<title>Application Of Maximum-Entropy Spectrum Analysis To The Measurement Of Low-Frequency Radiographic Noise Spectra</title>

Abstract

Application of maximum -entropy spectrum analysis to themeasurement of low- frequency radiographic noise spectraEugene A. Trabka and Phillip C. BunchResearch Laboratories, Eastman Kodak CompanyRochester, New York 14650AbstractThe maximum -entropy method (MEM) for computing noise spectra has been extended toaccommodate two- dimensional microdensitometric scans of uniformly exposed radiographs.Based on computer -simulated short scans, the MEM requires one -tenth the data of the con-ventional FFT method, for the same statistical stability. In particular, MEM estimatesof very low- frequency noise power have proved to be substantially more stable than FFTestimates. MEM spectra also have the advantage of being smoother than the FFT method.For measured data, we found the results to be quite sensitive to the nature of thedetrending algorithm which was applied to the raw data.IntroductionIn a recent overview of the history of spectrum estimation' the author points out thatat a meeting of the Society of Exploration Geophysicists in 1967 John Burg presented apaper that was to shake the foundations of spectrum estimation. The maximum- entropymethod (MEM) proposed in that paper has since been successfully applied in manyengineering disciplines. We propose here to evaluate its efficiency and utility to themeasurement of the Wiener spectrum of photographic image noise. In particular we areinterested in characterizing the quantum mottle arising in radiography due to the use ofintensifying screens.The type of spectrum we expect to find is shown in Figure 1. The inherent flat