This study shows the methods for estimating the irradiation by using the GMS images The brightness values of clouds were forecasted as a basis study of estimatihg the irradiation. In the first place, we have developed the method for deriving drift vector of clouds from the images and forecasting brightness at certain pixels. The drift vector stands for movement direction and speed of clouds are calculated with the correlation function that calculate with GMS cloud images. This time, the brightness values forecasted from one hour later to five hours later were calculated with the data from January 3 to March 10 in 1997. The root-mean-square error, RMSE was 7.37% for an hour later estimate. The RMSE was much the same as this model errors. This result may indicate that the spatial resolution of the images is not enough to detect clouds movement for an hour, statistically. Another RMSE were 12.9% for two hours later estimate, 13.6% for three hours later estimate, 13.8% for four hours later estimate and 13.9% five hours later estimate. When the weather is not changeable, estimated values were not affected by the size of area which the correlation function was calculated in. The method of estimating the brightness methods we have developed is available for forecasting irradiation for a few hour later. If the weather changes rapidly, would calculate estimated values by size to calculate correlation coefficient adapted to the weather.
This paper describes a slope failure detector based on mathematical morphological operations. Multisensor data, LANDSAT TM, ADEOS AVNIR, SPOT HRV XS and HRV-Panchromatic, are chosen as the original images. The proposed detector consists of morphological fundamental operation such as dilation and erosion. The formula was applied to the detection of earthquake-induced slope failures in Rokko Mountain side triggered by the Hyogoken-Nambu earthquake. The results lead two major conclusions as follows; (1) The accuracy of proposed slope failure detector depends on the resolution and high-resolution data made by wavelet transform is effective for the input data. Best result of proposed method showed almost perfect for failures of over 1000m2. (2) The accuracy of detection depends on the band (wavelength) of the input data rather than the resolution.
It has been reported that the clock onboard NOAA satellite has some time errors, which lead to the geometric distortion in the track direction. The time errors have been calculated based on the PaNDA system. The results indicated that the time errors from the PaNDA were much bigger than the reported results. It seemed because the perturbation was not considered in the PaNDA system. Even though, if the time error is corrected, the geometric distortion in the track direction will decrease drastically. And the time errors will not vary rapidly in a short period. The fact of both effects on the geometric accuracy and the stableness of time errors will lead to that the geometric correction of the images with only few GCPs or without any GCPs will be able to be carried out with some accuracy.
In this paper, we propose an improvement on the hierarchy method by using variable window size method to generate TIN (Triangulated Irregular Network) from DEM (Digital Elevation Model) . We calculate the inclination on an area to choose appropriate window size depending on type of the surfaces. By dividing the whole image into smaller window size that is different depending on type of the surfaces, and evaluate the grid points inside, we can interpolate the all grid points by using a small number of important grid points. We evaluated this proposed method by calculating the processing time, and also compared the regenerated DEM from TIN to the original DEM to know the elevation's differences in RMSE for keeping the performance of the hierarchy method. As the result, the elevation's differences in RMSE become small when using this proposed method.
We studied the effectiveness of coherence information for monitoring deforestation in tropical rain forests. First, the changing patterns of SAR backscattering intensity of JERS-1 SAR data in a tropical forest region in Sumatra island were investigated and it was found that the backscatter changes are rather complicated than those estimated from conventional studies. Then the coherence images by several JERS-1 SAR interferometric pairs were presented and the relationship between coherence and backscatter changing patterns was discussed. In addition, the accuracy for identifying deforested areas was assessed for the combined use of coherence with backscattering intensity. These results proved the effectiveness of the combined use of coherence for identifying deforested areas compared with the use of intensity only.
Using a digital height data of the tropical swamp peat forest trees which were derived from aerial stereophotos taken from a cabled balloon, we estimated the height and the brest bight diameter of every tree, then, we estimated the biomass per hector above the ground in the tropical swamp peat forest. Comparing it with the data from a literacy, it differed about 3% from the biomass data estimated by Hara and suzuki (1997) . As a result of this research, we conclude that the usage of low altitude stereo aerophotographs are useful for the estimation of the biomass above the ground in the dense natural forest of which deeper part we can not easily access into and we need some technical improvements for it.
This is to report the principle of data scanning and the distinctive features of the infrared image scanner that We have recently developed. The infrared image scanner has been developed by modifying the already marketed flathead type image scanner to visualize invisible information by infrared light. This paper will report some features of the infrared image scanner, comaprered with conventional infrared video camera, and conventional image scanner. In addition, the verified results of practical application of the scanner in the field of archaeological study are also included in this report. And more, this image scanner will be effective tool to get only “Carbon” image (i.e. the image which is dawn by pencil) from the colorfull documents.