红外技术, 2014, 36 (5): 372, 网络出版: 2014-06-03
利用NSCT分解的高光谱异常检测
Anomaly Detection Algorithm Based on NSCT Decomposition in Hyperspectral Imagery
NSCT分解 背景抑制 高光谱异常检测 NSCT decomposition background suppressed anomaly detection in hyperspectral imagery
摘要
针对因背景复杂而导致高光谱异常检测效果降低的问题, 提出了一种基于背景抑制的异常检测方法。首先对高光谱图像进行 NSCT分解, 得到各个波段的低频图像, 然后对图像的低频信息做差, 得到背景残差图像, 对背景残差图像在做突出目标信息处理, 最后使用 KRX算法对处理后图像异常检测, 并与其他方法进行比较, 证明本文方法的有效性。
Abstract
A novel anomaly detection algorithm was proposed to improve detection result by suppressing background of imagery. Every band of the hyperspectral imagery was decomposed by NSCT transformation at first, and the background residual error data which was the minus of the hyperspectral imagery and low frequency images was processed to get prominent target. At last, the images processed was detected by KRX algorithm. The approach was better than other three methods by the ROC comparison.
孟强强, 杨桄, 卢珊, 张俭峰, 童涛. 利用NSCT分解的高光谱异常检测[J]. 红外技术, 2014, 36(5): 372. MENG Qiang-qiang, YANG Guang, LU Shan, ZHANG Jian-feng, TONG Tao. Anomaly Detection Algorithm Based on NSCT Decomposition in Hyperspectral Imagery[J]. Infrared Technology, 2014, 36(5): 372.