http://dx.doi.org/10.5755/j01.eee.20.5.7102 ELEKTRONIKA IR ELEKTROTECHNIKA, ISSN 1392–1215, VOL. 20, NO.

5, 2014

Influence of 3D Scanning Data Scattering to

Volume Measurement of Horizontal Fuel Tanks
V. Knyva1, M. Knyva1
1
Department of Electronics Engineering, Kaunas University of Technology,
Studentu St. 50, LT-51368, Kaunas, Lithuania
[email protected]

1Abstract—Calibration of the fuel tanks must be made once for horizontal fuel tank),
in a 5 years. For calibration of the fuel tank 3D laser scanning  preparation of data for processing (merging, filtering
and data processing method can be used. This article describes etc.),
the influence of 3D laser scanning data scattering to volume
measurement of horizontal fuel tank by means of relative  calculation of tanks volume,
error. It describes how to eliminate data scattering effects  preparation of graduation table.
using filtering. From inside the fuel tank must be scanned at least 2–3
times from different scanner positions. The best way to scan
Index Terms—3D laser scanning, fuel tank calibration. the horizontal fuel tank is to put a scanner at different sides
of the tank. The main problem to make a full scan of the
I. INTRODUCTION horizontal fuel tank is that most tanks have only one hatch.
Calibration of the fuel tanks must be made once in a 5 If you want to put scanner at the end of the tank you have to
years. For the tank calibration geometric and volumetric get inside the tank and move the scanner. But this solution is
methods can be used [1], [2]. Last few years authors were not acceptable. So, 3D scanner can be placed only near
working with calibration of various fuel tanks using 3D laser hatches. This can be solved by putting a scanner at different
scanning method. In previous papers [3], [4] authors heights and merging scanned data or by creating z plane
presented a new approach to the calibration of horizontal from points around it (upper, downer z planes).
and vertical fuel tanks which is based on the 3D laser Figure 1 shows z plane of the fuel tank scanned from one
scanning technique. The main idea was that the tank can be position – scanner was placed at the bottom of the tank. As
scanned using an accurate laser and volume of the tank can it can be seen, only half of z plane is “normally” filled with
be calculated using 3D point clouds. points (-4÷1 on x axis). This can be explained by the
The main advantages of the 3D laser scanning method scanner position near the hatch which is on the right side of
are: the tank.
1. Time saving and water waste problems; Figure 2 shows the same z plane (merged data) of the
2. Lower measurement temperature limit; horizontal fuel tank scanned from two different scanner
3. Only 1 person is needed for calibration. height positions. The first scanner position was on the
Experimental verification of the algorithms which was bottom of the tank and the second position was near the top
performed with different types of fuel tanks showed that for of the tank.
all fuel tanks relational error stays around 0,3 %–0,4 %. This As it can be seen the left part of the tank still has big
can be explained by scattering of the data merged from breaks between points. These breaks influence volume
different scans. calculations.
In this article we will analyse the influence of data
2.5
scattering effect to fuel tanks volume measurements and
explain the method for data scattering effect decreasing. 2

1.5
II. CREATION OF THE CLOUD OF POINTS FROM 3D LASER
SCANNING 1

Calibration of horizontal or vertical fuel tanks is followed

y,m

0.5
by these steps [3], [4]:
 inspection of the fuel tank, 0

 preparation of measurement gear and fuel tank for -0.5

verification,
 fuel tank scanning (2–3 scans for vertical and 1–2 scans -1

-1.5
-8 -7 -6 -5 -4 -3 -2 -1 0 1
x,m
Manuscript received October 11, 2013; accepted December 18, 2013. Fig. 1. 3D scan of the fuel tank (1 scan).

