A radiometrically-accurate super-resolution approach to thermal infrared image data

ISSN： 1947-9832

Source： International Journal of Image and Data Fusion, Vol.4, Iss.1, 2013-03, pp. : 52-74

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

Super-resolution is an image processing and analysis technique used to improve the original (or native) spatial resolution of data. Super-resolution approaches have commonly sacrificed radiometric accuracy for visual appeal or vice versa. The results presented here are a significant modification and improvement of an algorithm originally applied to the thermal infrared (TIR) data from the Earth-orbiting Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument. The algorithm has been programmed in the Interactive Data Language (IDL) scripting but could be easily coded in other programming languages. The primary focus of these modifications was to adapt the algorithm for use with visible and TIR data from the Mars-orbiting Thermal Emission Imaging System (THEMIS) instrument in addition to other improvements, such as a user-defined Point Spread Function (PSF) using an alpha notation. In addition, the previous requirement for an intermediate spatial/spectral resolution dataset has been removed after determining it to be unnecessary for accurate and visually pleasing results. This super-resolution approach is now more transparent to the user, and provides data from the intermediate steps, which allows for more accurate analysis of the results. The super-resolved TIR data from both the ASTER and the THEMIS are radiometrically accurate, interpretable, reproducible and maintain an excellent qualitative appearance.