A Method for Chainsaw Sound Detection Based on Haar-like Features

Abstract

Illegal deforestation has long been a major environmental issue causing climate changes, flooding, landslides, global warming, species extinction, etc. This paper deals with the detection of actively-cutting chainsaw for the efficient real-time monitoring of forest areas. As a solution to this problem, we propose a novel technique based on the extraction of 2D Haar wavelet coefficients from the spectrogram of environmental sounds. Considering an accuracy of 97%, our method can be used to develop a good detection system for preventing the illegal logging.