Quality of inverse radon transform-based image reconstruction using various frequency domain filters in parallel beam transmission tomography

Inverse Radon Transform (IRT) based image reconstruction is of prime importance in various areas of science and engineering in this paper, we have used Filtered Back-Projection technique to find IRT estimate by employing various filters, namely Ram-Lak, shepp-Logan. Cosine, Hamming and Hanning filers for deblurring and emphasizing high frequency components in the projection domain. The head phantom and the lung phantom have been used for carrying out the simulation of 512 * 512 image size. Best results as reconstructed image have been obtained by using Hanning filter. Various image parameters have been used as quality measures including average error, maximum error, peak signal-to-noise ratio, Euclidean error, mean squared error, average error histogram. Variance histogram, mean squared error histogram. Standard deviation histogram, etc. The FBP technique using Hamming filter has been found as effective as FBP Hanning technique in both cases. (author)