3-D Reconstruction Algorithm of Flame Based on Inversion Calculation of Thermal Radiation

This paper presents the design, method, and evaluation of a 3-D field reconstruction of a combustion flame. Three images under three different wavelengths were captured together using charge-coupled device cameras, which were specially processed by filters. Three sets of data about the value of radiation under different wavelengths were transformed from the images that were obtained according to the law of thermal radiation and derivation. For a pixel point in flame image, its irradiance can be expressed as the accumulation of the emissive power from luminous flame with different temperatures along the rays through the pixel. Then, through the calculation of the thermal radiation inversion based on the data from pixels in images, the 3-D temperature field of the flame is obtained and displayed. Experimental tests were done using candle flame to evaluate the effectiveness of the system. Ultimately, this method can restore the temperature distribution of the flame temperature field effectively. Compared with other methods, such as the two-color pyrometric method, the device of this method can get synchronous monochromatic images under three different wavelengths, the beams of which were through a beam splitter at the same time to avoid the errors of spatiotemporal matching. Else, we can reproduce the validity and efficiency of temperature field using the reconstruction algorithm that is built on improved simulated annealing algorithm.