A new composite method is proposed. This method is designed to apply to the global data set and it is composed of three consecutive criteria which are based on NDVI, brightness temperature and scan angle, respectively. This method is applied to the global data set with 4km resolution which is processed by Center for Environmental Remote Sensing (CEReS), Chiba University. The comparison with Maximum Value Composite (MVC) and maximum brightness temperature composite showed some advantages of this method. Clouds over low NDVI region as desert are eliminated due to brightness temperature criterion. Neighboring pixels are more frequently selected with same data due to scan angle criterion. It shows that the atmospheric condition and the observation geometry is similar over a wide area. The scan angle distribution is centered on nadir, therefore, effects of atmosphere and bidirectional reflectance, size of footprint and overlap between adjacent pixels are reduced.

View full abstract

Corner reflectors are suitable as ground control points when they are placed in spaces that are very large, flat, and empty. Since their reflection points are known, they can be detected and identified with relative ease in SAR images through comparisons of images before and after their installation. However, in practical applications, large flat empty spaces are not often availble near the locations necessary for a particular application. Even when these corner reflectors are placed in a flat empty lot closest to the desired location, they are not always easily identifiable in SAR images due to insufficient size or influences from surrounding areas. Using existing singular corner reflectors which were installed as ground control points for JERS-1 SAR images, we examined the minimal conditions required for empty lots to be used as installation sites, considering factors such as the number of observations, identifiability of the area surrounding the installation site, identifiability of the installation site in the images, and existence of any confounding point targets surrounding the corner reflector. We also investigated whether the same corner reflectors could be used as ground control points for the C-band ERS-2 SAR, which has an incident angle relatively similar to that of the JERS-1 SAR. We found that the most crucial condition necessary for easy detection and identification of corner reflectors in both JERS-1 SAR images and ERS-2 SAR images was that the installation site be located with adequate accuracy in the JERS-1 SAR image by visually locating the site or estimating the location from nearby distinct features. It was not necessary for the site to be very large. As long as they were larger than 10 x 10 pixels (roughly 150 m in the range direction and 50-60 m in the Azimuth direction), we could detect these sites, including any adjacent areas that appeared very dark, such as agricultural fields, rice paddies, tea plantations, water bodies, grassland and athletic fields, from single look JERS-1 SAR images.

View full abstract

Digital archives or VR museums for structures of architectural significance and objects of importance to the World's cultural heritage have recently received more attention. However, there are some issues for effective operation of digital archives or VR museums. These problems include real-time imaging, spatial data aquisition and modelling. In particular, efficient spatial data aquision tecniques in sites should be developed or invetigated.
With this objective, and for multiple applications such as human motion analysis, auto tracking, real-time positioning and so on, Hybrid Video Theodolite (HVT) system was developed by the authors which consists of the 6 parts : sensor, pan head and tilt body, imaging, recording, control and monitor. The most remarkable points of this HVT system are its ability to obtain synchronized stereo image sequences and rotation parameters while tracking a moving object. As for further additional point of this system, automated camera calibration without target can be achived.
This paper describes the HVT system, and investigates the one of applications of this system to 3D modeling of panoramic objects in architecture and archeology.

View full abstract

In order to obtain the more detail land use information for wide area, the author examined subpixel classification method, which would be applied to coarse spatial resolution data such as NOAA/AVHRR. This method employs the features that each land use shows specific temporal profile of NDVI. Estimation of winter wheat cropped area and other land uses for the case study site located in the suburban area of Beijing in China was generally coincident with the results obtained from LANDSAT/TM data. The author also proposed an index, which evaluated appropriateness of combination of temporal data used in this method. This index was examined and proved to improve efficiency of process of subpixel classification. Although the method has not yet been established from the practical point of view, it can be expected to contribute to compilation of dataset of land use for global and regional scales.

View full abstract

We investigated the combined use of NOAA AVHRR-based NDVI and spaceborne SAR-based land cover information for monitoring of‘spring flush’, which is rapid growth of grass over Japanese dairy farming region. Comparison of SAR (JERS : 17 July, 30 August, ERS : 12 June 1995) and TM (14 June 1995) -derived land cover map showed that compiled data of JERS-1 SAR and ERS-1 SAR lead to an acceptable success (overall agreement : 77%) for classification of pasture, forest, and bare-field distribution in the dairy farming landscape. Temporal NDVI changes in pastures during the spring flush season (29 May-9 June 1996) were retrieved through the simple inversion of linear mixing modeling, using land cover information addition to the representative forest and bare's NDVI values and the observed AVHRR NDVI values. Validations with the high-resolution optical data (HRV : 30 May, HRV : 4 June, TM : 7 June, HRV : 10 June 1996) -derived pasture's NDVIs show the good agreements in spatial and temporal variety of the pasture's NDVIs. These results indicate that jointly use of AVHRR and SAR is an effective strategy for monitoring of short-period temporal change in pastures over temperate regions. Further exploration of additional uses of GIS complementally data to SAR based classification and applications with coordinated multi-sensor datasets in areas of even greater spatial heterogeneity are needed.

View full abstract