Identifying the population of animals through pitch, formant, short time energy-A sound analysis

For a long time humans have employed animal sounds to recognize and find them. This paper proposes an automatic sound classification system based on the features extracted from the animal sounds. This paper explores the use of sound analysis algorithms to extract distinct features of different animals. Pitch, formant and short time energy are the features extracted. Support Vector Machine (SVM), a recent classifier is used for classification of animal sounds. It also deals with the comparative performance analysis of pitch by means of Auto Correlation Function (ACF) and Average Magnitude Difference Function (AMDF). More specifically the objective of this paper is to classify animal sounds and to identify the location of animals.