Image resampling on a cylindrical sector grid
A cylindrical sensor image grid with equal area pixels for representing tomographic images is described which offers computational advantages for some algebraic and stochastic reconstruction strategies. An evaluation is included for techniques for resampling from the cylindrical representation to the standard square pixel representation. The resampling techniques of nearest-neighbors, bilinear, cubic B splines, two high-resolution cubic spline, and overlap weighting are evaluated by their noise propagation, resolution recovery, noise power spectra, and visual appearance. Visually, two techniques were superior: bilinear interpolation, and overlap weighting. In a comparison of resolution recovery vs. noise, the overlap weighting provided lower noise and better resolution. Examination of noise power spectra supported this observation. With prestored values of the weighting, computation time for all the techniques is approximately equal
