Leaf Recognition Based on DPCNN and BOW

Leaf classification is an interesting and important research. Current work focuses mainly on feature extraction, especially on textural feature extraction. In this case, we propose a new method of leaf recognition based on bag of words (BOW) and entropy sequence (EnS). In our method, EnS is firstly obtained by dual-output pulse-coupled neural network and then it is improved by BOW. Locality-constrained linear coding method is used for sparse coding. Then, the classification system is built where the linear support vector machine is taken as classifier. Some representative datasets and existing methods are employed to evaluate the effect of the proposed method. Finally, experimental results show that the accuracy of our proposed method is better than existing methods.