The use of digital filtering is explored as a means of minimizingartifacts in a general class of image reconstruction approaches commonlyreferred to as direct Fourier reconstruction methods. In thepolar-to-Cartesian conversion, applications of windowing operations forthe two-dimensional interpolation-type methods are seen to reduceartifacts according to a number of error measures. For theone-dimensional case, the use of a high-resolution spline interpolationgives the lowest error measures. The use of the chip-Ztransform is proposed as a way to generate transformed projection datain a concentric square grid to eliminate part of the one-dimensionalinterpolation currently needed for the processing
展开▼