Применение компьютерного зрения в сепарации семян зерновых культур

Abstract

The most outstanding agricultural problem is a cereal seeds sorting. The sorting reduces the number of impurity seeds, causing the acceleration of growth and amount of the crop. These actions lead to economy prosperity. Today there are a lot of mechanical types of sorting, but they don't provide the required product yield. The paper is concerned with using of the computer vision technology as an additional sorting stage for a better accuracy. It provides a list of sorting criteria and some experimental results. A basic sorting method and some universalized intelligent approaches for unknown cultures are given. The possibility of corrupted or tinted seeds identification is described.