453 
. Beijing 2008 
REMOTE SENSING IMAGE 
INTERPRETATION STUDY SERVING URBAN PLANNING BASED ON GIS 
Qian Wang 3 ’ b ’ c , Jianping Chen 3 ’ b ’ c , Yi Tian 3 ’ b ’ c 
a Department of Global science and Resource, CUGB, Xueyuan Road, Beijing - (wangqianl2031)@163.com 
institute of Land Resources and High Techniques , CUGB, Xueyuan Road, Beijing - (wangqianl 2031)@ 163.com 
c Laboratory of Beijing Land Resources Information Development, CUGB, Xueyuan Road, Beijing - 
(wangqian 12031 )@ 163 .com 
PS-13, WgS IV/3 
KEY WORDS: Remote Sensing, GIS, Visual Interpretation, Supervised Classification, Urban Planning, Present Situation 
ABSTRACT: 
The level of urbanism is an important parameter to measure a country’s civilization degree, so it is important to make a reasonable 
planning for the city. The primary stage of making an urban planning is the survey of present situation, which needed a lot of people, 
material and money in the past. But now remote sensing has been used as a new means to get the space information and some 
attribute information. That’s more speedy, accurate and economical. There are mainly two ways to do the remote sensing image 
classification. One is visual interpretation, and the other is computer automatic interpretation. Computer automatic interpretation is 
very fast and facile, but it is difficult to make use of people’s experience and knowledge. It is inclined to classify a subject in to a 
wrong class or omit some objects. Meanwhile, city is a very complex geographic synthesis of population, resources, environment 
and social economy. There are different objects may have same spectral character or the same class of objects have different spectral 
characters. And the emphasis of making an urban planning is to research city functional cells and the structure rationality. The result 
of computer automatic classifying can not satisfy the city plan. In contrast, though interactive visual interpretation needs a longer 
time, it can make full use of people’s experience and knowledge. The result is more reasonable and correct. The paper compares the 
supervised classification with interactive visual interpretation, and indicates that interactive visual interpretation can serve the urban 
planning better. 
1. INTRODUCTION OF URBAN PLANNING THEOTY 
The urban is a compound system of human and nature. It is also 
a high-dense geographical synthesis of population, resources, 
environment, social economic and so on. As one sign of 
civilization and social progress, the city’s effects on national 
politics, economics and culture become prominent increasingly. 
In other words, the urbanization’s level is a significant 
parameter to measure a country’s extent of civilization, social 
progress and economic. So it is very important to make 
reasonable and fit urban planning and management (Fan 
Wenbing, 2006). 
According to the urban development aims, urban planning 
constitutes the urban character, scale and development direction, 
makes use of the urban land reasonably. Urban planning relates 
to politics, economic, society, technology, art and 
comprehensive domains of human life. It is not only integrated, 
but also concerned with the policy and practice (Zheng Chaogui, 
2004). 
The primary phase of urban planning is present situation 
investigation. In the past, it always consumed a lot of labor 
force, material and money. The result was not timely and exact. 
Nowadays, the remote sensing technology can be used to 
investigate urban terrain, physiognomy, lakes, plants, sights, 
traffic, land utilization and building distribution quickly. As a 
main method to obtain and update urban geometric information 
and some attribute information, the remote sensing technology 
is quick, exact and economical (Xu Zhenhua, 2005). 
2. INTRODUCTION OF REMOTE SESING IMAGE 
CLASSIFICATION METHODS 
The important content of investigating present situation with 
remote sensing technology is the remote sensing image 
classification. The remote sensing image classification has 
mainly two ways. One is the visual interpretation, the other is 
computer automatic classification. The visual interpretation is a 
process of utilizing biology geosciences rules and the check 
analysis method to carry on generalized analysis and logical 
deduction according to the operating person's experience and 
knowledge, the sample's image characteristic and the spatial 
characteristic (shape, size, shadow, texture, graph, position and 
layout), together with many kinds of non-remote sensing 
information. The computer automatic classification uses the 
pattern recognition technology and the artificial intelligence 
technology to carry on the analysis and the deduction, 
understand the remote sensing image, and complete the 
interpretation of the remote sensing image finally, according to 
the land object’s characteristic in the remote sensing image and 
the goal land object’s interpretation experience and image 
formation rules in the expert knowledge library, based on the 
computer system (Liu Renzhao, 2005). 
The visual interpretation classification's application mainly has 
two forms. The first form is carrying on the artificial 
interpretation and drawing the land utilization chart manually to 
obtain the land utilization category information with the aerial 
image, the spatial image and many other kinds of remote 
sensing images. The second form is mainly the man-machine