TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
NOISE-ROBUST SPATIAL PREPROCESSING PRIOR TO ENDMEMBER EXTRACTION FROM HYPERSPECTRAL DATA
1. Noise-Robust Spatial Preprocessing Prior to Endmember Extraction from Hyperspectral Data Gabriel Martín, Maciel Zortea and Antonio Plaza Hyperspectral Computing Laboratory Department of Technology of Computers and Communications University of Extremadura, Cáceres, Spain Contact e-mail: aplaza@unex.es – URL: http://www.umbc.edu/rssipl/people/aplaza
2. Talk Outline: 1. Introduction to spectral unmixing of hyperspectral data 2. Spatial preprocessing prior to endmember extraction 2.1. Spatial preprocessing (SPP) 2.2. Region-based spatial preprocessing (RBSPP) 2.3. Noise-robust spatial preprocessing (NRSPP) 3. Experimental results 3.1. Synthetic hyperspectral data 3.2. Real hyperspectral data over the Cuprite mining district, Nevada 4. Conclusions and future research lines Noise-Robust Spatial Preprocessing for Endmember Extraction IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2011), Vancouver, Canada, July 24 – 29, 2011
3.
4.
5.
6. Talk Outline: 1. Introduction to spectral unmixing of hyperspectral data 2. Spatial preprocessing prior to endmember extraction 2.1. Spatial preprocessing (SPP) 2.2. Region-based spatial preprocessing (RBSPP) 2.3. Noise-robust spatial preprocessing (NRSPP) 3. Experimental results 3.1. Synthetic hyperspectral data 3.2. Real hyperspectral data over the Cuprite mining district, Nevada 4. Conclusions and future research lines Noise-Robust Spatial Preprocessing for Endmember Extraction IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2011), Vancouver, Canada, July 24 – 29, 2011
7.
8.
9. Spatial Preprocessing Prior to Endmember Extraction 5 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2011), Vancouver, Canada, July 24 – 29, 2011 Region-Based Spatial Pre-Processing (RBSPP) Developed by Martín and Plaza ( IEEE Geosci. Remote Sens. Lett. , 2011) Estimation of the number of endmembers p Hyperspectral image with n spectral bands Several possibilities: Chang’s VD; Bioucas’ HySime; Luo and Chanussot’s eigenvalue approach
10. Estimation of the number of endmembers p Spatial Preprocessing Prior to Endmember Extraction 5 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2011), Vancouver, Canada, July 24 – 29, 2011 Region-Based Spatial Pre-Processing (RBSPP) Developed by Martín and Plaza ( IEEE Geosci. Remote Sens. Lett. , 2011) Hyperspectral image with n spectral bands Unsupervised clustering ISODATA is used to partition the original image into c clusters, where c min = p and c max =2 p
11. Estimation of the number of endmembers p Unsupervised clustering Spatial Preprocessing Prior to Endmember Extraction 5 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2011), Vancouver, Canada, July 24 – 29, 2011 Region-Based Spatial Pre-Processing (RBSPP) Developed by Martín and Plaza ( IEEE Geosci. Remote Sens. Lett. , 2011) Morphological erosion and redundant region thinning Hyperspectral image with n spectral bands Intended to remove mixed pixels at the region borders; multidimensional morphological operators are used to accomplish this task
12. Estimation of the number of endmembers p Unsupervised clustering Morphological erosion and redundant region thinning Spatial Preprocessing Prior to Endmember Extraction 5 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2011), Vancouver, Canada, July 24 – 29, 2011 Region-Based Spatial Pre-Processing (RBSPP) Developed by Martín and Plaza ( IEEE Geosci. Remote Sens. Lett. , 2011) Region selection using orthogonal projections Hyperspectral image with n spectral bands An orthogonal subspace projection approach is then applied to the mean spectra of the regions to retain a final set of p regions
13. Estimation of the number of endmembers p Unsupervised clustering Morphological erosion and redundant region thinning Region selection using orthogonal projections Automatic endmember extraction and unmixing Spatial Preprocessing Prior to Endmember Extraction 5 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2011), Vancouver, Canada, July 24 – 29, 2011 Region-Based Spatial Pre-Processing (RBSPP) Developed by Martín and Plaza ( IEEE Geosci. Remote Sens. Lett. , 2011) Preprocessing module Hyperspectral image with n spectral bands p fully cons-trained abun-dance maps ( one per endmember )
14.
15.
16.
17. Talk Outline: 1. Introduction to spectral unmixing of hyperspectral data 2. Spatial preprocessing prior to endmember extraction 2.1. Spatial preprocessing (SPP) 2.2. Region-based spatial preprocessing (RBSPP) 2.3. Noise-robust spatial preprocessing (NRSPP) 3. Experimental results 3.1. Synthetic hyperspectral data 3.2. Real hyperspectral data over the Cuprite mining district, Nevada 4. Conclusions and future research lines Noise-Robust Spatial Preprocessing for Endmember Extraction IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2011), Vancouver, Canada, July 24 – 29, 2011
18.
19.
20.
21. AVIRIS Data Over Cuprite, Nevada Experimental Results with Synthetic and Real Hyperspectral Data 12 IEEE International Geoscience and Remote Sensing Symposium (IGARSS’2011), Vancouver, Canada, July 24 – 29, 2011
22. Experiments with the AVIRIS Cuprite hyperspectral image OSP (81 seconds) AMEE (96 seconds) SSEE (320 seconds) SPP+OSP (49+81 seconds) RBSPP+OSP (78+14 seconds) NRSPP+OSP (71+12 seconds) Experimental Results with Synthetic and Real Hyperspectral Data 13 IEEE International Geoscience and Remote Sensing Symposium (IGARSS’2011), Vancouver, Canada, July 24 – 29, 2011 RMSE=0.165 RMSE=0.265 RMSE=0.101 RMSE=0.067 RMSE=0.085 RMSE=0.129 Times measured in Intel Core i7 920 CPU at 2.67 GHz with 4 GB OF RAM ( p = 22)
23.
24. IEEE J-STARS Special Issue on Hyperspectral Image and Signal Processing IEEE International Geoscience and Remote Sensing Symposium (IGARSS’09), Cape Town, South Africa, July 12 – 17, 2009 15
25. Noise-Robust Spatial Preprocessing Prior to Endmember Extraction from Hyperspectral Data Gabriel Martín, Maciel Zortea and Antonio Plaza Hyperspectral Computing Laboratory Department of Technology of Computers and Communications University of Extremadura, Cáceres, Spain Contact e-mail: aplaza@unex.es – URL: http://www.umbc.edu/rssipl/people/aplaza