Development and evaluation of a strawberry sorting system using machine vision technique

Machine vision system is one of the emerging technologies for the quality assessment of agricultural products. Quality of agricultural products is a crucial factor in marketability. In this study, a device was fabricated and developed to evaluate the quality of strawberry using machine vision technology. Strawberry is a rich source of anthocyanins in the nature. Because of the beneficial effects of anthocyanins on health, especially due to its antioxidant, anticancer and anti-inflammatory activites, especial attention is paid to strawberry consumption. In the present study, the possibility of using image processing algorithms to determine the maturity of strawberry fruits based on color and its relationship with the features extracted by destructive methods (weight, antioxidants, sugar, chlorophyll and carotenoids) was evaluated. For this purpose, 12 single-channel color spaces were developed and 228 color features were extracted. To determine the most effective features, correlation analyses were performed between destructive and non-destructive properties at 5% confidence level . The results showed the color space R / (R G B) as the most desired space whose features including sweetness and mean resulted in correlation coefficients of 0.93 and 0.91, respectively with anthocyanins. In addition, the maturity of strawberry could be classified during different growth stages using machine vision with an accuracy of 90%.