Paper
15 October 2012 Hyperspectral target detection based on improved automatic morphological endmember extraction method
Xu-guang Sun, Jing-ju Cai, Zhi-yong Xu, Jian-lin Zhang
Author Affiliations +
Proceedings Volume 8415, 6th International Symposium on Advanced Optical Manufacturing and Testing Technologies: Large Mirrors and Telescopes; 841518 (2012) https://doi.org/10.1117/12.976015
Event: 6th International Symposium on Advanced Optical Manufacturing and Testing Technologies (AOMATT 2012), 2012, Xiamen, China
Abstract
At present, commonly endmember extraction of hyperspectral is mainly concentrated in the spectral region. Because the spatial information is not used enough, the endmember extraction is not precise which can lead to a bad result of mixed pixel decomposition and hyperspectral target detection. Actually, the distribution of endmembers in space has a certain shape and aggregation. By making use of these information we can extract more precise endmembers. Automatic morphological endmember extraction technology can make full use of abundant spectral information and spatial information. This paper based on the existed automatic morphological algorithm, presents a method in combination with maximum distance for morphological endmember extraction to solve the influence of spectral variations, which effectively extracts different classes of endmember curves. Based on the theory of orthogonal subspace projection, the authors propose an improved constrained energy minimization (CEM) algorithm, achieve better hyperspectral target detection results.
© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Xu-guang Sun, Jing-ju Cai, Zhi-yong Xu, and Jian-lin Zhang "Hyperspectral target detection based on improved automatic morphological endmember extraction method", Proc. SPIE 8415, 6th International Symposium on Advanced Optical Manufacturing and Testing Technologies: Large Mirrors and Telescopes, 841518 (15 October 2012); https://doi.org/10.1117/12.976015
Lens.org Logo
CITATIONS
Cited by 2 scholarly publications.
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Detection and tracking algorithms

Hyperspectral target detection

Algorithms

Hyperspectral imaging

Mathematical morphology

Dimension reduction

Electronics

Back to Top