New methods to correct band misalignment in satellite data

The JITTER project investigates new methods to measure and mitigate geometric errors in satellite images caused by high-frequency attitude variations. This wobble or “jitter” can contribute to reduced data quality and sensor performance since without correction such effects cause localised image distortions and misalignment between spectral bands. The effect is analogous to skewing distortions seen in the foreground of mobile phone images taken from moving vehicles.

Jitter can be a particular problem for small satellites because of their relatively low mass and less performant attitude sensors & control systems. By measuring and correcting such effects, smallsat missions are capable of generating image products fully compliant with market expectations. 

The project will result in techniques to measure and correct for jitter in satellite imagery. These methods will be tested for suitability both on the ground and in onboard processing so that AI and analytics can be fed with image data of optimal quality.