Fundamental Study on Ground-Based Sensor Integration for Spatial Data Acquisition
Mitsunori YOSHIMURA
Center for Southeast Asian Studies
Kyoto University
Japan
yosh@cseas. kyoto-u. ac. j p
Tetsuji ANAI, Hirofumi CHIKATSU
Department of Civil Engineering
Tokyo Denki University
Ryosuke SHIBASAKI
Center for Spatial Information Science
University of Tokyo
KEY WORDS: Ground-based Sensor Integration, Spatial Data, Field-level Observation
ABSTRACT
Recent several years, Photogrammetry, GPS and their related technologies have been advanced. These technologies can be
integrated under GIS. However GIS has an unsolved problem how cost-effective and accurate geo-data can generate. Here spatial
information is defined as land use distribution in spatial dimension.
According to the mentioned background, the authors developed a spatial data acquisition tool by combining ground-based
sensors as basic components. All of instruments are connected with note type PC in order to control the operations and exchange
measured data. This tool is supposed to be used mainly in field-level observations. Accordingly from the point of their volumes, weights
and what kind of electric power source is used, all of components were selected.
Our final goal is to identify the both of specific land cover condition and distribution as three-dimensional data. Three-
dimensional measurement can be introduced by close-range Photogrammetry. Furthermore obtained information is used as one ground
truth data in spatial analysis.
In order to estimate background data with sufficient accuracy for used camera calibration, Lab-level experiments were
conducted. Through its experiments, inner orientation parameters have obtained. GPS and 3D sensor could be confirmed to have enough
capabilities for providing external orientation parameters. As the future studies, the authors are planning to verify the developed system
potential as a tool for spatial data acquisition through an actual field-level observation.
1. INTRODUCTION
1.1 Why Spatial Data needed?
GIS related technologies, 1) Spatial Positioning, 2) Network
Technology and 3) Remote Sensing have been advanced since
recent years. They are becoming integrated, resulting in the
tremendous expansion and rapid growth of these markets.
However • the spatial data acquisition and its database
construction are most of bottleneck problems in GIS. If some of
them can be improved, total volume of actual research work
should decrease and as its result, most of GIS investigation can
extend to larger regions and longer time period than the previous.
On the other hand, at the most of GIS, Remote Sensing and its
related projects, all of used spatial and statistical data must be
gathered and constructed as the preliminary stage. At actual
foreign project, it is difficult to obtain existing detail maps or
some other spatial materials by us. Especially in developing
countries such as Asian or Africa regions, even if they are
available, most of them are subject to restriction by government.
Moreover times and opportunities to investigate them actually
are limited. Accordingly the authors reached to the decision that
we have to develop the spatial data acquisition tool by ourselves.
Multi-ground-based sensors that we suppose Video, GPS and
3D sensor, can be mounted on various platforms. Most common
platform is land-based vehicles and sometimes hand-carried by
individual surveyor. Accordingly vehicle or surveyor becomes a
potential data collectors.
1.2 Our Goals of Development
So far as this moment, photographs are given us good
impression with location and actual phenomena at each observed
point. However view photos can not used for measurement of the
area where some phenomena occurred.
Our developed concepts are supposed to be used in both fixed-
point and moving field observations and to identify 1) When, 2)
Where 3) What, 4) How many area for the specific land condition.
Their target area is Kuba National Park and its surrounding
regions in Sarawak, Malaysia.
The fixed-point observation is aimed to the total observation
system for forest ecosystem. Here daily forest condition is
obtained as the image by automatic camera system that put on the
forest monitoring tower crane with 80m height. Also this imagery
is used for the measurement or monitoring of the forest condition
by combining with observed other physical data such as climate