Share Email Print
cover

Proceedings Paper

Natural pixel decomposition for computational tomographic reconstruction from interferometric projection: algorithms and comparison
Author(s): Don J. Cha; Soyoung Stephen Cha
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

A computational tomographic technique, termed the variable grid method (VGM), has been developed for improving interferometric reconstruction of flow fields under ill-posed data conditions of restricted scanning and incomplete projection. The technique is based on natural pixel decomposition, that is, division of a field into variable grid elements. The performances of two algorithms, that is, original and revised versions, are compared to investigate the effects of the data redundancy criteria and seed element forming schemes. Tests of the VGMs are conducted through computer simulation of experiments and reconstruction of fields with a limited view angel of 90 degree(s). The temperature fields at two horizontal sections of a thermal plume of two interacting isothermal cubes, produced by a finite numerical code, are analyzed as test fields. The computer simulation demonstrates the superiority of the revised VGM to either the conventional fixed grid method or the original VGM. Both the maximum and average reconstruction errors are reduced appreciably. The reconstruction shows substantial improvement in the regions with dense scanning by probing rays. These regions are usually of interest in engineering applications.

Paper Details

Date Published: 29 September 1995
PDF: 9 pages
Proc. SPIE 2546, Optical Techniques in Fluid, Thermal, and Combustion Flow, (29 September 1995); doi: 10.1117/12.221527
Show Author Affiliations
Don J. Cha, U.S. Dept. of Energy (United States)
Soyoung Stephen Cha, Univ. of Illinois/Chicago (United States)


Published in SPIE Proceedings Vol. 2546:
Optical Techniques in Fluid, Thermal, and Combustion Flow
Soyoung Stephen Cha; James D. Trolinger, Editor(s)

© SPIE. Terms of Use
Back to Top