Classification of Oryza Sativa using Digital Image Analysis

Abstract

Rice grains with high percentage of chalk are prone to breakage and making them undesirable that affects their market value. Moreover, in relation to health issues; people who consume rice with high chalkiness value are prone to type two diabetes since chalk I the rice grain contains a lot of starch. This research describes an algorithm for analyzing the chalkiness percentage of different types of white rice variety using different image processing techniques. The rice grains will be put in the black background and scanned with 1200 dots per inch. The scanned image will be converted into a grayscale image, the next step is binarization of image using thresholding, the next phase is morphological transformation; opening to remove the islands and unnecessary parts. The result image for both the whole rice area and chalk area will be a binary image in which all white pixels present on the whole rice or chalk area that are desired pixels and are counted to measure the chalkiness. Different types and classes of rice grain varieties are considered in the study for getting the chalkiness level. The results about accuracy of the system are presented in details by different tables.