Principal

This paper presents the classification method of various paddy varieties as per the rice processing requirement. In first phase four morphological features of the individual as well as group's average features of paddy were extracted using image processing. Out of these four features only two features (minor axis and area) are providing sufficient information to classify the paddy as per the requirement of rice dryer and processing plant. In the second stage a feed forward neural network was applied to classify the extracted data. These data were classified in to large, medium and small samples. Another 10 sets of samples were tested using NN and it is found that all these samples are classified properly