查看文章

Rambutan (Nephelium lappaceum L.) is a widely grown and favored fruit in tropical countries such as Malaysia, Indonesia, Thailand, and the Philippines. This fruit is classified into tens of different cultivars based on fruit, flesh, and tree features. In this project, five different rambutan cultivars classification models using deep learning techniques were developed based on a 1000 rambutan images dataset. Common deep learning methods for the image classification task, Convolutional Neural Network (CNN), and transfer learning method were applied to recognize each rambutan variant. Results have shown that the VGG16 pre-trained model performed best as it achieved 96% accuracy on the test dataset. This indicates the model is reliable for the rambutan classification task.