Researchers at Sultan Qaboos University (SQU) have developed an AI-powered method to identify frankincense trees in Dhofar with 93.7% accuracy, offering a new way to monitor one of Omanβs most important natural and cultural resources.
The study combines field spectroscopy, high-resolution WorldView-3 satellite imagery and deep-learning models to identify Boswellia sacra trees in Wadi Dawkah, part of the UNESCO-listed Land of Frankincense.
Among the tested AI detection models, SSD performed best, achieving 83% precision and 93% recall across 701 validated tree detections.
The technology could help researchers and conservation authorities monitor tree distribution, detect changes over time and reduce the need for extensive ground surveys across difficult terrain. It could also support efforts to address pressures such as grazing, overtapping and climate stress.
The research demonstrates how Oman is applying AI, satellite technology and environmental science to protect its natural heritage.
