Please use this identifier to cite or link to this item: https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/2826
Title: A mobile based solution for automatic detection and classification of birds
Authors: Manamendra, M.A.S.C.
Issue Date: 30-Oct-2014
Abstract: Sri Lanka is a tropical island situated close to the southern tip of India. The bird life of Sri Lanka is very rich for its size, and about 433 species have been recorded. In addition to the many resident birds, a considerable number of migratory species winter in the country to escape their northern breeding grounds. With the invention of the new technologies like mobile applications, the environment lovers want to get the details of the different types of birds using the smart devices. To fulfil this need, the details should be available globally and an application should be there to identify the sound of the birds and predict what the bird is. This research basically looks at detecting the bird calls and predicting it. The research was conducted focusing on the birds that are local to the Sri Lanka context. The detection is highly subjective since bird call is considered to be the input signal to the algorithm. Many methods were experienced during the research time and wavelet method was selected for the detection purpose. Android was the platform for the mobile application and server application relies on matlab with .net framework. Noise effect had to be considered when the research was conducted and noise removal was one of the major tasks. For that signals were sent to the noise removal algorithm which has been developed in matlab, once that is finalized coefficient values were calculated for the noise free signals. Those values are considered in the matching process with a predefined threshold value. The application is looking for the values that can match at least 95% of a given signal. This document describes the basic the main research focus and along with many research backgrounds that have already been conducted. In addition to the basics, the details description is provided about the research process and the methods selected for this research. For this research ten birds were selected by considering many facts like similar calls and rare birds. They are Sri Lanka spur fowl, Black headed ibis, Shikra, White-Brested waterhen, Red-wattled lapwing, Common Snipe, Euransian Woodcock, Sri Lanka hanging parrot, Sri Lanka Red-Faced malkoha, Sri Lanka green-billed coucal.
URI: http://hdl.handle.net/123456789/2826
Appears in Collections:Master of Computer Science - 2014

Files in This Item:
File Description SizeFormat 
Thesis-11440383.pdf
  Restricted Access
2.08 MBAdobe PDFView/Open Request a copy


Items in UCSC Digital Library are protected by copyright, with all rights reserved, unless otherwise indicated.