Carabao Mango Export Quality Checker Using MATLAB Image Processing

Abstract

Purpose – Mango plays an important role in the Philippine exporting business. In the peak of mango exporting in the Philippines there are some exported fresh carabao mango to other countries that are ripped, damaged, not sweet and did not pass the safety regulations of other countries. The main reasons are pest, exporting damages, and wrong classification. Most of the exporting companies in the Philippines used the manual method or the traditional method of classifying mango by barely using their hands for touching and observation. With this being said, the objective of this research is to create a device that will automatically detect the quality and classify the carabao mango whether it is for export or not.

Methodology – The proponents used the image processing methodology. MATLAB, Arduino, and image processing techniques such as image enhancing, deblurring, edge detection, and color segmentation were adapted in the process to ensure the attainment of the objectives. The device was tested using several trials to meet designed objective metrics.

Results - As a result, the device was able to achieve the following data. The speed of classifying the quality of the mango was measured with an overall value of 8.132 seconds and for accuracy the device attained an accuracy of 91.2 %.

Conclusion – With the determined results, the proponents proved that the device successfully met the defined objective metrics. An “excellent” rating results were achieved for the speed of classifying the quality of the mango, while its accuracy was rated “good” as defined on the proponent’s metrics.

Recommendation - The proponents recommend for the improvement of the project is that future researchers should adopt better coding techniques to allow simultaneous quality checking and classification for different types of mangos.