Hyperspectral remote sensing is a flexible tool that can detect, identify and characterise surface features based on their optical reflectance properties. Our airborne platform collects scientific quality hyperspectral data faster than a UAV and at a higher spatial resolution than satellites. Our automated processing transforms raw data into exploitable outputs within hours and our bespoke state-of-the-art machine learning extracts information and delivers it in a format of your preference.
Our hyperspectral cameras can collect up to 7200ha/hr for as long as 5 hours. They collect 488 bands covering a wavelength range of 0.4μm to 2.5μm at spatial resolutions from 0.16 m. Collection is supplemented by our 80MPix survey camera which acquires true colour imagery at a resolution of a few centimetres.
Our automated workflow corrects for radiometric and atmospheric effects to transform raw data into reflectance. Imagery is subsequently geo-registered to our own orthomosaics. Advanced processing can also be supported to account for rough terrain, bidirectional reflectance distribution function (BRDF) effects, cloud and cast shadow.
As well as reflectance imagery, additional outputs, such as vegetative indices, are also generated automatically. Each image is accompanied by a quality report. The rich information in these outputs is ready for analysis.
Variable illumination and BRDF effects inhibit the generation of uniform mosaics, which is critical to the exploitation of large area collections. We have developed a statistical technique to improve the uniformity and hence value of large area collections.