An automatic segmentation algorithm for touching rice grains images

An automatic segmentation algorithm for touching rice grains images was proposed. Original gray image from rice inspection system was transformed into binary image and the edges of touching grains were extracted and de-noising was operated. This algorithm detected concave corner points of touching rice grains by a sector area and found corner point pairs by the shortest Euclidean distance. The touching rice grains were segmented by the line segment which connected two corner point pair. This algorithm was efficient to segment touching long- and round-shape rice grains.