Random Collective Representation-Based Detector with Multiple Features for Hyperspectral Images

Collaborative representation-based detector (CRD), as the most representative anomaly detection method, has been widely applied in field of hyperspectral (HAD). However, sliding dual window original CRD introduces high computational complexity. Moreover, HAD models only consider a single spectral or spatial feature image (HSI), which is unhelpful for improving accuracy. To solve these problems, terms speed and accuracy, we propose novel approach, named Random Collective Representation-based Detector with Multiple Feature (RCRDMF). This method includes following steps. first extract different features include feature, Gabor extended multiattribute profile (EMAP) morphological (EMP) matrix from HSI image, enables us to improve accuracy by combining multiple features. The ensemble random collaborative representation (ERCRD) then applied, can speed. Finally, an adaptive weight approach proposed calculate each feature. Experimental results on six datasets demonstrate that superiority over