Rubber Leaf Disease Detection from Low Altitude Remote Sensing Techniques

Abstract

The rubber tree is an important commodity crop in Malaysia that contributing about 20% of world natural rubber production.However,infection of several diseases will affect yield production and also the life of the plant.Recently,latex yield production in Malaysia shows decrement of 20% that caused by rubber leaf diseases since 2011.Therefore,this study aims to identify the health of rubber tree that affected by Oidium leaf disease using low altitude remote sensing images that acquired from unmanned aerial vehicle (UAV) platform. Based on remote sensing approach, the characteristics of rubber leaf disease can be identified from two groups of a spectral wavelength at visible (RGB: 0.4 μm–0.7 μm) and near infrared (NIR: 0.7 μm–2.0 μm),respectively using digital compact cameras.Both images RGB and NIR, respectively are acquired using low altitude UAV (120 m–170 m above ground level) at 10 hectares rubber plantation experimental plot in southern of Peninsular Malaysia.The images are then to be analyzed and calibrated with the handheld spectroradiometer in order to measure the spectral trends of infected leaf.As a result,this study successful investigates the new application using low-cost UAV payload digital compact camera which considered as low altitude remote sensing technique to be used in rubber leaf disease detection. Thus,NIR and red respectively are identify as the potential spectrum region from digital compact camera that suitable to be used in rubber leaf disease spectral analysis.