For the purpose of the best possible use of orthophotomaps formed by the digital photometry, the simple method of estimating the diffuse reflectance factor by photographic density is developed and tested. First, the model for evaluating effects of film characteristics, A/D conversion, and geographical feature distorting photographic density is constructed. Elements of the geographical affection in this model are the diffuse-reflection component of direct solar insolations and scattering solar insolations, the specular-reflection component of direct solar insolations, and the compensation for evaluating the shade produced with the irregularity in meshes, such as a crown, according to an observation angle. Next, the method of utilizing the balloon, whose picture was taken in the photo, as a color collimating mark is examined. Every slope is assumed to receive as much solar insolation as a portion of the balloon with the same normal as that slope. Then, the formula that presumes a relative diffuse reflection factor corresponding to each normal vector from the photographic density distribution of the balloon is derived on the basis of the model. The density distribution of balloons computed by this method is almost equal to the photographic density distribution of actual balloons.
A method for Sea Surface Temperature: SST estimation with thermal infrared radiometer data by means of non-linear iteration is proposed. Although it is essential that SST is estimated by solving of radiative transfer equation: RTE, it is not so easy to solve the equation because of a non linearity of the equation. In the proposed method, RTE is symplified and is solved with a iterative method such as Conjugate Gradient method, etc. Through a comparison of the Root Mean Square error of the proposed method with that of the well known Split Window method, it is found that 7.9 to 56.3% of improvement can be achieved by the proposed method for 1350 of atmospheric and ocean surface parameters for the MODTRAN 3.7. Further, the proposed method is applied to the ADEOS/OCTS data. It is also found that RMS error of the proposed method is around 0.382 [K] .
A full automatic processing system for compositing AVHRR mosaic image was developed. In the processing, each image is applied an atmospheric correction by using 6S code and geometric corrections by referring to orbit parameters and ground contol matching patterns. By using AVHRR-HRPT data received at Tokyo (Japan), Kuroshima (Japan), Ulaanbaatar (Mongolia) and Bangkok (Thailand) in 1998, 10-day composite mosaic images covering almost total Asian region have been under processing. The data set will be open to public through the satellite environmental data base in Institute of Industry Science, University of Tokyo.
Advanced procedure for presumption of underground slide surface by analysis of different-time-photogrammetry is shown by study of Hachimantai-Sumikawa landslide, at the northeastern part of Akita prefecture in Japan, occurred on May, 11th, 1997. This landslide area had no artificial objects which can be used for tracking point to find out the ground displacement, so that the photo interpretation technique based on canopy gaps in natural beech forest was newly introduced. Additionally, in the presumption of underground slide surface geometry, a new analytic procedure based on soil mechanics, i. e. the slide-force-ratio that is derived from the digital elevation model, raised the precision of presumed 3-dimentional slide surface. The result of analysis shows good coincidence with a slide surface presumed by the data of boring survey.
A limit of the RLS (Recursive Least Square) method for the parameter estimation of Kalman Filter is clarified based on the simulated time series of data from Time Variant AR (Auto-Regressive) model. Learning process is converged for the stational time series of data while the process is diverged for the non-stational time series of data. It is also found that the required time for the convergence depends on the degree of stationality of time series of data.
A hyper spectral video system (HSV) was developed to acquire the data of hyper spectral image in low cost and easy operation. The HSV composed spectrograph system and charged coupled device (CCD) analog video camera system. Spectrograph is based on combination prism and grating. The HSV can acquire high spectral resolution imagery in the range of 400nm to 950nm. This paper describes the outline of HSV and acquired images.