A computer vision system for coffee beans classification based on computational intelligence techniques

• Computer vision system was implemented to classify coffee beans using CIELAB space.
• Transformation model (RGB to CIELAB) was performed using three Neural Networks.
• Coffee beans were classified into groups defined by the Naive Bayesian classifier.
• Classification accuracy of 100% was found using different training/validation sets.
• Higher values of L*, a* and b* were found in poor quality coffee beans.

Evaluating the color of green coffee beans is an important process in defining their quality and market price. This evaluation is normally carried out by visual inspection or using traditional instruments which have some limitations. Thus, the objective of this study was to construct a computer vision system that yields CIE (Commission Internationale d'Eclairage) L*a*b* measurements of green coffee beans and classifies them according to their color. Artificial Neural Networks (ANN) were used as the transformation model and the Bayes classifier was used to classify the coffee beans into four groups: whitish, cane green, green, and bluish-green. The neural networks models achieved a generalization error of 1.15% and the Bayesian classifier was able to classify all samples into their expected classes (100% accuracy). Therefore, the proposed system is effective in classifying variations in the color of green coffee beans and can be used to help growers classify their beans.