Dictionary Based Temperature and Emissivity Separation Algorithm in LWIR Hyperspectral Data

This paper presents a new algorithm for Temperature and Emissivity Separation in Long Wave Infra Red hyperspectral data. The algorithm exploits the assumption that the emissivity spectra of natural and man-made materials can be well represented in a given subspace of the original data space. The basis matrix of such a subspace is obtained by means of a dictionary based estimation strategy. Results on simulated data are presented in order to discuss the performance of the algorithm with respect to several factors such as noise and the rank of the basis matrix adopted to address the emissivity subspace.