Evaluation of algorithms for tomographic reconstruction of chemical concentrations in indoor air.
Am Ind Hyg Assoc J 1994 May; 55(5):403-417
Algorithms for the tomographical reconstruction of pollutant concentrations in indoor air measured with an optical remote sensing system were evaluated and compared. The algorithms evaluated included ART, ART3, SIRT, and SART. The image quality of the four algorithm methods were compared using numerical studies. One hundred and twenty test concentration maps were reconstructed under ideal and nonideal sampling conditions. A reduced number density of sampling rays, and three levels of measurement noise were used as nonideal conditions. Image quality was evaluated using nearness, projection data distance, and exposure location analysis. A visual assessment of reconstructed maps compared to the original concentration maps. Results indicated that nearness values increased as the number of peaks in the test maps increased for all four algorithms. The peak centered exposure values in the reconstructed maps correlated well with the peaks in the test maps. The visual image quality was similar for the maps from all four algorithms. ART and ART3 were ranked as best under ideal conditions. Performance deteriorated with simulated measurement noise. SIRT and SART were sensitive at low noise levels. The author concludes that the success of the system depends on the choice of the algorithm which will depend upon the final purpose of the tomographic reconstructions.
NIOSH-Publication; NIOSH-Grant; Grants-other; Air-contamination; Air-flow; Air-quality-measurement; Indoor-air-pollution; Mathematical-models; Indoor-environmental-quality
Environmental Sciences & Engr University of North Carolina CB #7400 Chapel Hill, NC 27599-7400
American Industrial Hygiene Association Journal
University of North Carolina Chapel Hill, Chapel Hill, North Carolina