2. Thermal Imaging

111 The 3D visualization activity is implemented in software using the Open Inventor environment with the Silicon Graphics hardware technology. Textures can be realized through the use of three-component texture maps that include rgb values to specify characteristics in the texture coordinate (s,t) plane. In summary, this paper addresses the use of thermal imaging for 3D object characterization. The approach is currently conceptual and uses a stereo-based IR system and software simulations to generate live 3D object models with texture information. These concepts can be applied to any inspection processing system that can support thermal image acquisition. Fundamentals Temperature defines the state of energy within a given body and thermal radiation can be described as wave motion and follows the fundamental equation: λν = c (1) where λ represents the wavelength ν represents the frequency c represents the velocity Thermal energy is transmitted in the infrared wavelength between 1 µm and 100 µm. Radiation[10] is defined by Planck's blackbody radiation law given as: M e (λ,T) = 2πhc 2 /λ 5 e(hc/λkT) – 1 (W m-3) (2) where c represents the speed of light h represents Planck's constant K represents Boltzmann's constant T represents absolute temperature in Kelvin λ represents radiation wavelength in meters The radiation of a true object is calculated from the product of the blackbody radiation and the spectral emissivity [10] of the object which is denoted as ε(λ) in the equation: [M e (λ,T)] actual = ε(λ, T) M e (λ,T) (3) However, in this experimental case we suggest a method for obtaining temperature without the need of the emissivity value of the target object. The emissivity value for a given subject is not always readily available. Hence, an additional acquisition sensor can be introduced to the acquisition setup to alleviate the need for an emissivity value that is compulsory to a mono sensor-based setup. Using a dual-wavelength approach [2] as suggested by Legrand et al., in equation 12, the uncertainty of the true temperature is dependent upon the uncertainty in the emissivity variation and not the actual emissivity value. With this relationship, the accuracy of the temperature measurement is dependent upon the wavelength selection. Hence, a bandpass is defined where the two sensors operate and the emissivity ratio can be resolved to a discrete temperature value. A filter imposes restrictions on the allowable frequencies to pass through a given system. For this specific case, two …