Please use this identifier to cite or link to this item:
Title: Classification of Birds using FFT and Artificial Neural Networks
Authors: Abewardana, Anuradha
Sonnadara, D.U.J.
Keywords: Neural networks
Fast Fourier Transform
Issue Date: 2012
Citation: Proceedings of the Technical Sessions, Institute of Physics Sri Lanka, 28 (2012) 100-105
Abstract: The use of feed-forward artificial neural network to categorize a selected set of Sri Lankan bird species based on their vocalization is presented. The inputs to the neural network were frequencies of bird vocalizations where each vocalization was characterized by a frequency range. Out of the selected birds, only two birds showed peak frequency values below 1,000 Hz. The Sri Lanka Scaly Thrush has the maximum average peak frequency of 7,761 Hz and the Green Billed Coucal has the lowest of 334 Hz. The preliminary results show that the artificial neural network which was trained to classify individual birds based on their frequency features had an accuracy of greater than 90% for several bird types
Appears in Collections:Department of Physics

Files in This Item:
File Description SizeFormat 
2012Paper31.pdf669.95 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.