A Novel Approach to Temperature-Emissivity Separation Using a Multiple-Window Smoothness Criteria

A blackbody is any object that absorbs all the electromagnetic radiation incident upon it, and that emits spectral radiation that is solely dependent on the object’s kinetic temperature. Real-world objects seldom exhibit this behavior, and are termed spectral greybodies. A spectral greybody does not absorb all the incident energy, it differentially absorbs this energy as a function of wavelength. The ratio of energy radiated by a greybody to the energy radiated by a blackbody at the same temperature and wavelength, is called the spectral emissivity. Unlike spectroscopy in the reflective portion of the spectrum, the temperature of the object must be known in order to calculate the spectral emissivity. This presents a problem since a measure of the energy radiated by the object can be made at each wavelength, however, this only provides n measurements to solve for n+1 unknowns, including the object’s temperature. One is left with an ill-posed problem requiring an optimization approach. An optimization referred to as temperature-emissivity separation is used to best estimate the temperature of the object. There are previously published methods for accomplishing this, however, these methods only look at a small region of the full wavelength range for their optimization. This research proposes a novel approach that looks at multiple subregions, in multiple positions, using optimization to achieve a better estimate of the object’s temperature. A comparison of the results obtained using this new approach with methods previously employed are presented.