Semi-Supervised Discriminant Projection for Plant Leaf Classification

Plant leaf classification is important but very difficult, because the leaf images are irregular and nonlinear. In this paper, we propose a novel semi-supervised method, called Semi-supervised discriminant projection (SSDP) dimension reduction algorithm for leaf recognition. SSDP makes full use of both labeled and unlabeled data to construct the weight incorporating the neighborhood information of data. The labeled data points are used to maximize the separability between different classes and the unlabeled data points are used to estimate the intrinsic geometric structure of the data. The experiment results on a public plant leaf database demonstrate that SSDP is effective and feasible for plant leaf recognition.