Plant identification via multipath sparse coding

In this paper, we propose a novel plant identification method based on multipath sparse coding using SIFT features, which avoids the need of feature engineering and the reliance on botanical taxonomy. In particular, the proposed method uses five paths to model the shape and texture features of plant images, and at each path it learns the dictionaries with different sizes using hierarchical sparse coding. Finally, we apply the learned representation for plant identification using linear SVM for classification. We evaluate the proposed method on several plant datasets and find that multi-organ is more informative than single organ for botanist. Experimental results also validate that the proposed method outperforms the state-of-the-art methods.