Feasibility of a simplified handheld optical system for blueberries ripeness field evaluation

During last years, small fruits farming and marketing have increased widely in Italy in response to consumer's interest in bioactive and health protecting compounds fruits rich. At present, the ripening stage and harvest time is normally estimated by a visual analysis based on grower's experience. Farm growers could be supported in their decisions by simple and portable devices, used to directly assess the berries ripeness in the field and accordingly to plan the best harvest time. The aim of this work was to study the feasibility of a simplified handheld and inexpensive optical device, based on measurements and processing of diffuse reflectance at a few wavelengths appropriately selected. This study is focused on selecting the most significant wavelengths able to identify in the field the blueberries ready to be harvested. To this aim Vis/NIR spectra in the range 445-970 nm were acquired for Vaccinium corymbosum ('Brigitta' cultivar) during two different growing seasons (2005 and 2006), harvested in Valtellina area (Lombardy, Italy). Spectra measurements were taken in the field on individual berries along their equator region. Fruits samplings were performed weekly, and picked fruits were divided in four ripeness classes according to external colour. The PCA of 634 fruits spectra highlighted two principal spectral bands (around 680 nm and 740 nm) in which differences among fully (class IV) and not completely ripe (classes IIII) samples spectra are maximized. Reflectance values at these wavelengths were used to obtain spectral ratios normalised to 850 nm reflectance. Based on these relations, a simple fruits classification algorithm was proposed. Berries ripeness grading based on the proposed index showed a high ability in discriminating fully ripe fruits from partially ripe fruits and unripe fruits. More then the 93% of samples were correctly classified in validation phase. The results of this study demonstrate the feasibility of a simplified and low-cost handheld device, based on the use of only three wavelengths, able to quickly estimate blueberry ripeness in the field, with special reference to the last and most sensitive stages of ripening process.