×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
Remote sensing for resources development and environmental management
Author
Damen, M. C. J.

470
Photo 1. Computer classification of a severely damaged
spruce-fir stand near Bad Peterstal, 300 m altitude.
For better presentation spruce was not divided in
different damage classes.
red:
spruce
green:
fir healthy
yellow:
fir damaged
blue:
fir severely damaged
orange:
aeciduos trees
white:
roads, glades
black:
shadow
Photo 2. Digital crown map for checking the classi
fication.
Same colours like Picture 1 only
violett: spruce damaged
red: spruce severely damaged
Picture 1. Results of forest damage classification using ground check,
CIR-Photointerpretation and multispectral scanner data
frcm different altitudes
□ S0/S1 healthy
52 damaged
53 severely damaged
In table 4 examples of the analysis of a classifica
tion are given for differently damaged firs.
Table 3. Classification of forest damage classes used
throughout Germany
SO-1
52
53
54
less than 25% needleloss
26 to 60% needleloss
more than 60% needleloss
dead
(healthy - sickly)
(damaged)
(severe damaged)
Table 4. Example for differently damaged firs (tree
#1 to #3 damage class S2, tree #4 to #6 darhage class
S3). Results are given in percent of classified pixels
per tree crown.
Stand A Stand B
tree
SO-1
S2
S3
misclassi-
average damage
100
number
fied
class rating
1
60
23
17
0
1,5
90
2
57
23
16
4
1,8
80
3
44
19
16
21
1,6
4
18
19
39
24
2,3
c 70
5
23
24
37
16
2,1
8 60
6
26
31
23
20
2,0
From this altitude a damaged tree crown contains up
to 100 pixels which can belong to any given damage
class. For example a fir with approximately 50%
needleloss will be classified as healthy in the top
of the crown where it seems to be green and healthy,
the lower branches which show some needleloss and
discoloration, will be classified as sick and the part
of the crown which is defoliated, will be classified
as severely damaged or dead. The degree of misclassi-
fied pixels will also increase with the damage degree
because more shadow, soil or ground vegetation re
flects through the defoliated crown and increases the
mixed signatures. The situation is the same when other
species are evaluated.
£ 20
m
i
i
Stand A: Old spruce stand mixed with fir; severely damaged
with low density
Stand B: Old spruce stand severly damaged and partly snow
breakage
Stand C: Young spruce stand with snow breakage
Stand D: Yeung spruce stand