Farm Biosecurity Hot Spots prediction using very high resolution Big Data Ana- lytics

Detection and early intervention of salad leaf related disease is a challenging problem for the salad-growing farmers with severe economic consequences. Sudden change in environmental condition, extreme weather condition, such as hail storm, high wind, very low temperature with humid condition or severe solar radiation can cause a series of scenarios, which can cause different diseases in salad leaves. In this paper a novel application of salad leaf disease detection has been developed using a combination of big data analytics and on field multi-dimensional sensing. We propose a cloud computing based intelligent big data analysis platform to predict farm hot spots with high probability of potential biosecurity threats and early monitoring system aiming to save the farm from significant economic damage. In this study we focus upon developing methodologies to detect the presence of a particular disease or pest infestation damage of salad leaf through machine learning based analysis of spectral profiles recorded by a Spectroradiometer in the in-situ field environment. The motivation behind this study was to demonstrate a proof of concept of the effectiveness of the hyperspectral sensing of salad leaf physiology to create ground truth data along with the climatic data (temperature, rainfall, humidity, wind speed etc.), in accordance with the very high resolution remote sensing data integration for rapid plant disease hot-spot prediction. Success from this study would be used to validate the current predictions of disease or pest likelihood on the farm and improve the model through adaptive learning accordingly.