The HYPSO-1 mission is collecting hyperspectral data across the world. We want to develop a toolbox building on existing frameworks to generate standard and intercompatible data products. 

The HYPSO-1 mission is collecting hyperspectral data across the world. We want to develop a toolbox building on existing frameworks to generate standard and intercompatible data products, for easier comparison with other earth observation satellites. This project will be about the georeferencing processing step, which is concerned about finding geographic location of pixels and generating map products. 

Georeferencing methods can be grouped into two kinds of methods: Direct methods do not use pre-existing map products and use purely the satellite metadata to compute locations. Indirect methods find locations by comparing spatial features with already existing maps. 

The accuracy of direct georeferencing methods is highly dependent on attitude and timing knowledge. Augmenting direct methods with indirect methods could improve the quality of the processing algorithms and the resulting data products, when attitude and timing knowledge is imperfect. This project is about theorizing and implementing processing algorithms or tools, that combines both direct and indirect georeferencing methods for the Hypso-1 satellite data processing chain. 

For more information or questions, contact Dennis Langer (dennis.d.langer@ntnu.no) or Joe Garret (joseph.garret@ntnu.no)