displayed, shown as the right figure in Figure 8. In this view,
the image is magnified to better visibility for the user to decide
the trees one by one in the selected plot rectangle region. User
clicks the mouse in the range of individual tree crown and
counts the number of trees in the selected plot region. In this
way, the number of trees per unit area in the sub-compartment,
that is, tree density can be estimated. At this time, the system
estimates tree height at each selected point in the tree crown
region based on the height of DSM and DTM. The average
value of tree height on these selected points is used as average
tree height in the sub-compartment.
lick Tree Position by Human |
Number : 73
Area :0,.10ha
E
Tree Density Estimation UI
Figure 8. Estimation UI of Tree Density
3.2. Forest Resource Estimation
The proposed system estimates resource volume automatically
using estimated tree density and average tree height of the
target sub-compartment. The estimation is carried out under the
existing "system yield table", based on the forest age and site
class acquired from forest registration, average tree height, tree
density and the area of the sub-compartment. When user clicks
the "harvest" button, the window to set the thinning rate appears.
After user sets the thinning rate (0 — 100%), the estimation
results such as the number of timber and tree volume in every
diameter class, income expected from unit price set in advance
and volume of lumber remnants are output in excel format,
shown in Figure 9.
In this way, by using the proposed forest management GIS
system, user can obtain the volume of lumber acquired from
thinning in certain sub-compartment very easily by himself.
Compartment || Diameter Number of Tree Income Lumber
Number Class (cm) | | Timber Volume Remnants
Figure 9. Result of Forest Resource Estimation
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B8, 2012
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia
4. EXPERIMENT
We build the proposed GIS system and apply it to the
experiment forest of Mie University (East to West : 4km, North
to South : 1.5km, Area : 460ha ). We select 82 sub-
compartments as experimental area where Japanese cedar and
Japanese cypress grow. We experiment to estimate volume of
lumber and lumber remnants after acquiring tree density and
height of each sub-compartment.
DSM is generated from digital aerial photograph (3 courses, 72
photographs) by stereo processing. As DTM, ground height
data generated from LIDAR is used. The information of tree
species, forest age, site class and area of sub-compartment is
obtained from the forest registration. Table 10 shows the
specification of aerial photograph and Figure 11 shows the
image overlaying orthophoto and sub-compartment map used in
this experiment.
Camera DMC (Z/I Imaging)
Image Size 13,824 x 7,680
Pixel Size 12 micrometer
Focal Length 120 mm
Scale 1/5,000
Spatial Resolution 10 cm
Overlap Rate OL:75%, SL:55%
Table 10. Specification of Aerial Photograph
Figure 11. Aerial Photograph used for Experiments
In this experiment, volume of lumber and that of lumber
remnants are estimated. The result is shown in Figure 12 and
Figure 13. The result indicates that the forest resource
management system can estimate the forest resource volume.
Furthermore, because it takes only almost 2 hours for
estimating resource volume of target sub-compartments on the
second stage, the result indicates that the system succeeds in
cost reduction.
Lumber Volume
(m?/ha)
EC 120
201-400
BE 401-600
E 601-300
BE o0
Curr
fores
fores
and |
On t
exan
trees
the :
num
to e:
accu
com
regis
fores
the
prob
large
intrc
in th
estir
nece
prac
In t
Syst
pho
aeri:
inte;
reso
The
mar
of f
timl
was
In 1
esti;
fore
prac
duri
rela