Computer-based recognition of severity of apple blue mould using RGB components

ABSTRACT: In order to estimate the severity of apple blue mould disease which controlled by chemo- biological approach, using RGB channels; the experiment was done using two antagonistic yeasts Pichia guilliermondii (A6) and Candida membranifaciens (A4) in combination with different concentration of silicon (Si) 0.1%, 0.3% and 0.5% against apple blue mould caused by Penicillium expansum Link, in the Plant Pathology Laboratory of Abourihan Campus, University of Tehran in 2010. The results indicated that combination of the yeasts with Si could control the blue mould significantly (P < 0.05). Extraction of statistical moments of RGB channels in infected area showed that Red standard deviation was the most important feature which separated the A and B groups with accuracy of 80% and 100%, respectively. Estimation of disease severity using Standard Deviation and Kurtosis of Red channel achieved the accuracy of 100% for the both groups. The results of this study emphasize the role of color in automated estimation of the severity of apple blue mould in storages