In this paper, the performance of an automatic visual
inspection system of rice variety is studied. During the
automatic inspection process, rice seeds are positioned, carried
through photographing sections for image inspection, and then
sorted into collection containers according to their color
feature and appearance quality. Special rice variety inspection
software is developed to prepare grading parameters, and to
tune the sorting precision and machine operation. The
parameters relating to rice appearance characteristics are used
to categorize rice seeds. In the experimental results, it
concludes that the system is accurate enough to use for
inspection of different rice seed based on its appearance
characters of the seeds