Signature extension in remote sensing

摘要

This paper considers the problem of signature extension in remote sensing. Signature extension is a process of increasing the spatial-temporal range over which a set of training statistics can be used to classify data without significant loss of recognition accuracy.Methods are developed for the selection of segments for obtaining the training data. Selection of the number of segments is treated as the problem of expansion of rectangular matrix with basis matrices. Computational algorithms based on mean minimum square estimation error are developed for the selection of best segments. Furthermore, a combinatorial algorithm for generating all possible r combinations of S in Scr steps with a single change at each step is presented.