Enhanced Medicinal Plant Identification: Leveraging Machine Learning and Image Processing for Accurate Raw Material Recognition

Abstract

Data on medicinal plants and raw materials encompasses details about diverse botanical resources employed for their healing attributes in both traditional and contemporary medical practices. The difficulties in improving the identification of medicinal plants include fluctuations in image quality caused by elements such as lighting, background clutter, and camera angles, which affect the accuracy of algorithms, as well as the shortage of varied training data, potentially resulting in biases or inaccuracies during recognition. To address the challenges mentioned earlier, utilize customized Extended Convolutional Neural Network (CNN) techniques capable of managing variations in image quality and utilizing diverse training data to enhance the precision of medicinal plant identification. The Extended CNN approach provides better flexibility in handling varied image attributes and heightened precision in medicinal plant recognition, thanks to its specialized architecture and utilization of extensive training datasets.