ABSTRACT

Consumption of blueberry across the world has risen mainly due to its wellknown health bene¤ts, such as low calorie and the presence of anticancer and antioxidant properties that prevent various diseases, becoming an important component of a healthy diet (Sinelli et al. 2008). Blueberries are blue little fruits from the genus Vaccinium with a short shelf life. It has been stated that under refrigeration temperatures (∼0°C), blueberries’ shelf life is about 14-20 days (Yommi and Godoy 2002, Nunes et al. 2004). Penetration of global markets and retention of the existing markets depend on the ability to deliver consistently high-quality products. Ÿe color of this fruit ranges from light blue to deep black depending on the cultivar and presence of an epicuticular

22.1 Introduction 211 22.2 Materials and Methods 212

22.2.1 Blueberry Cultivars and Storage Conditions 212 22.2.2 Image Analysis Using Computer Vision 213 22.2.3 Sensory Evaluation 213 22.2.4 Statistical Analysis 213

22.3 Results and Discussion 214 22.4 Conclusions 217 Acknowledgments 217 References 217

wax on the skin, which gives it an attractive appearance (Nunes et al. 2004). Ÿe main quality indicators of blueberry are fruit appearance (color, size, and shape), ¤rmness or texture, ¨avor (soluble solids, titratable acidity, and pH), and nutritive value (vitamins A and C and antioxidants) (Duarte et al. 2009). However, changes in color during storage can have a profound e•ect on consumer acceptability, the darkening of the color being the limiting factors for berries stored at 10°C or 15°C (Nunes et al. 2004). Computer vision (CV) is a nondestructive technology used for acquiring and analyzing digital images to obtain information of a product or to control processes in manufacturing (Brosnan and Sun 2004). It has been regarded as a valuable tool which helps to improve the automatic assessment of food quality (Pedreschi et al. 2006, Mery et al. 2010). CV has been used in the food industry for quality and color evaluation, detection of defects, grading and sorting of fruits and vegetables, bakery products, and potato chips, among other applications (Gunasekaram and Ding 1994, Leemans et al. 1998, Pedreschi et al. 2006, Mery et al. 2010). Since color measured by CV can easily be compared to that obtained from instruments, the instrumental color spaces o•er a possible way of evaluating the performance of CV systems in measuring color of objects. Commercial colorimeters measure in CIELAB space standard only over a very few square centimeters, and thus their measurements are not very representative of heterogeneous materials such as most food products (Segnini et al. 1999, Papadakis et al. 2000).