Spectral–Spatial Anomaly Detection of Hyperspectral Data Based on Improved Isolation Forest

Anomaly detection in hyperspectral image (HSI) is affected by redundant bands and the limited utilization capacity of spectral–spatial information. In this article, we propose a novel improved Isolation Forest (IIF) algorithm based on assumption that anomaly pixels are more susceptible to isolation than background pixels. The proposed IIF modified version (iForest) algorithm, which addresses poor performance iForest detecting local anomalies high-dimensional data. Furthermore, detector (SSIIFD) make full use global information, as well spectral spatial To be specific, first, apply Gabor filter extract features, then employed input relative mass forest (ReMass-iForest) obtain score. Next, original images divided into several homogeneous regions via entropy rate segmentation (ERS) preprocessed Finally, fuse scores combining them linearly predict experimental results four real datasets demonstrate outperforms other state-of-the-art methods.