ENHANCED IMAGE PREPROCESSING FOR AUTOMATED MAIZE GRAIN VARIETY IDENTIFICATION

Abstract

Image identification is based on identifying the basics of geometry and shapes of the objects. The implementation of computer technology for identifying agricultural products based on their visual characteristics has been increasing popularity in the last few years. Methods for image processing based on product visual characteristics are used in a variety of fields for identification and analysis purposes. For the development of hybrid model for automated identification of maize grain varieties, several processes involved including image preprocessing, segmentation, feature extraction and identification. This paper is about the preprocessing of maize grain images after capturing images. Image preprocessing was necessary to resize images, augment images, gray-scale conversion and eliminate the noise from images in order to obtain more accurate feature information.