In the present grain-handling scenario, grain type and quality are identified manually by visual
inspection which is tedious and not accurate. There is need for the growth of fast, accurate and objective system
for quality determination of food grains. An automated system is introduced which is used for grain type
identification and analysis of rice quality (i.e. Basmati, Boiled and Delhi) and grade (i.e. grade 1, grade 2, and
grade3) using Probabilistic Neural Network. This paper proposes a model that uses color and geometrical
features as attributes for classification. The grading of rice sample is done according to the size of the grain
kernel and presence of impurities. A good classification accuracy is achieved using only 6 features, i.e. mean of
RGB colors and 3 geometrical features. The total success rate of type identification is 98% and total success
rate of quality analysis and grading of rice is 90% and 92% respectively.