Enhanced classification of medicinal plants using deep learning and optimized CNN architectures

Area of Science:

• Botany
• Computer Science
• Biodiversity Conservation

Background:

• Medicinal flora is vital for global biodiversity and health.
• Accurate classification of medicinal plants is essential for conservation and utilization.
• Traditional classification methods struggle with plant complexity and limited annotated datasets.

Purpose of the Study:

• To develop a deep learning framework for accurate classification of medicinal plant images.
• To address the limitations of traditional methods in classifying complex plant features.
• To improve the efficiency and accuracy of medicinal plant identification.

Main Methods:

• Utilized a deep learning framework based on convolutional neural networks (CNNs).
• Employed CNN architectures with residual and inverted residual blocks.
• Applied data augmentation techniques to enhance the dataset.
• Integrated Binary Chimp Optimization and serial feature fusion for feature selection.

Main Results:

• The proposed deep learning framework significantly outperformed conventional classification methods.
• Achieved high accuracy in classifying medicinal flora images.
• Demonstrated the effectiveness of the chosen CNN architecture and feature selection methods.

Conclusions:

• Deep learning models show great potential for automating and improving medicinal plant identification.
• The framework offers a robust solution for accurate classification of medicinal flora.
• Suggests potential for extension to the identification of other plant species.