The aim of this study is to demonstrate the validation of the original polarization components of Pi-SAR data for land cover mapping using a standard method. For this purpose, the original polarization and Pauli components of the Pi-SAR L-band data are used and the results are compared. As a method for the land cover discrimination, the traditional method of statistical maximum likelihood decision rule is selected. To increase the accuracy of the classification result, different spatial thresholds based on local knowledge are determined and used for the actual classification process. Moreover, to reduce the speckle noise and increase the spatial homogeneity of different classes of objects, a speckle suppression filter is applied to the original Pi-SAR data before applying the classification decision rule. Overall, the research indicated that the original Pi-SAR polarization components can be successfully used for separation of different land cover types without taking special polarization transformations.
PSInSAR technique was applied to non-urbanized area in Japan to monitor the land deformation caused by earthquakes, volcanic activities, land slides, and so on. PSInSAR technique has been developed to monitor the land subsidence over urbanized area. In Japan, however, to monitor land deformations over non-urbanized area are often required. In this paper, after the principles of PSInSAR technique are explained, PSInSAR study for Tokai earthquake risk assessment is described. This was a joint research project for ESA. PSInSAR technique monitors the 2-dimensional distribution of land deformation as the time series. These results are useful to estimate the underground structure around Tokai area. Before PSInSAR technique is applied to real projects, the feasibility of the study should be evaluated. Several conditions and limitations for PSInSAR were studied in this paper and two methods were proposed to evaluate the PS density-feasibility of PSInSAR. The PSInSAR feasibility of the area along Riv. Kakkonda near to Mt. Iwate was evaluated by the proposed two methods. PSInSAR study was then applied to the area and showed that the feasibility study was enough adequate. PSInSAR study around the area also showed the land deformation caused by an earthquake clearly and revealed the location of the active faults near the surface.
The objective of this study was modeling of causal relationship between water quality and spectral reflectance. In this paper, linearity between water quality (chlorophyll-a and suspended solid) and spectral reflectance were firstly evaluated to choose non-linear transformation of the water quality variables. Next, the direct influences from each water quality to spectral reflectance ware evaluated by partial correlation analysis. Finally, the modeling of causal relationship between water quality and spectral reflectance was performed by using path diagram for Lake Kasumigaura with latent variables possibly corresponding to SS quality and quantity.
A supervised classifier for satellite images by the Modified Counter-Propagation(MCP) is proposed. The MCP is a neural network for the competitive learning and it is the modified version of the Counter-Propagation, whose competitive layer is replaced by the Self-Organizing map(SOM). The Landsat image data are adopted as the input data of the MCP, and the output layer consists of the pixel values, which represent categories to be classified. Our result shows that the MCP can classify the data more accurately, objectively, and stably than the SOM only.
A supervised classification method using a self-organizing map (SOM) is proposed for the classification of remote sensing data. We adopt a three-layered SOM network and counter propagation learning method for multi-spectral data. The proposed classification method is employed to identify liquefied area in Kobe (Japan) that was damaged by the 1995 Hyogoken Nanbu earthquake, using SPOT XS data. The proposed method provides a category map to visualized the SPOT XS data and offers higher classification accuracy than either the maximum likelihood or back-propagation methods (BP).