Anomaly Detection in Hyperspectral Images Based on an Adaptive Support Vector Method

摘要: Recently, anomaly detection (AD) has attracted considerable interest in a wide variety of hyperspectral remote sensing applications. The goal this unsupervised technique target is to identify the pixels with significantly different spectral signatures from neighboring background. Kernel methods, such as kernel-based support vector data description (SVDD) (K-SVDD), have been presented successful approach AD problems. most commonly used kernel Gaussian function. main problem using methods optimal setting sigma. In an attempt address problem, paper proposes direct and adaptive measure for K-SVDD (GK-SVDD). proposed based on geometric interpretation GK-SVDD. Experimental results are real synthetically implanted targets blind-test sets. Compared previous measures, demonstrate better performance, particularly subpixel anomalies.