Detection of internal browning in ‘Scilate’ apples with a high-speed linear-rail dual-laser scanning system

Abstract

This study reports on a linear-rail, dual-laser scanning system for high-speed, non-destructive detection of internal quality of fruit. The system was used to detect internal browning of ‘Scilate’ apples. A sample of 200 ‘Scilate’ apples with four different (healthy, slight, moderate, and severe) levels of browning was investigated. The new system’s performance was compared to that of a bench-top near infrared spectroscopy (NIRS) system. Apples with moderate and severe browning were easily detected by both systems. A comprehensive binary classification was made between healthy apples and those with slight browning. The classification results showed that the dual-laser scanning system performed very well for classifying apples with slight internal browning, achieving high accuracies of around 90 % compared to that of 82 % with the NIRS method. The dual-laser system operated successfully while fruit were moving at high speeds of 1.125 m/s and was able to identify small or localised defects.