ere 
is 
ers 
nal 
lere 
the 
Was 
laic 
ion 
Ss 
D 
ble 
ear 
new 
4) 
QO 
3 
(D 
weighted coefficients determined according to a 
actual image. If we want to rise the vesstatton 
information, B shouid be positive, otherwise. B 
should be negative. With this modifi cation. “the 
vegetation information between the different 
temporal adjacent images tended to the same and the 
other area kept the original state. 
CR ORC the Color of Snow or Odd Cioud: As the 
therma! infrared channel assigned as a red color 
had peer negatived, the lower temperature objects 
appeared as a cyan color in the composite. This 
goes against the human habit and results a 
decrement of color contrast or a destruction of 
color balance also. 370 restore the snow or odd 
cloud to white, Use the snow or could detection 
method as the same as that in could removing step 
and take a negative operation for CH4 on snow part 
of a image and increase the brightness of CH! and 
CH2 property so that the snow is pure white. 
Change the Color of Sea: Like the color change 
of snow, the color of sea in the original composite 
image appeared as red because of its related higher 
temperature compared with the land in winter. 
Although the thermal infrared data shows the cold 
and warm currents clearly, but that may be beyond 
the comprehension for most readers who does not 
have the enough know i edge about geosciences. 
Therefore, it is necessary to change the color of 
sea to fit an overwhelming majority reader’s habit. 
A land mask was created by a bands arithmetic 
operation and a piecewise linear transform was used 
on the sea area to change the color from red to 
dark blue and remove thermal infrared information. 
in the illustration of South China Sea, thermal 
infrared information was kept but its contrastv and 
brightness was reduced extremely. 
image Sharpening Processing: The goal of this 
processing was to improve the image subjective 
quality. A Laplace transform for extraction of the 
edge feature was used to form a enhanced edge 
image. The resultant sharpening image was achieved 
by the means of a weighted addition of the origina: 
image and its enhanced edge image. This can be 
expressed as below; 
NEWIMAGE - A * IMAGE * B * EDGE!MAGE (5) 
{ BC A, A,BOO, A cB 5-1) 
A and B are the weighted coefficients. 
Integral Processing of the Image Base Map: With 
the advantage of digital image processing, a 
integra! image base map was made. Because the 
images outside China were not coilected compeletely, 
the area outside China in the image base map was 
assigned a grey tone instead of image except a zone 
along the international boundaries of China. A 
white contour along the continental edge, including 
the islands, was added using a erosion algorithm of 
mathmorphology. A mini cross mark was located at 
the point of intersection of latitude and longitude 
every 5 degrees for the purpose of register with 
other geographic elements maps. The legends, 
margins and the illustration of South China Sea 
were inserted at once according to the design. 
PHOTOMAP DESIGNING AND MAPPING 
After image processing, the satellite image mosaic 
is not a real map, only the following steps are 
finished. There are four parts in photomap design; 
expression of geographic elements, lettering, 
legend design and preliminary layout. 
Geographic elements are very important for photomap. 
Expression of geographic elements involves mainly 
three problems: i.e., selection, location, design 
of color and symbol of geographic elements. 
Satellit photomap is a new type of maps with 
383 
providing a? tot of "information. For genera! 
readers, a photomap itself provides a color 
photography reflecting native land so that they are 
interested in reading and interoreting; For readers 
with geographic knowledge, it provides much more 
information than many general-purpose mans, so that 
they can analyse and "taste" it. There is no doubt 
that geographic elements are first of al! contents. 
It is necessary and confined to express them. in 
this photomap, geographic elements included rivers, 
lakes, reservoirs, settlements, railways, highways. 
the tops of mountain and  internaiona! boundaries 
and so on. in the river system, all first-level 
branches and some secondary branches in China were 
selected, such as Han Shui river, Xiang Shui river; 
large lakes and reservoirs were selected, such as 
Boyang lake, Dongting lake and Danjang reservoir, 
so that the structure and density of rivers were 
represented. Cities above regional level were 
selected in principle, but parts of them were 
deleted in the east part of China with high density 
of population. On the contrary, in the west part of 
with low density besides regional cities, some 
county cities were also represented properly. In 
road nets, main lines of railways and national 
level highways ( in western China, trunk highways ) 
were selected. The grades of all social elements of 
selection were based on up-to-date published maps 
from China Cartographic Publishing House. For 
mountain elements, some famous and typic ones were 
selected ( for example,  Zhumulangma peak, Hange!i 
peak and yellow mountain and so on >. Witn 
geographic elements of objective reality, however, 
it is difficult to display szeosraphic elements 
because of the limited resoiution of image. To 
display the se!ected position correctly, we used 
visual interpretion and overtaping depiction to 
locate geographic elements with the same scale ona 
the same projection overlaping on the map 
image. In order to enhance the accuracy of 
depiction, it was necessary to depict correctively 
5 by 5 degrees geographic grids in the overiaped 
map one by one. As a matter of fact, - even though 
Some rivers and settlements, railways and highways 
were too small and unclear, there was no problem to 
locate them. International boundaries were based on 
standard nationa! boundaries published by China 
Cartographic Publishing House and refered to 
correlation of image information. That were 
examined and approved by National Surveying and 
Mapping Service. in order to stress on the 
geographic elements, we must apply symbois to 
express some trivial linear geographic elements 
(such as rivers, railways and highways) and very 
small settlements, some famous peaks and So on. 
Because the main color backgrounds of the image 
base map were blue, green and yellow, we decided to 
use violet for stream System, black and white 
sharply contrasted lines for railways, pinkish red 
lines for highways, black isosceles triang!e 
symbols for the tops of mountain. In settlements, 
some important cities were exaggerated by their 
out!ine figures and the rest were represented by 
circle pinkish symbols, like red flowers on the 
background. International boundaries were decorated 
with violet bands so that the main area was 
reinforced. Lettering is an important part in " map 
languages ". Because of big chinese characters, 
lettering in number restricts the selection of 
geographic elements to a great extent. in the 
photomap, lettering of first level rivers, a part 
of secondary level rivers and large lakes were 
selected in principle. All of selected settlements 
and the tops of mountain had lettering. Railways 
and highways were not lettered. Great geographic 
units ( such as Huabei plain, Dongbei plain and 
Sichuan Basin and so on ) and some mountain ranges 
( such as Ximalaya Range, Tanggula Range and