An alternative approach to two-dimensional power spectrum estimation incorporating the Radon transform in conjunction with each of the one-dimensional periodogram, Blackman-Tukey, and Autoregressive parameter estimation algorithms is examined. The Radon transform is used to express a two-dimensional data set in terms of its projections onto a set of one-dimensional radial lines, effectively reducing the two-dimensional estimation problem to a series of one-dimensional problems. The resulting two-dimensional power spectrum estimates are compared to the known power spectra for a variety of data types. The Radon transform approach combined with autoregressive parameter estimation can provide a high-resolution power spectrum estimate, effectively surpassing the resolution limitations of the Fourier methods without the cumbersome implementations of the more direct high resolution estimation methods in two dimensions.

Library of Congress Subject Headings

Spectral theory (Mathematics)Image processing--Mathematical models; Estimation theory; Radon transforms

Publication Date


Document Type


Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)


Easton, Roger

Advisor/Committee Member

Ninkov, Zoran

Advisor/Committee Member

Rhody, Harvey


Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: TA1637 .W54 1996


RIT – Main Campus