72
 ELEKTRONIKA IR ELEKTROTECHNIKA, ISSN 1392–1215, VOL. 20, NO. 5, 2014

2.5 merging from upper and downer z planes. After some

2
inspection of scanned data of different sizes horizontal fuel
tanks we found out that some z planes has more points on
1.5 one part and some on another. By merging scanned points in
these planes we can receive z plane without big “breaks”
1
between points in the point cloud. Figure 4 illustrates z plane
created from its neighbors z-0,5 cm and z+0,5 cm with step
y, m

0.5
of 0,1 cm. 10 z planes were merged into 1. Calculations of
0
area of this z plane gives results (Fig. 5) close to the real
-0.5 ones measured using liquid fill method (19,64 liters) which
is described in [3].
-1
2.5
-1.5
-8 -7 -6 -5 -4 -3 -2 -1 0 1
2
x, m
Fig. 2. 3D scan of the fuel tank (2 scans bottom and top merged).
1.5

A volume calculation of the scanned fuel tank was 1

described in previous author’s publications [3]. The main

y,m
formulas are 0.5

0
1 N 1
Aj   ( xi , j yi 1, j  xi 1, j yi , j ), [m 2 ], (1) -0.5
2 i 0
-1

where Aj –area of layer, j  1, H max / h ,

-1.5
-8 -7 -6 -5 -4 -3 -2 -1 0 1
H  1, H max [cm] fuel tanks filling level, H max - x, m

maximal filling level, h - scanning step [mm], xi and yi – Fig. 4. 3D scan of the fuel tank (1 scan different z plane layers merged).

point coordinates [m], N - point number; Area of z plane = 19.763

2.5

H M 2
VH  VH 1   A j  h  K H , [dm3 ], (2)
j  ( H 1)M 1 1.5

10
M  , (3) 1

h
y, m

0.5

where M – number of layers in 1 cm; K H – sum of 0

volumes of all inside construction elements.

-0.5
Algorithm of the calculations can be described as follows:
calculations of the area from point cloud of z plane must be -1

made using (1) and volume can be calculated as sum of all

-1.5
areas using (2). Figure 3 shows calculations of z plane area -8 -7 -6 -5 -4 -3 -2 -1 0 1
x, m
when 2 scans (laser at different heights) were used. It is Fig. 5. Calculations of z plane area (1 scan different z plane layers
clear that at the left side we have area cut offs. merged).

Area of z plane = 19.1293 TABLE I. DEPENDENCIES OF RELATIVE ERROR FROM DIFFERENT

2.5
DATA MERGING METHODS.
2
z plane area (2 z plane area (1
z plane area
scans bottom scan different z
(liquid fill
1.5
and top plane layers
method) liters
merged) liters merged) liters
1
1 mm 19,12 19,76 19,64
Relative
-2,7 +0,6 -
error %
y, m

0.5
1 cm 191,3 197,6 196,52
0 Relative
-2,6 +0,62 -
error %
-0.5 10 cm
Relative
-2,6 +0,8 -
-1 error %
Full volume 23401,68
22653,35 23530,45
-1.5 (240 cm)
-8 -7 -6 -5 -4 -3 -2 -1 0 1
x, m
Relative
-2,63 +1,02 -
error %
Fig. 3. Calculations of z plane area (2 scans bottom and top merged).

Another method for getting “full” scan of z plane is data Differences between methods used for creation of point
73
 ELEKTRONIKA IR ELEKTROTECHNIKA, ISSN 1392–1215, VOL. 20, NO. 5, 2014

