Etude des potentialités de la vision artificielle pour la reconnaissance optique des semences immatures de chicorée industrielle (Cichorium intybus L.)

Abstract

Study of the potentialities of machine vision used for optical selection of immature seeds of industrial chicory (Cichorium intybus L.). The commercial production of industrial chicory seeds (cypselae) includes the postharvest elimination of non-viable seeds by non-destructive tools. For this purpose, two machine vision methods are described for the detection of non-viable seeds: color vision and fluorescence imaging. The analysis of color images of 1,500 seeds of the 'Nausica' variety allows only the recognition of desiccated and undeveloped seeds. This is caused by the large variability of seed color, shape and texture. Fluorescence imaging is set up in order to analyze the repartition of chlorophyll fluorescence, a marker of seed maturity, on different areas of the seed (pericarp, radicle tip and pappus). A very sensitive system is needed due to the low chlorophyll content and the large amount of seeds to be sorted. A fluorescence imaging system is proposed. Its distinctive feature is the possibility to modify the spectrum of the light source (in order to optimize the sensibility of the machine vision system) and to record the evolution of fluorescence repartition with time. The system is functional and delivers images of fluorescence repartition within external cypsela tissues. It could allow to analyze the fluorescence of a large sample of seeds to correlate fluorescence features to germinability and maturity.

Keywords

• Chicory
• chlorophyll
• fluorescence
• machine vision
• maturity
• seed
• sorting