BASAL STEM ROT (bsr) DISEASE DETECTION AT DIFFERENT SEVERITY LEVELS OF INFECTIONS USING MACHINE LEARNING WITH VEGETATION INDICES AND THERMAL IMAGERY

The oil palm industry in Malaysia experienced substantial growth in 2021, reaching over 5.7 million ha [1]. However, G. boninense pathogen causing basal stem rot (BSR) disease has posed a severe threat to the industry.

Remote sensing, particularly through ground-based [2,3], airborne [4] and satellite platforms [5], has shown promise in efficiently detecting the BSR disease. Ground-based sensing is impractical for big plantations and has limited data coverage. Satellite images are limited since Malaysia's location at the equator makes it hard to have a cloudless sky. Hence, this study proposes a solution to the threat of BSR disease by leveraging unmanned aerial vehicles (UAVs) equipped with multispectral and thermal sensors, combined with machine learning techniques.