cloud (z plane) can be analysed by relative error. Table I  Data scattering with boundary -5 mm to 5 mm.
illustrates how relative error changes when calculating fuel Bigger boundary levels were not analysed because the
tanks volume at different heights. For illustrations only one influence to final volume value and the relative error is too
horizontal fuel tank was used, but our analysis of different big. Relative error was calculated at 4 different levels.
size horizontal fuel tanks shows that the same behaviour Starting in the middle of the fuel tank (1,20 m) with the
occurs in all volume measurements. As it can be seen h  1 mm (2):
relative error when using 2 scans (bottom and top merged)  Volume of 1 mm;
stays at “–” zone. This means that using this point merging  Volume of 1 cm;
method we always get lower volume. This can be explained,  Volume of 10 cm;
as mentioned earlier (Fig. 3), by breaks between points.  Full volume 240 cm.
On the other hand, when creating point cloud from 1 scan
and merging points from neighbouring layers always gives a TABLE II. DEPENDENCIES OF RELATIVE ERROR FROM DATA
little bit increased volume (1 %–1,2 %). This can be SCATTERING.
Volume Volume Volume Volume
explained by data scattering. Normally relative error must Measured
(scattering (scattering (scattering (no data
stay around ±0,3 % [5]. fuel tanks
from -1mm -2.5mm to -5mm to scattering
height
to 1mm ) l 2.5mm) l 5mm ) l )l
III. INFLUENCE OF THE DATA SCATTERING TO VOLUME 1 mm 16,35 16,37 16,51 16,31
Relative
MEASUREMENTS error %
0,25 0,4 1,2 -
Data scattering appears after merging different z planes. 1 cm 163,84 164,15 165,7 163,42
This happens because of natural cylindrical horizontal fuel Relative
0,26 0,45 1,4 -
error %
tanks shape. After inspection of 50–60 3D scans of different 10 cm 1687,78 1693,04 1709,82 1682,9
fuel tanks we found out that the best z planes merging result Relative
0,29 0,6 1,6 -
is received when 10 planes is merged (z-0,5 cm and error %
Full volume 37333,67 37501,09 37761,52 37203,46
z+0,5 cm). The main criteria to receive best result were to Relative
decrease breaks between points. Data merging gives side 0,35 0,8 1,5 -
error %
effect – data scattering. Figure 6 shows zoomed section of
the fuel tank z plane which was used for volume As it can be seen from Table II data scattering with
calculations. boundary of -1 mm to 1mm gives acceptable results; relative
error stays around 0,3 %. But at increased boundaries (-2,5
1.18
mm to 2,5 mm or -5 mm to 5 mm) relative error increases to
0,9 %– 1,5 %. Figure 7 illustrates how different data
1.175
scattering boundary levels effect volume calculation going
1.17
through all the fuel tank from bottom to top (50÷240 cm).
1.165 First curve from the top represents relative error when data
scattering boundary is -1 mm to 1 mm, second – -2,5 mm to
y, m

1.16

1.155
2,5 mm, and third -5 mm to 5 mm.
1.15

1.145

1.14

1.135
-0.27 -0.26 -0.25 -0.24 -0.23 -0.22 -0.21 -0.2 -0.19
x, m
Fig. 6. Data scattering of merged layers.

For better understanding of how data scattering influences

volume calculations and for finding the acceptable limit of
scanned data scattering mathematical model of horizontal
fuel tank was created. We adopted model to real sizes of the Fig. 7. Dep1endencies of relative error from data scattering.
fuel tank and created the possibility of imitation of data
scattering. The model creates the array of coordinates (x, y, This data scattering after merging problems can be solved
z) of points which form cloud of points. For data scattering using data filtering methods.
imitation for each coordinate random number from
determined interval (for example from -0,5 cm to 0,5 cm ) is IV. METHOD OF ELIMINATION OF 3D LASER SCANNED
generated and the point is moved from its ideal position. DATA SCATTERING AFTER MERGING – EXPERIMENTAL
Algorithm for ideal fuel tanks mathematical model was RESULTS
created in MATLAB language. Effect of data scattering can be minimized using data
For experiments with ideal fuel tank model 4 data sets filtering. As can be seen from Fig. 4–Fig. 5, the biggest data
were created: scattering can be found on the fuel tank side where the
 No data scattering; scanner was placed. For data scattering filtering we used
 Data scattering with boundary -1 mm to 1mm; algorithm based on regression, which creates boundary
 Data scattering with boundary -2,5 mm to 2,5 mm; levels calculated by equations for each point coordinate:

74
 ELEKTRONIKA IR ELEKTROTECHNIKA, ISSN 1392–1215, VOL. 20, NO. 5, 2014

xi   x  i  n   x  i  n  1    x  i  n   / S n , [m], (4) elimination method can be compared with works presented

