659 
GLOBAL TERRAIN DATA ORGANIZATION AND COMPRESSION METHODS 
LI Yi * *, GONG Jianhua* 
State Key Laboratory of Remote Sensing Science Institute of Remote Sensing Applications, CAS 
3 Datun Road, Chaoyang District Beijing 100101, P.R. China 
Commission V, WG V/4 
KEY WORDS: Multi-resolution terrain model, Global terrain data, Improved SPIHT wavelet, Terrain simplification 
ABSTRACT: 
Terrain data is important data source in the 3D geographic visualization system. The key problem to browse global terrain data in 
network environment is to organize, transform and display data in a quick and efficient way. This paper introduces some of the key 
techniques in the global terrain visualization. The simplified global terrain data partition and organizing method is adopted. In order 
to reduce the size of datasets, the human-vision related improved SPHIT wavelet compression method is also introduced in this 
paper. The experiment shows that these technique solutions are effective to solve the problem in global terrain visualization in C/S 
mode. 
1. INTRODUCTION 
Since the digital Earth was first proposed by the USA former 
vice president Gore [1], the information technologies and 
collection of spatial data is developing very quickly. National 
Spatial Data Infrastructure (NSDI) and “digital earth” strongly 
desire the digital expression for a region, a nation, and even 
global area. 
Virtual reality (VR) and GIS are important technologies for 
digital earth building and massive spatial information 
visualizing. Recently, as advancement in computer hardware 
and computer graphics, VR and VR-GIS get their great 
development, the traditional 2D GIS is promoted to Web and 
3D GIS [2]. 
Global virtual terrain scene, which bears the weight of all kinds 
of spatial information and cultural information, is the basic 
framework of digital earth. A great number of 3D GIS 
prototype systems have been developed, which is widely 
applied in such areas as underground mine resource 
management, digital cities and digital ocean. In 1997, 
Lindstrom put forward a 3D digital earth prototype [3], which 
greatly contributed to the development of 3D virtual digital 
earth. EarthSystem developed by Keyhole Company of 
America in 2001 is so far an excellent web-based 3D 
information system [4]. ESRI Company released a multi 
resolution global 3D data visualization model ArcGlobe in 
ArcGIS 9.0. This module allows users to visualize and analyze 
large amount of 3D data. And users can browse the earth in 
different scale smoothly. Stanford Research Institute 
International (SRI) developed a distributed terrain visualization 
system called Terravision. It is open source and it can manage 
and browse large amount of terrain and image data. Google 
earth can access the expanded Keyhole aerial image database 
and provide 3D virtual earth display and map search service. 
NASA developed global visualization system WorldWind 
which can display not only high quality of imagery data of the 
Earth, but also provided the imagery about Moon, Mars and so 
on. 
Although some research has been made on browsing globe 
image and terrain data in net environment, some key problems 
such as effective organization and management of global large 
dataset, spatial data compression and simplification methods, 
data dispatch strategies and so on are still urged to be solved. 
For the terrain data is important basic geographic data. It has 
widely applications and its data amount is very huge. This 
paper mainly focuses on the optimizing strategies for the 
browsing of terrain data in networking environment, such as the 
simplified global partition and organizing method of terrain 
data, and the improved SPHIT wavelet compression method to 
reduce the size of datasets, which is relevant with human-vision. 
The rest of the paper is organized as follows: In section 2 the 
global partition of terrain data and the storage and index method 
is introduced. In section 3, we introduce the improved SPIHT 
wavelet compression method to reduce the size of datasets, 
which is relevant with human-vision. The experimental results 
are shown in section 4. Some concluding remarks are given in 
section 5. 
2. GLOBAL TERRAIN DATA PARTITION AND 
ORGANIZING METHODS 
2.1 Existing partition and organizing method 
The terrain surface distributed scope based on globe is very 
large compared to the visible scope which human eye can 
clearly observe the terrain surface. In order to make data 
dispatch more convenient and reduce the detail levels of terrain 
data. We need to do partitions to the earth surface. Nowadays 
the researchers around the world select many kinds of earth 
surface partition method according to different situations. 
* LI Yi. Ii_yi81@126.com 
* GONG Jianhua. jhgong@irsa.ac.cn