The detailed characterization of interfaces between essentially homogeneous materials is important for ct nde of solid rocket nozzles and related components. Speed and image size limitations of current ct imaging and image evaluation hardware leave such features under-resolved. This results in unacceptable ambiguity at both the single-pixel and multiple-pixel levels. The objective of this project is to develop algorithms which utilize a priori information about the ct system and about the object being imaged to achieve super-resolved estimates of nozzle ct interface geometry. In this phase i project, adaptation of theory and algorithms developed for the manual and automatic detection of interface anomalies in solid rocket motors will be investigated. The project will include (1) an initial review of pertinent work to date, (2) an analysis of nozzle ct data measured by the advanced rocket nozzle inspection system (arnis) to ascertain its unique measurement system characteristics, (3) an examination of typical nozzle design tolerances to validate assumptions made by the proposed analysis algorithms, (4) demonstration of these algorithms using both synthetic and measured nozzle ct data, and (5) a detailed plan for phase ii followup.