The development of automated identification system for selected species of monogeneans using digital image processing, K-nearest neighbour and artificial neural network approaches / Elham Yousef Kalafi

Elham Yousef, Kalafi (2017) The development of automated identification system for selected species of monogeneans using digital image processing, K-nearest neighbour and artificial neural network approaches / Elham Yousef Kalafi. PhD thesis, University of Malaya.

[img] PDF (The Candidate's Agreement)
Restricted to Repository staff only

Download (1806Kb) | Request a copy
    [img]
    Preview
    PDF (Thesis PhD)
    Download (2623Kb) | Preview

      Abstract

      One of the key challenges to control diseases in fish population is achieving precise and correct identification of fish parasites. Monogenean parasites are flatworms (Platyhelminthes) that are primarily found on gills and skin of fishes. Organizing and preserving specimens of monogenean is a time consuming and difficult task. In addition, classification and identification of these specimens requires assistance of taxonomy experts. Since last two decades, improvements in developing computational tools made significant motivation to classify biological specimens` images to their correspondence species. These days, identification of biological species are easier for taxonomists and non-taxonomists due to the development of models and methods that are able to characterize species` morphology. Monogeneans have categorical homogeneous morphology, hence, pattern recognition techniques can be used to identify them. In this study, fully automated identification model for monogenean images based on the shape characters of their haptoral organs is developed. The morphological features were extracted from anchors and bars of monogeneans by adoption of digital image processing techniques. The Linear Discriminant Analysis (LDA) method was used to transform extracted feature vector to lower dimension feature vector and the transformed features were put into K-Nearest Neighbour (KNN) and Artificial Neural Network (ANN) classifiers for identification of monogenean specimens of eight species, Sinodiplectanotrema malayanus, Diplectanum jaculator,Trianchoratus pahangensis, Trianchoratus lonianchoratus, Trianchoratus malayensis, Metahaliotrema ypsilocleithru, Metahaliotrema mizellei and Metahaliotrema similis. Considerably, this is the first fully automated identification system for monogenean with the accuracy of 86.25% using KNN and 93.1% using ANN classification techniques. Images are classified based on monogenean diagnostic organs which are haptoral bars and anchors.

      Item Type: Thesis (PhD)
      Additional Information: Thesis (PhD) – Faculty of Science, University of Malaya, 2017.
      Uncontrolled Keywords: Automated identification system; Monogenean diagnostic organs; Digital image; Homogeneous morphology
      Subjects: Q Science > Q Science (General)
      Divisions: Faculty of Science
      Depositing User: Mr Mohd Safri Tahir
      Date Deposited: 19 Sep 2019 01:57
      Last Modified: 19 Sep 2019 01:57
      URI: http://studentsrepo.um.edu.my/id/eprint/7691

      Actions (For repository staff only : Login required)

      View Item