DocumentCode :
692811
Title :
Improvement of background characterization for hyperspectral target detection
Author :
Ben Ma ; Qian Du
Author_Institution :
Dept. of Electr. & Comput. Eng., Mississippi State Univ., Starkville, MS, USA
fYear :
2012
fDate :
4-7 June 2012
Firstpage :
1
Lastpage :
4
Abstract :
In this paper, several background characterization methods are reviewed and evaluated together with widely used unstructured matched filters. Two modified methods of background characterization based on spectral clustering are proposed and implemented. Experiments with in-field hyperspectral dataset are conducted to show the stability and performance improvement of the proposed methods.
Keywords :
filtering theory; hyperspectral imaging; object detection; pattern clustering; background characterization method; hyperspectral target detection; in-field hyperspectral dataset; spectral clustering; unstructured matched filters; Abstracts; Detectors; Hyperspectral imaging; Matched filters; Object detection; Radio frequency; background characterization; hyperspectral target detection; matched filter;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS), 2012 4th Workshop on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4799-3405-8
Type :
conf
DOI :
10.1109/WHISPERS.2012.6874264
Filename :
6874264
Link To Document :
بازگشت