3-2-2 Classification based on textural feature 
Some classes of grassland are still seriously confused after 
classification based of spectral feature, such as bushland and 
vegetable wood-grassland and grassland or bushland. But their 
texture feature are quite different. So we use texture measuring 
to distinguish the various confused classes. For this reason every 
4x4 or 8x8 window of image with confused classes is one by one 
transformed into spatial frequency domain. 
F(u,v)= FFT (f(x,y)) (5) 
Then we calculate their measures of texture: 
TX F = <5 si /F si (0,0) (6) 
or the measures of texture direct in spatial domain : 
TX3= 5^/Msi (7) 
Table 2 shows some measures of texture in 2 typical areas ( Qiyu 
mountain and Fubao mountain). 
Class 
Measure of texture 
TXs 
TX F 
Wood-grassi and 
12 
12 
Forest 
6 
15 
Grassland 
8 
10 
Vegetable plot 
10 
11 
Bush-grass 1 and 
8 
18 
Table 2. Texture measures of 5 classes 
We using feature of spectrum and texture in image 3 classes of 
grassland at first level are extracted. 
3-2-3 Classification based on non-remote sensed data 
There is second level classification, each class is classified 
into 3 groups by DHM and soil classes, and natural extension of 
grassland should be considered at the same time.Finally, a grass 
resources distribution map of Lichuan county at scale 1:100000 has 
been output by computer. Table 3 shows the area of 9 groups in 
grass resources map. 
Grassland 
( ha ) 
Bush-grassland 
( ha ) 
Wood-grassi and 
( ha ) 
Total 
( ha ) 
High-mountain 
Groups 
34444 
30432 
51277 
116153 
Mid-mountain 
Groups 
607 
825 
7954 
9386 
Low-mountain 
Groups 
877 
20 
0 
897 
Total 
35928 
31277 
59231 
126436 
TABLE 3. Area of 9 groups grassland in Lichuan county 
196