This paper addresses the change detection capabilities of fully polarimetric synthetic aperture radar (SAR) for the L-band frequency in comparison with single- and dual-polarization and fully polarimetric SAR data. All polarization combinations are investigated quantitatively for unsupervised change detection under different topographic characteristics. In particular, a highly urbanized area, a vegetated area, and a mixed topographic area were examined. This allows optimal selection of polarization combinations that provide the highest change detection accuracy. The unsupervised change detection method applied in this study was based on a closed-loop process. Firstly, adaptive iterative filtering was used to determine an optimal filter size such that the speckle noise was sufficiently reduced. Secondly, the log-ratio image was generated from the filtered SAR images and was modeled according to a Gaussian distribution. Thirdly, the modified Kittler-Illingworth minimum error thresholding (KI) algorithm was applied under generalized Gaussian (GG) assumption to select double threshold that discriminates the positively and negatively changed classes from the unchanged class. Experimental results reveal that the most suitable data used for the change detection was the combined cross-polarized (HV+VH) power image, because it can achieve high correct change detection rate for any topography. The selection of filter size affects the change detection accuracy, and was dependent on the topographic characteristics. In addition, the use of the combined polarized power data, which were generated after filtering the single-polarization data at each filter size, was found out to increase the change detection accuracy.-
It is effective to control the water stress of satsuma mandarin trees artificially in order to yield fruits with high sugar content. In the remote sensing, it is known that the water stress will cause the blue shift of the red edge wavelength in spectral reflectance of leaves. In the present paper, the spectral reflectance measurement device with two optical systems and angular velocity sensor, which the spectrum characteristic of tree equator part can be acquired by scanning, was developed, and its basic performance was evaluated. As the result, it was confirmed that the spatial resolution was about 20mm in the diameter, and that it has the ability to catch the spectrum reflection characteristic of the measurement object excellently. Additionally, the author thought the average characteristic of the row of the spectrum characteristic that had been acquired in the scanning to be a standard value. Next, the square mean of the rest error of an individual spectrum characteristic was evaluated, and the spectrum characteristic that had come off from the standard values most was deleted one by one. The spectrum characteristic that had remained at the end was judged to be the spectrum of green leaves. And this procedure was proven to be correct as a result of processing the spectrum characteristic of the tree equator part. And finally, the water potential and the spectrum characteristic that had been acquired using the spectral reflectance measurement device were comparing evaluated with the satsuma mandarin in the experimental field. As a result, it was confirmed that a time response to the water stress was late, and it was confirmed to be able to presume the water potential by the accuracy of SEP=0.12MPa and R2=0.50 if the spectrum characteristic was used.
This paper reports the representation method of topographic features using detail digital elevation model. In previous research, topographic features index that compounded the over-ground and under-ground openness index (Yokoyama et al., 1999) and slope gradation map was used as well. However, these methods have some problems such as, restriction of its usage by the patent and complicated calculating processes. Therefore, many users are not able to use it for topographical feature interpretation work freely and easily. The proposed method in this report, adopting Laplacian of Gaussian filtering used in the machine vision technology, is able to calculate the same effect as in the openness index, without any concern of patent restriction. As this proposed method can be used for making landslide inventory map, consideration of road network or planning for forest resources management, it is expected to be used in various fields by various users.
We developed a 3D system which estimates the damages of disaster and quantity of house debris by fusing multi fixed camera images. It is very important to estimate the quantity of debris for grasping the damages of disaster and for promoting the restoration activities. This 3D system estimates the damages of disaster by superimposing the feature points, extracted from multi camera images taken before and after the earthquake disaster, on the 3D-model. Those structure parts in the 3D model projected with many feature points are judged as damaged parts and the volume will be calculated. In our model experiments, the calculated volume of the damaged parts was reasonable when the dimension of the house model was considered.
Rokko formula is a discriminant function for risk assessment of slope failures by earthquake. There are three problems to apply the formula for real time anticipation : too narrow target range of PGA (peak ground acceleration), lack of PGA data, and too long calculation time. We developed modified Rokko formula, estimation method of PGA from JMA (Japan Meteorological Agency) seismic intensity scale, and calculation method using precomputed histogram of real-time-independent part. We applied the methods to past earthquakes using shake maps (estimated seismic intensity distribution data) by JMA, and compared the results with slope failure positions by air photo interpretation. The results were appropriate in the interpretation area except narrow steep-slope-strips, but the system sometimes anticipated higher risk at the outside of the interpretation area. The difference was mainly caused by lack of seismometer near the epicenter.