The Ratio of Double Bounce Scattering over Surface and Volume Scatterings for Vegetation-Covered Archeological Object Detection
-
- HAMKA Ibrahim Muhammad
- Universitas Sebelas Maret Chiba University
-
- PRAMONO Subuh
- Universitas Sebelas Maret
-
- ADRIYANTO Feri
- Universitas Sebelas Maret
-
- HARYUATMANTO Gregorius
- Waindo SpecTerra
-
- KASHIHARA Hisato
- Chiba University
-
- IZUMI Yuta
- Muroran Institute of Technology
-
- SRI SUMANTYO Josaphat Tetuko
- Universitas Sebelas Maret Chiba University
この論文をさがす
説明
<p>Polarimetric synthetic aperture radar (PolSAR) contains rich data that is potentially exploited to detect hidden features, e.g. archeological artifacts in the forest. In this paper, further analysis is conducted on post-polarization decomposition by exploiting the surrounding area of the archeological sites. This research proposed the double-bounce over surface scattering area (DoS) and double-bounce over volume scattering area (DoV) as novel descriptors to extract information in PolSAR images. These novel descriptors acted as filters to gain the presence of double bounce scattering from the dihedral-shaped archeological objects, distinguishing it from the surrounding surface scattering and volume scattering. Further, three variants of the surrounding area are examined, i.e. fixed area, area with homogeneity, and area with both homogeneity and connectivity. The proposed descriptors were validated using the scale model of the manmade structure or model of archeological objects under the vegetation canopy. Then the proposed descriptors are applied to the ALOS-2 PALSAR-2 image in the archaeological site in Muaro Jambi, Sumatera island, Indonesia to investigate the buried and vegetation-covered structures. Quantitatively, the scattering mechanism is analyzed on both scale model and satellite data. The proposed descriptors were evaluated for classifying: restored site, unrestored sites and vegetation in Muaro Jambi site. The information quality and its ability to classify between oriented structures and natural targets are evaluated. The information quality shows that the proposed descriptors have richer information up to 204%, especially in vegetation areas, and potentially indicate hidden features. For classification, the DoS descriptor provides a better classification of oblique buildings versus natural targets. With further research, the proposed descriptors could be employed to detect other hidden archeological artifacts in vegetation-covered areas.</p>
収録刊行物
-
- IEICE Transactions on Communications
-
IEICE Transactions on Communications E108.B (4), 348-361, 2025-04-01
一般社団法人 電子情報通信学会
- Tweet
詳細情報 詳細情報について
-
- CRID
- 1390866647408709248
-
- ISSN
- 17451345
- 09168516
-
- 本文言語コード
- en
-
- データソース種別
-
- JaLC
- Crossref
-
- 抄録ライセンスフラグ
- 使用不可