A statistical detection of an anomaly/target from noisy tomographic projections
Lionel Fillatre, Igor V. Nikiforov · 2004
The problem of detecting an anomaly/target from a limited number of noisy tomographic projections is addressed from the statistical point of view. We study an unknown two-dimensional scene composed of an environment, considered as a nuisance parameter set, with a possibly hidden anomaly/target. An optimal statistical invariant test is proposed to solve such a problem.