Apply Deep Learning techniques on Date palm trees

Abstract

The objective of this project is to implement a deep learning model for observing the state of the date palm flower, hoping to improve the production of date fruit, to detect the state of the (spathe) in images using ”Deep Learning” and ”Transfer Learning”. We gathered a dataset containing more than 19000 images for training. We then used Transfer Learning by modifying the weights of a pre-trained YOLO model. Satisfactory learning, testing, and validation results were obtained.