Generalized constrained energy minimization approach to subpixel target detection for multispectral imagery

Subpixel detection in multispectral imagery presents a challenging problem due to relatively low spatial and spectral resolution. We present a general ized constrained energy minimization (GCEM) approach to detecting targets i n multispectral imagery at subpixel level. GCEM is a hybrid technique that combines a constrained energy minimization (CEM) method developed for hyper spectral image classification with a dimensionality expansion (DE) approach resulting from a generalized orthogonal subspace projection (GOSP) develop ed for multispectral image classification. DE enables us to generate additi onal bands from original multispectral images nonlinearly so that CEM can b e used for subpixel detection to extract targets embedded in multispectral images. CEM has been successfully applied to hyperspectral target detection and image classification. Its applicability to multispectral imagery is ye t to be investigated. A potential limitation of CEM on multispectral imager y is the effectiveness of interference elimination due to the lack of suffi cient dimensionality. DE is introduced to mitigate this problem by expandin g the original data dimensionality. Experiments show that the proposed GCEM detects targets more effectively than GOSP and CEM without dimensionality expansion. (C) 2000 Society of Photo-Optical Instrumentation Engineers.