers. Often, these images carry with them important information. One example is the use of lichen or tree moss images. Lichen researchers frequently perform classifications of such images into their respective species and subsequently store them into a database. Lichens are kept in digital databases for taxonomy studies and referencing. Current classification methods are manually performed by the researchers. An automated image recognition system can therefore be developed to simplify this time-consuming task. In this study, we describe how an Artificial Neural Network (ANN) based template matching algorithm may be used to classify lichen species from twodimensional images. We also propose a viewfinder algorithm to extract shapes from the images in order to perform the ANN classification. Additionally, we also describe how pre-processing techniques can be used to improve the quality of the images and how to overcome problems related to variable feature-shape sizes, colour, background noise and variable placement angles of the shapes. We were able to achieve more than 90% accuracy in the tests performed. The benefits to understanding how we extracted the features/shapes from 2-D images, how we used ANN to perform classification, and how we improved classification accuracy will definitely help researchers in creating innovative solutions for digital archival systems or in general pattern recognition.

Item Type: Thesis (Masters)
Subjects: Z Bibliography. Library Science. Information Resources > Z665 Library Science. Information Science
Depositing User: MS NOOR ZAKIRA ZULRIMI
Date Deposited: 10 Jul 2013 00:37
Last Modified: 10 Jul 2013 00:37
URI: http://repository.um.edu.my/id/eprint/51

Actions (login required)

View Item View Item