作者: André R. Backes , Jarbas Joaci de Mesquita Sá Junior , Rosana Marta Kolb
DOI: 10.1007/978-3-319-25751-8_7
关键词: Plant taxonomy 、 Fractal dimension 、 Artificial intelligence 、 Epidermis (botany) 、 Discriminative model 、 Color texture 、 Mathematics 、 Pattern recognition 、 Texture (music) 、 Discriminant 、 Fractal
摘要: The leaves are an important plant organ and source of information for the traditional taxonomy. This study proposes a classification approach using adaxial epidermis tissue, specific cell layer that covers leaf. To accomplish this task, we apply high discriminative color texture analysis method based on Bouligand-Minkowski fractal dimension. In experimental comparison, success rate obtained by our proposed (\(96.66\%\)) was highest among all methods used, demonstrating is very suitable to extract discriminant features from epidermis. Thus, research can significantly contribute with other studies computer vision.