An application of genetic algorithms to the geometric correction of HypSEO hyperspectral data

The proposed paper describes a method to correct geometric distortions of raw images based on the optimization of a satellite geometric model. The geometric model depends on a wide set of parameters such as detector offset and orientation in the focal plane and coefficients of the polynomials fitting orbital and attitude data. The optimization of the model is achieved in two steps: at first, images obtained by different channels in the focal plane are coregistered with respect to each other. Then, the image of a reference channel is geolocated with the aid of ground control points (GCP).