Quality evaluation of Foeniculum vulgare (Fennel) seeds using colorization

Color based classification is widely used in food industry. Using this paper we propose a non destructive, efficient and novel technique of colorization and then propose an efficient quality analysis for the spices to be exported. Prior to export, Foeniculum vulgare (Fennel) seeds are subjected to inspection for the purpose of quality control and grading. For quality grading, the fennel seeds are cleaned and screened manually hence the practice is tedious, time consuming and labor intensive. We have developed of a machine vision system for quality inspection using combined measurement analysis. The methodology involves collecting the RGB images and then converting them to binary images using histogram analysis. Some colorization procedures are involved for image enhancement to distinguish the species from the background. Then the combined measurements consisting of area, minor axis length are extracted from the ISEF applied colorized images. The quality evaluation of the specified samples is done on the basis of quality table so formed.