Early detection of disease infection in chilli crops using sensors

Abstract

Control of diseases is a key aspect of profitable chilli (Capsicum annuum L.) crop production, and early detection of disease incidence is therefore an important aspect of crop management. Visual crop assessment is the most commonly used approach, but it is expensive where labour costs are high and tends to be unreliable, especially at low levels of infection. Alternative cost-effective approaches for detection of diseases and pests at an early stage are therefore desirable. This trial focused on the potential of sensor technologies to detect diseases in a chilli crop earlier than is currently possible with visual assessment. Experiments were conducted to determine whether multispectral data could be used to detect disease infection at the individual leaf and whole plant level. A multispectral camera mounted on an unmanned aerial vehicle (UAV) and a hand-held NDVI sensor were used to collect weekly data on plants in a fungicide field trial. Bacterial spot incidence in all treatments was low (<20%) but was detectable using the sensors. The hand-held NDVI sensor was able to detect diseased plants between 5 and 20 days before significant disease symptoms were visually recorded. The hand-held sensor was found to be much more sensitive in detecting diseased plants than the UAV mounted sensor.