Improved data processing techniques for the High Efficiency HyperSpectral Imager (HEHSI)
Saved in:
| Title: | Improved data processing techniques for the High Efficiency HyperSpectral Imager (HEHSI) |
|---|---|
| Authors: | Arabatti, Anand |
| Committee Members: | Boreman, Glenn |
| Summary: | This thesis presents improvements in data processing techniques for the High Efficiency Hyperspectral Imager (HEHSI), a novel imaging spectrometer with high signal collection ability (or throughput) and no moving parts. We examined data processing procedures, with the goal of improving signal-to-noise ratio (SNR) in the processed dataset and the ability to process larger datasets. Two dimensional interpolation was implemented to handle cross-track motion introduced in datasets due to mechanical misalignment. In order to improve the signal-to-noise ratio (SNR), we increased signal strength by oversampling and the use of a triangular weighting function. The techniques previously implemented were optimized based on data redundancy to handle larger datasets. A comparison of the results due to the processing improvements is presented with relevant metrics. Procedures to improve processing have been recommended . |
| URL: | https://stars.library.ucf.edu/rtd/743 |
| Database: | OpenDissertations |
| Abstract: | This thesis presents improvements in data processing techniques for the High Efficiency Hyperspectral Imager (HEHSI), a novel imaging spectrometer with high signal collection ability (or throughput) and no moving parts. We examined data processing procedures, with the goal of improving signal-to-noise ratio (SNR) in the processed dataset and the ability to process larger datasets. Two dimensional interpolation was implemented to handle cross-track motion introduced in datasets due to mechanical misalignment. In order to improve the signal-to-noise ratio (SNR), we increased signal strength by oversampling and the use of a triangular weighting function. The techniques previously implemented were optimized based on data redundancy to handle larger datasets. A comparison of the results due to the processing improvements is presented with relevant metrics. Procedures to improve processing have been recommended . |
|---|
