Measurement of the colour index of early-season citrus fruits using computer vision

Abstract

Aspect and colour of food surface is one of the first quality parameters evaluated by consumers and it is a key factor in the acceptance of a particular product for the consumers. In the Mediterranean area, early-season citrus fruit reach acceptable internal maturity standards for marketing, while the skin of the fruit is still green. A degreening treatment is widely used as a postharvest practice to improve the external colour. This treatment depends on the initial colour of the fruit. Therefore, an application where the inspection of the colour is needed is the assessment of these citrus fruits in order to determine accurately the colour of the citrus at harvest. The classification to determine if the fruit needs to be treated for degreening is based on the citrus colour index (CCI). In this work, the potential for the in-line CCI assessment of an industrial computer vision system is studied and compared with two other devices; a characterized computer vision system and a spectrocolorimeter used as reference in the analysis of colour on food. The results obtained prove that the industrial computer vision system predicts the colour index of citrus with a good reliability (R2 = 0.975) and is effective for classification of the fruit according to its colour.