Dual color thermography is a non-contact measurement temperature technique used mainly when the emissivity of surface is unknown; it is based on ratio of monochromatic emissive power calculated by means Planck’s radiation equation and allows measuring the temperature of gray body surface objects without being assigned their emissivity and without approximations.For real surfaces, the emissivity varies with the temperature of surface as well as the wavelength and the direction of radiation. In this case, the dual color thermometry is executed by equipping the IR camera of two narrow band pass filters, so as to consider the surface emissivity of a quite constant value. This allows calculating the ratio between the radiative fluxes of the two different emission wavelengths that is almost independent to the surface emissivity.One of the crucial factor in this technique is the choice of the two narrow filter wavelengths. In fact the measurement errors depends directly on the two wavelengths and the variation of spectral emissivity related to the wavelength chosen and it also depends inversely on distance between central value of filters.In this paper, the authors have developed and validated a mathematical model of experimental setup to measure object surface temperature by means IR thermo-camera. This mathematical model was used to quantify the temperature measurement error in the dual-color technique. A novel correlation to estimate temperature measurement error was provided.