MATLAB-Based Coconut Maturity Classifier using Audio and Image Processing Coconut Maturity Classifier

Abstract

A key phase in the post-harvest process of a coconut is its classification according to age or stage of development. Currently, the manual tapping method, as well as using the coconut’s exterior appearance as basis, is a customary practice among sellers to determine the fruit’s age. However, both have drawbacks when it comes to accuracy, as retailers have noted losses in income because these methods do not guarantee correctly classified coconuts. To address this concern, a coconut maturity classifier that uses MATLAB for processing the fruit’s thermal image and audio, produced by a mechanical tapper, was developed. After a series of tests, the proponents were able to develop a system that has an accuracy rate of 93.33% for audio and 60% for image. For the execution time, it takes an average of 75.37 seconds to complete the whole process. The user-friendly rating of the system is 4.62.