Project/Area Number |
17360194
|
Research Category |
Grant-in-Aid for Scientific Research (B)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
Measurement engineering
|
Research Institution | Kochi University of Technology |
Principal Investigator |
TAKAGI Masataka (2006) Kochi University of Technology, Department of Infrastructure Systems Engineering, Professor, 工学部, 教授 (50251468)
大内 和夫 (2005) 高知工科大学, 工学部, 教授 (10289259)
|
Co-Investigator(Kenkyū-buntansha) |
HORISAWA Sakae Kochi University of Technology, Department of Environmental Systems Engineering, Associate Professor, 工学部, 助教授 (20368856)
SHIMADA Masanobu Japan Aerospace Exploration Agency, Earth Observation Research Center, Chief Scientist, 地球観測利用推進センター, 主任研究員 (90358721)
WATANABE Manabu Japan Aerospace Exploration Agency, Earth Observation Research Center, Scientist, 地球観測利用推進センター, 研究員 (10371147)
高木 方隆 高知工科大学, 工学部, 教授 (50251468)
|
Project Period (FY) |
2005 – 2006
|
Project Status |
Completed (Fiscal Year 2006)
|
Budget Amount *help |
¥10,100,000 (Direct Cost: ¥10,100,000)
Fiscal Year 2006: ¥5,000,000 (Direct Cost: ¥5,000,000)
Fiscal Year 2005: ¥5,100,000 (Direct Cost: ¥5,100,000)
|
Keywords | measurement engineering / remote sensing / forestry / imaging radar / algorithm |
Research Abstract |
The purpose of this research is to develop techniques to extract forest information from the high-resolution multi-wavelength and multi-polarization synthetic aperture radar, Pi-SAR, developed jointly by the Japan Aerospace Exploration Agency (JAXA) and the National Institute of Information and Communications Technology (NICT). The test site is the coniferous forests in Tomakomai, Hokkaido. First, regression analyses are carried out between the Pi-SAR images acquired in November 2002, and the forest data measured simultaneously on the ground and the forest data measured in August 2003. Using the conventional technique that utilizes the radar cross section (RCS), it is found that there is no significant correlation between the X-band RCS and forest parameters at all polarizations. The L-band RCS, however, is found to increase with increasing forest biomass up to approximately 40 tons/ha. These trends are similar to those already reported by several researchers. Next, because the high-re
… More
solution SAR images appear to show the structures of the forests, the relation between the image texture and forest information is sought. As a result, the image amplitudes are found to obey the K-distributed probability density function; and that strong correlation exists between the order parameter of the K-distribution in the cross-polarized images and the forest biomass. Further, it is found that the order parameter increases with increasing biomass up to around 100 tons/ha which is well beyond the saturation limit of the conventional RCS method. From the regression curve, the biomass values of unknown forests is estimated and compared with those measured on the ground in 2005-2006. The comparison yields the model accuracy of 86%. Finally, the regression model is updated using all biomass data measured on the ground. This model is considered to be effective for estimating the biomass of coniferous forests on flat ground in the entire areas of Hokkaido; and the accuracy of estimating the forest biomass can be improved to much higher levels by combining the conventional RCS technique and the texture analysis developed in this study. Less
|