APPLICATION OF AN ADAPTIVE SPECTROGRAM SIGNAL PROCESSING ALGORITHM IN DETECTION OF THE RED PALM WEEVIL

摘要

Red Palm Weevil (RPW) is the major devastating insect for date palm trees. This type of insects causes a threat to palm trees and permanent damages at late stages. Consequently, early discovery of the infestation is a key in decreasing the spread. In this paper, novel signal processing system is implemented to detect the survival RPW using its bioacoustics features. Selective set of features are extracted, including unconventional features which are Mean Instantaneous Frequency (MIF), and Instantaneous power (IP), extracted from chirplet adaptive spectrogram of the acoustic RPW signal after a background wavelet denoising process. Features are extracted and correlated to give the decision criteria of RPW existence in the received bioacoustics sound. The results imply the competence of the implemented algorithm for the chosen features.