Remote Sensing for Detection of Ganoderma Disease and Bagworm Infestation in Oil Palm

Two major disease and pest in oil palm are Ganoderma disease and bagworm infestation. Ganoderma disease caused by Ganoderma boninense and bagworm infestation caused by Metisa Plana has caused significant loss to oil palm industry. Therefore, early detection and control are important to reduce the losses. This paper reviewed the existing approaches, challenges and future trend of aerial remote sensing technology for Ganoderma disease and bagworm infestation in oil palm. The aerial remote sensing technology comprises of multispectral, hyperspectral camera and radar which have different platform such as satellite, aircraft and Unmanned Aerial Vehicle (UAV). The aerial multispectral and hyperspectral remote sensing analysed spectral signatures from visible and near infrared spectrum range for detection of the disease and pest attacks. Studies showed that satellite-based multispectral remote sensing only provide moderate accuracy (<70%) compared to UAV-based multispectral remote sensing (>80%) for detection of disease and pest infestation. Meanwhile, our study using UAV showed 90% of accuracy for moderate and severe Ganoderma disease detection in oil palm. Meanwhile, application of aerial hyperspectral remote sensing for Ganoderma disease showed potential for early detection of Ganoderma disease in oil palm and also can be used to detect early pest infestation in oil palm based on field spectroscopy results. Other than that, radar remote sensing has also able to differentiate healthy and Ganoderma-infected oil palm and also pest infestation by analysis of radar backscatter image of the foliar, frond and crown of oil palm. The challenges for the implementation of aerial remote sensing technology for disease and pest detection in oil palm is in tackling problems from shadows, mixed-class from single canopy and false-positive classification and also producing equipment at a lower and affordable price and also a user-friendly data analysis system that can be used by the plantations for a fast disease and pest detection works. The introduction of Artificial Intelligence (AI), Machine Deep Learning (MDL), low-cost remote sensing camera and light-weight UAV has opened the opportunity to tackle the challenges. As a conclusion, aerial remote sensing provides better and faster disease and pest infestation detection system compared to ground-based inspection. The advancement of the aerial remote sensing technology can provide more economic and efficient disease and pest infestation detection system for large oil palm plantation areas.