A Case Study: Experiment of Artificial Intelligence for Thai Fruit Recognition

This research uses image processing technology with the deep learning methods, which can simulate the human vision to develop an artificial model for examining and classifying the types of Thai fruit. The experiment is conducted to classify five types of Thai fruits, namely CustardApple, Rambutan, RoseApple, Mangosteen and Durian,. These fruit samples are collected as 300 images of each type in 2D digital photographs from various views. The experimental results shown that the use of deep learning by the convolutional neural network (CNN) based on the model of EfficientNet-B0 architecture provides the excellent classification accuracy at 99%.