Identification of Sukun (Artocarpus altilis) and Kluwih (Artocarpus camansi) Leaves using Transfer Learning

It’s critical to correctly identify Sukun (Artocarpus altilis) and Kluwih (Artocarpus camansi) leaves to avoid misunderstandings about their traditional medical use and to prevent fraud in the sale of plant seedlings. This study uses the well-known deep-learning algorithms VGG16 and Inception v3 to identify Sukun and Kluwih leaves. A varied collection of leaf photos in a range of sizes was gathered and used in the experiment. The results of the experiment demonstrate how profoundly useful deep learning is at identifying leaves. A remarkable accuracy rate of 99.52 percent was demonstrated by the VGG16 and Inception v3 models. The models’ dependability for practical, real-world applications is highlighted by this level of accuracy. Due to the excellent accuracy attained by both models, it can be concluded that they are suitable for practical applications such as preventing the improper use of leaves and preventing potential fraud in the trading of plant saplings. The work underlines the viability of adopting these models to address relevant societal concerns while showcasing the promise of deep learning approaches in botanical classification.