Key points are not available for this paper at this time.
Detecting and identifying targets or objects that are present inhyperspectral ground images are of great interest. Applicationsinclude land and environmental monitoring, mining, military, civilsearch-and-rescue operations, and so on. We propose and analyze anextremely simple and efficient idea for template matching based onl₁ minimization. The designed algorithm can be applied inhyperspectral classification and target detection. Synthetic imagedata and real hyperspectral image (HSI) data are used to assess theperformance, with comparisons to other approaches, e. g. spectralangle map (SAM), adaptive coherence estimator (ACE), generalized-likelihood ratio test (GLRT) and matched filter. Wedemonstrate that this algorithm achieves excellent results with bothhigh speed and accuracy by using Bregman iteration.
Guo et al. (Sat,) studied this question.