Using real-time image processing and active thermography with artificial neural network modeling for non-destructive mango quality assessment

In Thailand, fresh mango fruit is an important export product. The country is among the largest net exporters of mangoes worldwide. In respect to its appearance and outstanding taste, the mango cultivar Nam Dokmai is the most favoured cultivar for export. During postharvest handling, manual quality control takes place which, besides the elimination of defective fruits, includes also sorting according to size and shape. The present research covers the determination of size and shape for real time mass estimation of irregular shaped mango cultivars for an automated sorting prototype. Furthermore, research was performed in the field of infrared technology to automatically detect and characterize defects in mango fruits and to integrate this application into a sorting prototype.