Apple ripeness identification from digital images using transformers

Abstract We describe a non-destructive test of apple ripeness using digital images multiple types apples. In this paper, fruit are treated as data samples, artificial intelligence models employed to implement the classification fruits and identification maturity levels. order obtain classifications fruits, we make use deep learning conduct our experiments; evaluate results proposed models. ensure accuracy experimental results, created own dataset, obtained best by comparing Transformer model YOLO in learning, thereby attaining recognition. At same time, also combined with attention module gave fast object detection improved model.