Nonlinear Cook distance for Anomalous Change Detection

TL;DR

This paper introduces a nonlinear extension of Cook distance using random Fourier features to improve anomalous change detection in remote sensing images, demonstrating effective empirical results across various multispectral datasets.

Contribution

It proposes a novel nonlinear Cook distance measure for anomalous change detection, enhancing impact assessment in remote sensing image analysis.

Findings

Effective detection of anomalous changes in multispectral images

Improved ROC performance over baseline methods

Visual and quantitative validation of the proposed method

Abstract

In this work we propose a method to find anomalous changes in remote sensing images based on the chronochrome approach. A regressor between images is used to discover the most {\em influential points} in the observed data. Typically, the pixels with largest residuals are decided to be anomalous changes. In order to find the anomalous pixels we consider the Cook distance and propose its nonlinear extension using random Fourier features as an efficient nonlinear measure of impact. Good empirical performance is shown over different multispectral images both visually and quantitatively evaluated with ROC curves.