Deep Learning Techniques on Very High Resolution Images for Detecting Trees and Their Health Conditions

Very high-resolution remote sensing imagery and imagery from unmanned aerial vehicles have been acknowledged as well as valued in recent years for a variety of purposes, especially in object detection. On the other hand, deep learning (DL) has evolved as a tool for assessing pattern recognition applications and standard machine learning techniques. This study applied several DL applications in the Sultanate of Oman for detecting trees and examining their health status using very high-resolution satellite imagery data. The DL model efficiently distinguished the date palm trees from other plants and other land uses, according to our results. Aside from date palms, the model developed in this study can serve as a starting point for models to identify other types of diseased plants and trees.