Applicazione di un sistema di analisi d'immagine alla differenziazione di cultivar di fagiolo

Nowadays new automated industrial image analysis technologies, using quick capture image systems by flat scanner and software for processing and statistic calculation, are available. These methods are able to exceed the slow and subjective vision to evaluate agricultural products samples that are also composed of a high number of individuals (e.g., seeds). The use of these non-destructive methods to differentiate cultivars, ecotypes, old varieties etc. of different species is particularly suited for agro-biodiversity studies. They can substitute or support common laboratory analysis methods. Similarly to what just realised for lentil, a new image analysis application was realised for bean. It can discriminate commercial varieties and can be used in future to characterise and distinguish Italian bean ecotypes. Eight bean cultivars (Phaseolus vulgaris L.), belonging to different types, were analysed and discriminated by a specific Macro of an image analysis system. This system measures seed morphological parameters such as area, perimeter, length, width, roundness (shape factor), RGB and HLS components of colour seed coat. Macro bean proved to work quickly and accurately