# Full text: Close-range imaging, long-range vision

```
re reduced as a labeling

: noise reduction

"| Noise data

P| Topographic data

phic and noise data
3.2 Interpolation
There are many lacks of 3D data such as shown in Figure 6,
then interpolation should be performed by the following
procedures.
‘Step (1)
+ Search a point in neighbouring the removed point.
+ If detect such a point, search another point in side direction.
+ If detect these point, the point is judged that belong to the
same slope, and lacks are interpolated by the point.
With this process, removed areas were interpolated, but
diagonally direction is not yet interpolated. Therefore,
'Step (2)
+ Search a point which was interpolated.
* If detect such a point, search another interpolated point to
diagonally direction.
+ If removed points were found between the points, the lack
space was interpolated using the detected points.
Here lack of 3D data was interpolated.
3.3 Unification of Coordinate System
In generally, many lack of data which is caused by blind spot
are estimated from a traverse station. Then, measurements from
multiple stations for the same scene become necessary.
When laser scanning sensor measure the mirror seal, the seal
can be found easily in intensity image since the intensity for a
seal have high brightness. So, mirror seals were set around the
sites as markers, and coordinates system for multiple traverse
points were unified automatically as the following procedures.
+ Set mirror seals at around sites.
+ Extraction of mirror seal (3D coordinate) in each intensity
image.
* Generation of TIN models using 3D data of the mirror at
every stations.
* Comparison with TIN models between the stations.
*Matching conjugate mirrors using area, corner angles and side
length for triangulations.
*Unification of coordinates system using conjugates mirrors.
Table 1 shows the equations which were used to unify the
coordinates system. This expression is generally used for three
dimensional transformation.
Table 1 Conversion Expressions

Coordinate Expressions
X x'—((x-x,) *cos 0 -(y-yj) *sin Q ) *x,,
Y y'-((x-xy) "sin  * (y-yy) *cos 0 ) +Yım
Z z'=(z-z;) +z,
Where,
(x,y,z) = Coordinate before conversion,
(x,y,z) - Coordinate after conversion
(x»ys) = Origin Coordinate at before conversion
(XmYınZm) = Origin Coordinate at after conversion
0 = Rotation angle
4. RESULTS OF THE PREMEASUREMENT
Figure 6 shows the image which noises were reduced for the
Figure 3. It can be seen that effective noise reduction were
performed by the above mentioned methods, similarly it can be
recognized that the necessity of interpolation. Figure 7 shows
interpolated image. From this figure, it can be said that the
interpolation was performed successfully.

Figure 6. Result of noise reduction

Figure 7. Interpolated image
-07—

```

## Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.