Classifying Adult Mango Pulp Weevil Activity using Support Vector Machine

Abstract

After the discovery of the existence of mango pulp weevil in Palawan the island has been under quarantine for exporting mangoes. Detection of the pest prove to be a difficult task as the pest do not leave a physical sign that a mango has been damaged by the pests. Infested mangoes are wasted as it cannot be sold due to the damages. This study serves as a base study for a non-invasive mango pulp weevil detection by using machine learning and audio feature extraction tools of MATLAB. Audio is recorded using a MEMS microphone and is placed inside a soundproof chamber to minimize the noise. The study was able to achieve a high accuracy on characterizing the adult mango pulp weevil activity by using MFCC as features extraction for identifying its activity.