Very High Spatial and Spectral Resolution Remote Sensing
A Novel Integrated Data Analysis System
The Project in a Nutshell
The "Very high spatial and spectral resolution remote sensing: a novel integrated data analysis system" project, funded by the Italian Ministry of Education, University, and Research within the PRIN 2012 program aims at providing the remote sensing community with novel methods and tools for image analysis at very high resolution in either the spatial or the spectral domain.
The pervasive use of information and communication technologies (ICT), as tools to increase the effectiveness and efficiency of human activities, and an increasing attention toward the environment, security, and quality of life of citizens are among the distinctive characteristics of advanced countries.
The moderate spatial and spectral resolutions of satellite images available in the last decades substantially limited their applications. Presently, the technological scenario is dramatically changing thanks to a new generation of sensors that acquire images with very high spatial resolution (up to 30 cm) and very high spectral resolution (up to 10 nm and 2 nm for spaceborne and airborne sensors, respectively). These data opens the possibility of new challenging applications of great social impact, such as high precision damage assessment after natural disasters, precision farming, urban monitoring (up to the detail of single buildings), infrastructure analysis, surveillance of ports and airports.
However, this technological ground-breaking improvement can be transferred into operational systems only by developing a new generation of data analysis techniques.
This is the scenario addressed by this project, with focus on the development and experimental validation of methods and processing chains for the following strategic and challenging problems:
- Very high spatial resolution (VHR) image analysis, for both optical and synthetic aperture radar (SAR) imagery. The focus is on classification, change detection, segmentation, and feature extraction through multiscale approaches that model the spectral signatures and the radar backscattering of complex objects at detailed resolution.
- Hyperspectral (i.e., very high spectral resolution) image analysis. The focus is on target and change detection through the merging of geometrical shape information and spectral signatures. Critical problems are also characterization and reduction of noise.
- Data fusion techniques for multiple heterogeneous sources. Multisensor, multiresolution, multitemporal data are tackled to benefit from the complementary information conveyed by the available remote sensing systems.
The project involves 5 remote sensing, image processing, and telecommunication labs (research units) in Italian universities. The theoretical bases of the proposed algorithms stem from random process theory, pattern recognition, and remote sensing.
A software collection integrating the techniques developed by these labs within the project, is released to the international scientific community for research, benchmarking, and didactic purposes.
Project Team and Research Units
Four project reports have been produced on the conducted research, development, and validation activities:
- First report (months 1 thru 6): state of the art analysis, algorithm specifications, and planning
- Second report (months 7 thru 18): algorithm development and preliminary experimental validation
- Third report (months 19 thru 34): algorithm integration and validation of processing chains
- Final report (months 35 and 36): final project outcomes and software delivery
► Click here to register and download the reports
The following source codes, developed within the project, are made available for free for research, didactic, and benchmarking purposes:
RBkernel_SVM_MRF_GC: region-based kernel for VHR optical images and SVM-MRF-Graph cut classifier for VHR multisensor images [C++ | University of Genoa]
MR_EdgeMarkFill: multiresolution edge-mark-fill segmentation for panchromatic and multispectral images [Matlab with precompiled C++ routines | University of Naples]
despeckling1: speckle filtering by soft classification for VHR SAR intensity images [Matlab | University of Naples]
HBDT: feature extraction, feature selection, feature fusion, and classification framework for VHR multisource images [Python | University of Pavia]
destriping_MLR_based_lbl_MAIN.m: multinomial logistic regression based residual striping estimation and reduction for hyperspectral images [Matlab | University of Pisa]
HYNPE_lbl_MAIN.m: hyperspectral noise parameter estimation algorithm for random noise in hyperspectral images [Matlab | University of Pisa]
Noise_whitening_lbl_MAIN.m: noise whitening procedure for hyperspectral images [Matlab | University of Pisa]
ACD_ProcessingChain.m: anomalous change detection processing chain for hyperspectral images [Matlab | University of Pisa]
SSBTD_ProcessingChain: spectral signature based target detection processing chain for hyperspectral images [Matlab | University of Pisa]
hc2va: graphical user interface for interactive analysis of hyper/multispectral images and semisupervised identification of changes in the spectral signature [Matlab | University of Trento]
► Click here to register and download the code
Publications from the Project
International Journal Papers
- Acito N., Diani M., Corsini G., and Resta S. (2017) “Introductory view of anomalous change detection in hyperspectral images within a theoretical Gaussian framework” IEEE Aerospace and Electronic Systems Magazine, in print
- Acito N., Matteoli S., Rossi A., Diani M., and Corsini G. (2016) “Hyperspectral airborne “Viareggio 2013 trial” data collection for detection algorithm assessment” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 9(6):2365-2376 DOI: 10.1109/JSTARS.2016.2531747
