Examining Change Detection Methods For Hyperspectral Data

Abstract

The requirement for change detection in hyperspectral data appears to be an important and necessary tool in a variety of fields such as military, medical, geology, etc. The main objective of change detection is to observe changes of the probability distribution of a stochastic process. In this paper, we analyze two detection methods which were introduced by Schaum & Stocker: chronochrome and covariance equalization. We observe the viability of both methods for when there is misregistration between the images and determine which one is better than the other at finding anomalies.