1235 
OBJECT IDENTIFICATION BASED ON MULTI-REMOTE SENSING IMAGES IN THE 
LACKING AREA OF 1:50 000 TOPOGRAPHIC MAP OF TIBETAN PLATEAU 
Zhigang Hong a ' * 
a Chinese Academy of Surveying and Mapping, Beijing 100039, China - hongzg@casm.ac.cn 
KEY WORDS: SPOT Image, 1:50000 Topographic Map Schema, Ground Objects Interpretation 
ABSTRACT: 
The lacking of 1:50000 topographic maps in western China is inconvenient to the national economic development. The rapid 
development of space technology, computer technology and information technology provides technological support for the remote 
sensing application. In addition, the enhancement of ability of obtaining hyper-resolution data provides data for feature extraction of 
ground objects. This paper employs SPOT5 RS image, based on the schema standard of 1:50000 topographic maps, to interpret the 
Qixiangcuo region in the Tibetan Plateau. The results indicate that the SPOT5 can meet the needs for 1:50000 topographic mapping. 
In order to further improve the interpretation multi-temporal and multi-resolution image are suggested to be used to interpret ground 
objects in 1:50000 topographic maps. 
1. INTRODUCTION 
1:50000 topographic maps are indispensable fundamental 
geographical information which can be used in national 
economic construction, geologiccal investigation and surveys, 
environment protection, water and biological resources 
exploration and national security. In western China, around 
200* 10 4 Km 2 area in Sorth Xinjiang desert region, Tibetan 
Plateau and the Traverse Mountains region does not have 
1:50000 topographic maps due to the weather condition, 
environment limitation, traffic inconvenience as well as the 
restrictions of the mapping technology and equipment in the 
past. The lacking of these maps not only restricts region 
economic and social development, the traffic and water 
conservancy construction, natural resources and mineral 
surveys, but also seriously impacts the economic development 
and national security of the western region of China. Along 
with the implementation of the western exploitation and the 
national sustainable development strategy, the development of 
the western infrastructure, resources survey and exploitation, 
ecological construction, environmental protection, the 
development of modem digital military directive system, 
antiterrorism and maintenance of the border security, the 
western scientific and cultural progress etc, all urgently need 
the 1:50000 topographic maps. Due to the need of large data 
sets for mapping 1:50000 topographic maps, high data 
acquisition technology is required. Especially for the routhless 
area, the spot reconnaissance is more difficult. In recent years, 
the rapid development and the popularization of remote sensing 
(RS) technology makes it possible to solve this problem. Using 
RS images as the data source to update the topographic map is 
increasingly used in land resources survey and monitoring 
(Jiang, 2001;Wang et al., 2002; Feng et al., 2004; Jiang et al., 
2004). 
In recent 20 years, the RS application has been highly improved 
by means of the development of space technology, computer 
technology and information technology (Liu et al., 1997; Lin et 
al., 2003; Zeng, 2004). Different RS images have different 
characteristics in feature extraction. SPOT-5 satellite was 
launched on May 4 th , 2002 by the French Ariane 4 launch 
vehicles. It is the last one of French Space Agency SPOTS 
satellite series which is obviously different from the former 4 
satellites. There are two HRG and a HRS instruments on it and 
it mainly provides 2.5m panchromatic image and 10m 
multispectral data. The characteristics of each satellite channel 
are shown in Table 1 and Figure 1. Currently, SPOT images are 
the main RS images used by the National Ministry of Land and 
Resources on large-scale national resource construction, the 
land use investigation, the digital territory project, the updating 
of soil resource and the land-use information system (Li, 2000; 
Deng et al., 2005). 
Sensor Band 
Ground 
Resolution 
Spectral range 
Panchromatic 
2.5m 
0.48-0.71 jum 
B1: Green 
10m 
0.50-0.59 jum 
B2:Red 
10m 
0.61-0.68 jum 
SPOT5 B3:Near Infrared 
(NIR) 
10m 
0.78-0.89 jum 
B4: Short 
wavelength 
infrared (SWIR) 
20m 
1.58-1.75 jum 
Table 1. SPOT5 data spectral bands and resolution 
2. THE OVERVIEW OF STUDY AREA 
The central part of the lacking area of 1:50000 topographic 
maps is Tibetan Plateau which is the highest one in the three- 
class ladder. It is bordered by the Pamirs in the west, Traverse- 
Mountains in the east, the Himalayas in the south, the Kunlun 
Mountain, Aerjin Mountain, and Qilian Mountain in the north. 
It covers the area from N 42°45' to N67°10' in latitude and from 
E81°00' to E106°00' in longitude. It is 2500km in length of EW 
and 1200km in width of SN. The area is 23 0 X 10 4 km 2 . In 
addition, the terrain of Tibetan Plateau is high because the 
average altitude is higher than 4500m and it is sloped from 
northwest to southeast. 
* Corresponding author. Zhigang Hong, hongzg@casm.ac.cn.