摘要：Corn kernels can have their nutritional and economic value hindered by factors such as breakage and rot on the grain. The grain classification, in general, is made by visual inspection of licensed professionals, and is a tedious and tiresome process. Besides, the possibility of error is large, due to visual human subjectivity. In an attempt to reduce this subjectivity by using image processing techniques, this paper presents a computational method for acquiring, preprocessing, and segmentation of corn kernels of digital images. It is expected that, with the proposed program, researchers can have either their focus or their efforts on feature extraction, recognition and interpretation of grains under analysis without spend time with steps related to image manipulation issues.
