Detection of Red Palm Weevils against Rhinoceros Beetle in Coconut Plantations of India by means of Signal

Abstract

Economic damage due to pest infestation on palms could be controlled by early detection and treatment. Several tests were conducted with acoustic instrumentation and software to assess the sound pulses produced by larva’s locomotary and feeding activities. Consistent activity patterns of hidden pest could be identified based on analysis of sound using MFCC and VQ. Mel frequency coefficients were extracted and matched to available sound codebooks that have been stored. The code book which determines the lowest quantization error will belong to the one which in turn results in confirmation of detecting the particular species of interest. This project dealt with two different species whose sound activities were recorded and further analysis was done by determining mean ,standard deviation and variance for each signal taken as input. In this paper, we have compared two different species present mainly in palms. We finally conclude that the variance , standard deviation and mean calculated for the two different species were different. Matching the obtained result online against the data determined we can prove the existence of the particular species.