This research supports Hummingbird Technologies drone-enabled imagery analytics business for precision agriculture.

Pixel by pixel comparison of image data collected over multiple dates requires calibration to separate changes caused by illumination, weather condition and platform geometry from crop canopy measurements. The very high resolution datasets require normalisation of imagery from multi-temporal collections to ensure high quality products for downstream analytics. Building on robust calibration, the computer vision techniques deployed in the cloud, deliver actionable insights on crop health directly to the field in less than 24 hrs.