Date of Award


Document Type


Degree Name

Master of Science


Department of Electrical and Computer Engineering


This thesis is part of a research effort to automate the task of characterizing space objects or satellites based on a sequence of images. The goal is to detect space object anomalies. Two algorithms are considered - the feature space trajectory neural network (FST NN) and hidden Markov model (HMM) classifier. The FST NN was first presented by Leonard Neiberg and David P. Casasent in 1994 as a target identification tool. Kenneth H. Fielding and Dennis W. Ruck recently applied the hidden Markov model classifier to a 3D moving light display identification problem and a target recognition problem, using time history information to improve classification results. Time sequenced images produced by a simulation program are used for developing and testing the anomaly detection algorithms. A variety of features are tested for this problem. Features are derived from the two dimensional (2D) discrete Fourier transform (DFT) with various normalization schemes applied. The FST NN is found to be more robust than the HMM classifier. Both algorithms are capable of achieving perfect classification, but when shot noise is added to the images or when the image sample spacing is increased, the FST NN continues to perform well while the HMM performance declines. A new test is presented that measures how wdl a test sequence matches other sequences in the database. The FST NN is based strictly on feature space distance; but if the order of the sequence is important, the new test is useful.

AFIT Designator


DTIC Accession Number