| Full text | |
| Author(s): |
Liesenberg, Veraldo
;
IEEE
Total Authors: 2
|
| Document type: | Journal article |
| Source: | 2022 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS 2022); v. N/A, p. 4-pg., 2022-01-01. |
| Abstract | |
Multi-frequency SAR scenes were evaluated for classification purposes of successional forest stages in Eastern Amazon. Advanced Land-Observing Satellite (ALOS) Phased Array L-band Synthetic Aperture Radar (PALSAR) data, RADARSAT-2 (C-Band), and TERRASAR-X (X-Band) data were used independently and combined in a dual SAR system approach. An image of the Advanced Land Imager (ALI) onboard the EO-1 was used as an additional ground reference and input dataset in the classification scheme. The backscattering intensity, polarimetric features, interferometric coherence, and texture attributes were considered as input data into the random forest (RF) classification algorithm. L-band data alone performed better than C- and X- bands alone but overall accuracies remained low. The addition of polarimetric features and texture at any frequency increase classification accuracies when compared to backscattering data alone. The best combination of two SAR systems were found when considering L-band with either C- or X- band. L-band also performed best when combined with optical data indicating a high potential to provide spatial information about successional forest stages. (AU) | |
| FAPESP's process: | 13/05081-9 - Characterizing secondary forest stages in different tropical rain forest environments under the recent REDD+ protocols: an attempt with multiple sources of remote sensing data |
| Grantee: | Veraldo Liesenberg |
| Support Opportunities: | Scholarships in Brazil - Post-Doctoral |