**THE STUDY OF RELIEF GENERALIZATION BASED ON DIGITAL ELEVATION
MODEL**

D. Xu, H. Liu

*Department of Cartography, Institute of Surveying and Mapping*

*liu2000@vip.sina.com*

The automatic relief
generalization is one of the conventional parts of the map generalization. This
paper studies the theories and algorithms of relief generalization which based
on digital elevation model, and gives a brief discussion on the advantages and disadvantages
of the algorithms mostly. Then, this paper chooses some of the representative algorithms
for thorough research and realizes them with visual C++. And also, this paper
tests all the programs with 1:250000 relief map data and analyzes the effect
and efficiency of each kind of the programs, and proposes some improve methods
for these programs. The main subjects of this paper are as follows:

1. Check-up of the original contour data. Because there are some inevitable errors in the original contour data, such as contour line with the error elevation and contour line intersect with each other. These errors will affect the effect of the programs, they even to bring the program to collapse, so we must inspect the data before we use it. This part of the paper attempts to figure out the criterion of the data and the method of data examination.

2. The algorithms of relief
generalization which based on digital elevation model and its realization. This
part of the paper studies the algorithms
of relief generalization which based on digital elevation model and programs
some of the representative algorithms with C programming language, such as
generalization of TIN based on the distance between the point and its corresponding
plane, multi-scale generalization of digital elevation model based on wavelet analysis,
generalization of digital elevation model based on three dimensional Douglas-Peucker algorithm and so on.

3. Analysis of the effect of
the algorithms. With the real 1:250000 relief map data, all the realized algorithms
were tested to analyze the efficiency of each algorithms, different generalization
effect of each algorithms with different parameter and what kinds of situation
does a given algorithm with given parameter could be applied to.

4. Improvement of the algorithms.
From the analyses above, we know that there are still some disadvantages on the
efficiency of each algorithms and generalization effect of each algorithms,
this paper proposes some improve methods for further study.