Design of a computational application for the evaluation of the quality of export lemon

Abstract

The automation of food processes increases the quality, productivity, and economy of the companies. This study aimed at the development of a Computer Vision System (CVS) to classify lemons in real-time according to their diameter. The CVS consisted of a software, coded in Java programming language, which covers the acquisition, pre-processing, segmentation, description, recognition, and interpretation of the images. The classification criteria were in accord with the CODEX STAN 213 standard for lime-lemon. The CVS reached a 0% error in the classification of the diameter of the lemons and required 0.33 seconds as processing time to detect and classify each lemon. The CVS showed high performance in the automatic classification of lemons according to their diameter. This CVS could avoid the disadvantages of manual classification.