Photometric Analysis Techniques Study

The report describes some studies made on the application of pattern recognition techniques to the Space Object Identification SOI problem using passive photometric signature data. Both simulated and measured passive photometric data primarily, the latter on satellite objects were compared using a variety of techniques in both the time and frequency domains. Different sample spaces with dimensionalities ranging from 14 to 50 were tested, with Principal Components Analysis being used in an effort to reduce the effective number of dimensions to a small subset with minimal loss of information. The GRC Mode Determination algorithm was then applied to the sample points in the reduced subsets to compare data from the different satelliteobserver pairs in order to ascertain the extent to which this type of data processing is able to separate objects which are different and group together those which are in fact similar.