Prospek Pengembangan Metode Identifikasi Spesies Burung berdasarkan Spektrum Suara menggunakan Analisis Spektogram

Abstract

Bird species can be identified using bird’s sound. Research about sound bird identification is lacking in Indonesia. This research is initial study to classify bird based on sound and its spectrum. The research was done in Laboratory of Animal Anatomy to record bird sample sound and Laboratory of Computation Physics for data analysis. This research used five recorded sound birds, i.e. Blue-Crowned (Loriculus galgulus), Greater Green (Chloropsis sonnerati), Black-Crested (Pycnonotus melanicterus), Long-Tailed (Lanius schach), and Asian Pied Starling (Sturnus contra). Identification of bird sound spectrum has been done using Fast Fourier Transform (FFT) method. Bird sounds and spectral that plotted in time-domain and in frequency-domain, were indicated unique pattern lead to bird identification method. The result showed that five birds recorded had different sound characteristics, both in time domain and sound spectrum frequency domain. The sound spectrum identification method can be used as initial step toward bird species classification called sono-taxonomy.