- Di Martino G., Di Simone A., Iodice A., Poggi G., Riccio D., and Verdoliva L. (2016) “Scattering-based SARBM3D” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 9(6):2131-2144 DOI: 10.1109/JSTARS.2016.2543303
- Gaetano R., Masi G., Poggi G., Verdoliva L., and Scarpa G. (2015) “Marker controlled watershed based segmentation of multi-resolution remote sensing images” IEEE Transactions on Geoscience and Remote Sensing 53(6):2987-3004 DOI: 10.1109/TGRS.2014.2367129
- Gragnaniello D., Poggi G., Scarpa G., and Verdoliva L. (2016) “SAR image despeckling by soft classification” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 9(6):2118-2130 DOI: 10.1109/JSTARS.2016.2561624
- Liu S., Bruzzone L., Bovolo F., Zanetti M., and Du P. (2015) “Sequential spectral change vector analysis for iteratively discovering and detecting multiple changes in hyperspectral images” IEEE Transactions on Geoscience and Remote Sensing 53(8):4363-4378 DOI: 10.1109/TGRS.2015.2396686
- Moser G., De Giorgi A., and Serpico S. B. (2016) “Multiresolution supervised classification of panchromatic and multispectral images by Markov random fields and graph cuts” IEEE Transactions on Geoscience and Remote Sensing 54(9):5054-5070 DOI: 10.1109/TGRS.2016.2547027
- Zanetti M., Bovolo F., and Bruzzone L. (2015) “Rayleigh-Rice mixture parameter estimation via EM algorithm for change detection in multispectral images” IEEE Transactions on Image Processing 24(12):5004-5016 DOI: 10.1109/TIP.2015.2474710
- Zanotta D. C., Bruzzone L., Bovolo F., and Shimabukuro Y. E. (2015) “An adaptive semisupervised approach to the detection of user-defined recurrent changes in image time series” IEEE Transactions on Geoscience and Remote Sensing 53(7):3707-3719 DOI: 10.1109/TGRS.2014.2381645
International Conference Papers
- Acito N., Diani M., and Corsini G. (2015) “Illumination and atmospheric conditions invariant transform for object detection in hyperspectral images” Proceedings of the 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2015), Milan, Italy, July 26-31, pp. 3731-3734 DOI: 10.1109/IGARSS.2015.7326634
- De Giorgi A., Moser G., and Serpico S. B. (2014) “Contextual remote-sensing image classification through support vector machines, Markov random fields and graph cuts” Proc. of IGARSS 2014, Quebec, Canada, July 13-18, 2014, pp. 3722-3725 DOI: 10.1109/IGARSS.2014.6947292
- De Giorgi A., Moser G., and Serpico S. B. (2015) “Parameter optimization for Markov random field models for remote sensing image classification through sequential minimal optimization” Proc. of IGARSS 2015, Milan, Italy, July 26-31, 2015, pp. 2346-2349 DOI: 10.1109/IGARSS.2015.7326279
- Gragnaniello D., Poggi G., Scarpa G., and Verdoliva L. (2015) “SAR despeckling based on soft classification” Proc. of IGARSS 2015, Milan, Italy, July 26-31, pp. 2378-2381 DOI: 10.1109/IGARSS.2015.7326287
- Iannelli G. C., and Gamba P. (2016) “Hierarchical hybrid decision tree multiscale fusion for urban image classification” Proc. of IGARSS 2016, Beijing, China, July 10-15, 2016, pp. 1800-1803 DOI: 10.1109/IGARSS.2016.7729462
- Masi G., Gaetano R., Poggi G., and Scarpa G. (2015) “A ground truth design tool for multi-resolution images” Proc. of IGARSS 2015, Milan, Italy, July 26-31, 2015, pp. 4999-5002 DOI: 10.1109/IGARSS.2015.7326955
- Masi G., Gaetano R., Poggi G., and Scarpa G. (2015) “Superpixel-based segmentation of remote sensing images through correlation clustering” Proc. of IGARSS 2015, Milan, Italy, July 26-31, 2015, pp. 1028-1031 DOI: 10.1109/IGARSS.2015.7325944
- Moser G., and Serpico S. B. (2014) “Kernel-based classification in complex-valued feature spaces for polarimetric SAR data” Proc. of IGARSS 2014, Quebec, Canada, 13-18 July 2014, pp. 1257-1260 DOI: 10.1109/IGARSS.2014.6946661
Prof. Sebastiano B. Serpico
University of Genoa, DITEN Dept., Via Opera Pia 11a, 16145 Genoa, Italy
Phone: +39 010 353 2752
Fax: +39 010 353 2134
To request the software code or the reports, click here to proceed to registration.
By submitting the registration form, users acknowledge that they agree to the following terms and conditions:
- The owners of the copyright on the reports and the software are the research units based on the Universities of Genoa, Pavia, Pisa, Naples, and Trento (Italy) that were involved in the “Very high spatial and spectral resolution remote sensing” project funded by the Italian Ministry of Education, University, and Research within the PRIN 2012 program.
- Any dissemination or distribution of the reports and the software by any registered user is strictly forbidden.
- The software is provided by the research units involved in the project exclusively for research, didactic, and benchmarking purposes.
- This software is provided “as is” and any express or implied warranties, including, but not limited to, the implied warranty of fitness for a particular purpose, are disclaimed. In no event shall the research units involved in the project be liable for any direct, indirect, incidental, special, exemplary, or consequential damages (including, but not limited to, loss of use or data) however caused and on any theory of liability, whether in contract, strict liability, or tort (including negligence or otherwise) arising in any way out of the use of this software, even if advised of the possibility of such damage.
- Any scientific publication using the software shall include a section “Acknowledgement.” This section shall include the following sentence: “The authors would like to thank the PRIN 2012 ‘Very high spatial and spectral resolution remote sensing’ project funded by the Italian Ministry of Education, University, and Research and coordinated by Prof. S. B. Serpico, University of Genoa, Italy, and <NAME(S) AND AFFILIATION(S) OF SOFTWARE AUTHOR(S)> for providing the <SOFTWARE NAME> software code.”