Analysis of SSME inspection imagery using AI approachesThe automated analysis of SSME injector assemblies has been investigated for the cases of LOX post surface defects and injector-baffle deterioration. Defects are isolated via 2D feature extraction from borescope and camera images; temporal-frequency transforms are then used to create a multiresolution set of feature vectors representing image contents. The potential flaws thus discriminated are then segmented and classified according to known categories. AI is applied in the form of a blackboard architecture that is controlled by a rule-based production system.
Document ID
19930032406
Acquisition Source
Legacy CDMS
Document Type
Conference Paper
Authors
Finegan, Michael K., Jr. (NASA Headquarters Washington, DC United States)
Wee, W. G. (Cincinnati Univ. OH, United States)
Date Acquired
August 15, 2013
Publication Date
January 1, 1991
Publication Information
Publication: In: Annual Health Monitoring Conference for Space Propulsion Systems, 3rd, Cincinnati, OH, Nov. 13, 14, 1991, Proceedings (A93-16401 04-20)