by W. Jintao [6], [7]. Researchers received measured
yi   y  i  n   y  i  n  1    y  i  n   / S n , [m], (5) relative error around 0,4 % for horizontal and 0,6 % for
vertical fuel tanks. Using our data scattering elimination
where xi , yi – coordinates of i point, n –number of points method we managed to decrease relative error to 0,3 %.
taken from one and other i-th point side, Sn – a span equal
to
1
Sn  . (6)
2n  1
Procedure of filtering can be described as follows:
 Creation of z plane data points array (Fig. 4);
 Calculation of boundary levels;
 Filtering point which are outside boundary; Fig. 10. Dependencies of relative error from data scattering after data
 Creation of new filtered z plane. filtering using real scanned data.
We calculate how many points n to use from the size of z
plane data array. This algorithm didn’t remove all data This article describes the influence of 3D laser scanning
scattering effect, but it decreased it to acceptable level. data scattering to volume measurement of horizontal fuel
Figure 8 illustrates how relative error changes after tank by means of relative error. Authors present the method
decreasing data scattering. As expected, changes of relative for elimination of data scattering effects.
error when data scattering has boundary -1 mm to 1mm is
unnoticeable (dot line), but when boundary increases to V. CONCLUSIONS
±2,5 mm (dash line) or ±5 mm (full line) relative error stays 1. Analysis of the different fuel tank scanning methods
around desired 0,3 % level. showed that the best result is received when the fuel tank
is scanned from one point and data merging is used. Data
merging method forms a side effect – data scattering.
2. Experiments with mathematical model of the fuel tank
showed that scattering with boundary of -1 mm to 1 mm
gives still acceptable results by means of relative error
(stays around 0,3 %), but bigger boundary levels gives
unacceptable results.
3. Experiments with ideal fuel tank model showed that
the influence of data scattering can be minimized to
Fig. 8. Dependencies of relative error from data scattering after data acceptable level of relative error using data filtering
filtering. algorithm.
4. Experiments with real scanned data proved that
filtering of all unnecessary data improves relative error
and helps to keep it in satisfactory ranges ±0,3 %.
REFERENCES
[1] V. V. Nosach, B. M. Belyaev, “The calibration of large vertical
cylindrical tanks by a geometrical method”, Measurement
Techniques, Vol. 45, No. 11, 2002 [Online]. Available:
http://dx.doi.org/10.1023/A:1022062532494
[2] V. V. Nosach, B. M. Belyaev, “Salient features of procedures for
geometric calibration of large vertical cylindrical tanks”,
Fig. 9. Dependencies of relative error from data scattering after data Measurement Techniques, Vol. 46, No. 1, 2003 [Online]. Available:
filtering using real scanned data. http://dx.doi.org/10.1023/A:1023413606206
[3] V. Knyva, M. Knyva, “New method for calibration of horizontal fuel
For experiments with real scanned data we ported tanks”, Elektronika ir Elektrotechnika, vol. 18, no. 9, pp. 91–94,
2012. [Online]. Available: http://dx.doi.org/10.5755/j01.eee.
scattered data filtering algorithm to LABWINDOWS CVI 18.9.2816.
environment which is based on ANSI C language for faster [4] V. Knyva, M. Knyva, “New method for calibration of vertical fuel
performance. Cloud of points was formed using only one 1 tanks”, Elektronika ir Elektrotechnika, vol. 19, no. 8, pp. 37–40,
2013. [Online]. Available: http://dx.doi.org/10.5755/j01.eee19.8.5391
scan using data merging and scattered data filtering [5] “Skystų produktų talpyklų patikra 3d lazerinio skenavimo metodu“
algorithm. Experimental verification of the algorithm was BPM 300637345-01:2013. – Lithuania: KTU Metrology institute,
performed with different fuel tanks of various sizes. Figure 2013. (in Lithuanian).
9–Fig. 10 represent dependency of relative error from data [6] W. Jintao, L. Zhang, L. Guo, X. Bao, L. Tong, „Precise measurement
of liquid petroleum tank volume based on data cloud analysis“, in
scattering after data filtering using real scanned fuel tank Proc. Sixth International Symposium on Precision Engineering
data. As it can be seen from figures, filtering all unnecessary Measurements and Instrumentation, China, 2010
data improves relative error and helps to keep it in [7] W. Jintao, L. Zhang, L. Guo, X. Bao, L. Tong, „Research on 3D laser
scanning method for tank volume metrology and its comparison
satisfactory ranges ±0.3 % [5] (bottom curve). experiment analysis“, Chinese Journal of Scientific Instrument,
The experimental results of 3D scanned data scattering China, 2010.

75