Histogram Features for Recognition Species of Birds

Abstract:

Despite extensive studies on relevant recognition of song, automatic description of the specific field of sounds of nature still needs an intensive research. This publication presents methods of recognition of different species of birds by spectral features obtained from "bird speech". For the improvement of automatic recognition a subband intensity histogram has been used. Values extracted from the histogram spectrum of sound signals form the feature vector for sound classification in an audio surveillance application. The results presented here are a continuation of the research on this subject, however, for the first time histogram has been used for the assessment of the efficiency of the objects of classification. The effectiveness of recognition has been verified by standard algorithms classification provided by WEKA system. The analysed sounds of birds come from 10 different species of birds: Corn Crake, Hawk, Blackbird, Cuckoo, Lesser Whitethroat, Chiffchaff, Eurasian Pygmy Owl, Meadow Pipit, House Sparrow and Firecrest.