Optimized Convolutional Neural Networks and Binary Chimp Feature Selection for Robust Medicinal Plant Image Classification

This study emphasizes the importance of medicinal plants as a key component of global biodiversity conservation and human health. It specifically stresses the necessity for precise identification of medicinal plant species to ensure their sustainable protection and proper application. Traditional classification techniques face difficulties due to the intricate nature of plant characteristics and the scarcity of annotated datasets. To overcome these limitations, this work introduces a deep learning–driven model for recognizing medicinal plant images using Convolutional Neural Networks (CNNs). The framework utilizes a CNN design integrating both residual and inverted residual blocks, supported by extensive data augmentation to strengthen the dataset. For feature selection, the system employs Binary Chimp Optimization in combination with serial feature fusion to enhance both accuracy and computational speed. Experimental findings indicate that the proposed method markedly surpasses conventional classification techniques in identifying medicinal flora, and it provides a strong foundation for future extensions to other plant groups. Overall, the results demonstrate how deep learning architectures can significantly advance automated plant identification when paired with botanical research.