Plants are essential for human beings to survive on this planet earth. Numerous plant species exist out of which medicinal plants play an important role as they are used for a wide range of humanoid ailments. At least 80 % of individuals now use herbal supplements and medications for some part of primary healthcare, a huge increase in use over the previous 3 decades. Therefore, proper detection of medicinal plants is a challenging task. This paper hence proposes a framework that efficiently and effectively classifies medicinal plants. The proposed framework is divided into 2 stages: (i) image pre-processing and (ii) classification network. The former utilizes the Kuwahara filter and also introduces a novel Hybrid Whale Cat Optimization Model (HWCOM). The latter employs a deep-learning-based classification model to classify medicinal plant images. The proposed framework also leverages fused gray-level co-occurrence matrix (GLCM) and principal component analysis (PCA) features to automatically classify medicinal plants by using multi-feature extraction of leaf images. Further, the proposed framework is trained and tested on medicinal plant dataset. The proposed framework can classify 30 different classes of medicinal plant leaves and provides an accuracy of 99.81 % when compared with other state-of-the-art.
Astha et al. (Thu,) studied this question.