Skip to main content
SearchLoginLogin or Signup

Replenishing the sparse data cubes from the hyperspectral imager Hyperscout-H of the Hera mission

Presentation #506.01 in the session Asteroids: Near-Earth Objects (Oral Presentation)

Published onOct 23, 2023
Replenishing the sparse data cubes from the hyperspectral imager Hyperscout-H of the Hera mission

On 26 September 2022, NASA’s Double Asteroid Redirection Test (DART) mission impacted Dimorphos, the moonlet of near-earth asteroid (65803) Didymos, performing the world’s first planetary defence test. ESA’s Hera mission will be launched in October 2024 and rendezvous with the Didymos system end of 2026 or beginning of 2027. It will closely investigate the system and in particular the consequences of the DART impact. Hera carries the hyperspectral imager Hyperscout-H. Its sensor consists of 2048 x 1088 pixels which are arranged in macro pixel blocks of 5 x 5 pixels. The 25 pixels of each block are covered with filters in 25 different wavelengths from 650 to 975 nm. Therefore, each of the 2048 x 1088 pixels provides only the brightness information for one wavelength and hence the theoretical 2048 x 1088 x 25 data cube is only sparsely populated. In order to test retrieval algorithms, we have created a fully populated test data cube by folding Chandrayaan-1/SIR-2 near infrared spectra with a Rosetta/OSIRIS image. This is then reduced to the actual Hyperscout-H data content. We replenish the reduced data with different methods and compare the results to the full data cube. A simple straight forward approach is to move a 5 x 5 pixel window with one pixel steps over the whole frame and assign the obtained 25 wavelengths spectrum to the center pixel of the window. The accuracy of this method is limited by pixel to pixel variations of the spectra and even more by varying albedo and shading effects caused by varying surface inclination. The reconstructed cube explains only 69% of the variance of the original cube. In a second more comprehensive approach, we employ linear algebra methods, in particular Principal Component analysis, to fully replenish the data cube. We reconstruct full spectra from 3 x 3 pixel windows and assign these appropriately scaled to the center pixel. This increases the spatial resolution of the reconstructed spectra and takes into account albedo and shading effects. The full cube reconstructed in this way explains 99.7% of the variance of the original cube. We will further test our approach with more simulated data and real Hyperscout-H images obtained in the laboratory.

No comments here