Snake species identification using image processing technique / Nur Farhani Azmi

Azmi, Nur Farhani (2020) Snake species identification using image processing technique / Nur Farhani Azmi. Degree thesis, Universiti Teknologi MARA, Cawangan Melaka.

Abstract

The purpose of this project is to develop the application of classifying the snake species. The classification was conducted by using Inception-V3, a trained model of Convolutional Neural Network (CNN) by retraining the model with two (2) species of snake which are Reticulated Python from non-venomous snake species and Malayan Pit Viper from venomous species. This project was guided by using a Modified Waterfall methodology that consist of five (5) phases which are Planning, Analysis, Design, Development and Testing. This application was built using Android Studio where a retraining model process done on using Anaconda Command. The model that has been chosen can be applied to mobile application as it will be easy to be used by all users. This application has been tested with 20 images of snake. The result of the testing shows 90% accuracy rate and all the testing images were classified correctly and successfully. The perception survey also has been evaluated by giving list of questionnaires among authorize person who are directly involve with snake such as Angkatan Pertahanan Malaysia (APM) and Bomba. The questionnaire of the survey form is based on I/S Success Model. The purpose of this survey is to get authorize person perceptions toward the application where 69.60% of 15 authorize person agree that the application produce a correct result as the information quality of the application has the highest mean value. For the future, more species of snake should be added, and user will be able to save and share the result.

Metadata

Item Type: Thesis (Degree)
Creators:
Creators
Email / ID Num.
Azmi, Nur Farhani
2017732651
Contributors:
Contribution
Name
Email / ID Num.
Thesis advisor
Othman, Zainab
UNSPECIFIED
Subjects: Q Science > QA Mathematics > Instruments and machines > Electronic Computers. Computer Science > Operating systems (Computers) > Android
Q Science > QA Mathematics > Instruments and machines > Electronic Computers. Computer Science > Neural networks (Computer science)
Divisions: Universiti Teknologi MARA, Melaka > Jasin Campus > Faculty of Computer and Mathematical Sciences
Keywords: Convolutional Neural Network; Image processing technique; Application
Date: 2020
URI: https://ir.uitm.edu.my/id/eprint/31578
Edit Item
Edit Item

Download

[thumbnail of 31578.pdf] Text
31578.pdf

Download (242kB)

Digital Copy

Digital (fulltext) is available at:

Physical Copy

Physical status and holdings:
Item Status:

ID Number

31578

Indexing

Statistic

Statistic details