Machine vision-based cocoa beans fermentation degree assessment
Résumé
Fermentation degree is one of the main important indicators of cocoa bean quality. Therefore, accurate estimation of fermentation degree is very important for ensuring the quality of final products. This paper presents a quantitative method for assessing the cocoa beans fermentation degree by image analysis.
In this approach, the image of cocoa beans are acquired using a camera and processed to obtain the bean’s target. Then, the target’s pixels are clustered into red, green and blue regions where each region’s pixel presents respectively a maximal value of R, G and B in RGB color space. After that, the first three color moments of each region are calculated from RGB space and used to describe the fermentation degree of the bean. Finally, multi-class support vector machine algorithm is used as classifier to discriminate cocoa beans sample into unfermented, partly fermented and well fermented categories. Experimental results show that the proposed method can be used as a fast, accurate and a reliable tool for cocoa beans inspection and their quality control.
Fichier principal
Aubin-2020-Machine vision-based cocoa beans fermentation degree assessment (1).pdf (750.4 Ko)
Télécharger le fichier
Origine | Fichiers produits par l'(les) auteur(s) |
---|