=Paper=
{{Paper
|id=Vol-2035/paper15
|storemode=property
|title=System GRANICS for Particle-Size Distribution Determination by Photoanalysis
|pdfUrl=https://ceur-ws.org/Vol-2035/paper15.pdf
|volume=Vol-2035
|authors=Vasiliy N. Kruglov,Artem V. Kruglov,Alexey G. Gruh
}}
==System GRANICS for Particle-Size Distribution Determination by Photoanalysis==
https://ceur-ws.org/Vol-2035/paper15.pdf
Proceedings of Information Technologies, Telecommunications and Control Systems (ITTCS) - 2017
System GRANICS for Particle-Size Distribution Determination by
Photoanalysis
Vasiliy N. Kruglov Artem V. Kruglov Alexey G. Gruh
Ural Federal University Ural Federal University AO “Evraz KGOK”
Yekaterinburg, Russia, Yekaterinburg, Russia, Kachkanar, Russia
v.krouglov@mail.ru avkruglov@yandex.ru
Abstract
The system for determining the grain-size distribution of bulk material
particles is described by the example of estimating the dimensions of crushed
pieces of iron ore. The stages of the algorithm for processing of iron-ore
particle images are illustrated. The results of full-scale testing of the system
at the mining and processing plant are analyzed.
1 Introduction
Grain-size analysis of particles of bulk material is one of the most important indicators of the quality of many technological
processes. Regarding to the process of ore crushing at mining enterprises, it allows operator to control the operation modes
of large, medium and fine crushers depending on the obtained material distribution by grain-size classes. Thus, the
optimization of a technological process for crushing solid material and the reduction of the company's costs for electricity
and consumables are possible.
To estimate the grain-size of crushed ore pieces, it is necessary to construct the devices with advanced accuracy,
reliability and speed as well as multifunctional properties. The devices based on vision systems are mostly met all these
requirements. An important advantage of such systems is the non-contact method of measuring the parameters of the object
of interest. This circumstance provides such devices with reliability and durability.
2 Materials and methods
The urgency of the problem of estimating the grain-size distribution of bulk materials particles is confirmed by the number
of publications devoted to its solution [1-9]. This work presents the results of algorithmic and design-technological studies
on the development of the hardware-software complex GRANICS. The complex is designed to determine the dimensions
of bulk materials particles, in particular, pieces of medium-sized iron ore, moving on a conveyor or in bolting mill. The
complex includes video sensors and a PC-station with a specific software which calculates the sizes of crushed pieces and
performs their statistical processing. The principle of the GRANICS system operation is shown in Fig. 1.
Figure 1. Scheme of the non-contact measurement of bulk material
96
Proceedings of Information Technologies, Telecommunications and Control Systems (ITTCS) - 2017
Using the video capturing sensor (VCS), the original image of the crushed ore particles lying on the conveyor belt is
captured (Fig. 2).
Figure 2. The initial image of crushed ore particles on the conveyor belt before processing
The algorithm for determining the dimensions of bulk particles is as follows. At the first stage, the original image
F (i, j ) is processed by a median filter. The aperture of the median filter W (i, j ) is a square matrix of size 3x3.
Traditionally, the filter response is calculated as:
MED[W (i, j )] MED[{ f (i k , j l )}; k , l 1,1] (1)
The nature of the impulse distortions on the resulting video images is in the form of "pepper" with an area of not more
than 3 pixels. Given this fact, a sliding hybrid median filtering is used, which is as follows.
W (i, j ) {clm(i, j 1), clm(i, j ), clm(i, j 1)}, (2)
where clm(i, j ) { f (i k , j ), k 1,1}
Then the response of the sliding hybrid median filter will be calculated according to the formula
HMED[W (i, j )] MED[ MED[clm(i, j 1)], MED[clm(i, j )], MED[clm(i, j 1)]] (3)
When shifting W (i, j ) by one position, we get
HMED[W (i, j 1)] MED[ MED[clm(i, j )], MED[clm(i, j 1)], MED[clm(i, j 2)]] (4)
It can be seen from (1) and (2) that with W (i, j ) of 3x3 size, the number of sorting operations is clearly reduced by 2
times. The resulting image is subjected to smoothing filtering using a 17x17 square mask consisting of unit elements. The
next stage is borders emphasizing which can be implemented by a nonlinear method of detecting the differences, based on
the homomorphic image processing proposed by Wallis [10]. According to this method, the element of the contrasted image
is defined as
1 [ F ( j , k )]4
G ( j , k ) log (5)
4 F ( j 1, k ) F ( j 1, k ) F ( j , k 1) F ( j , k 1)
97
Proceedings of Information Technologies, Telecommunications and Control Systems (ITTCS) - 2017
As a result of these transformations, a contour image of the visible layer of crushed ore pieces F c (i, j ) is obtained
(Figure 3), which is the aggregate of M disjoint regions
M
F c (i, j ) S k . (6)
k 1
Figure 3. Contour image of crushed ore particles
It is noted in [9] that one of the main features that allow us to classify an isolated region S k as a particle of crushed ore
is a sign of roundness calculated by formula
L
(d avg (d ))
i
k
i
k 2
Rk i 1 (7)
L
where di ( xi C x ) ( yi C y ) is a distance from each boundary point ( xi , yi ), i 1, L of the region S k to
k k k 2 k k 2 k k
its center (C xk , C yk ) (mean( xmk ), mean( ymk )),{xmk , ymk } S k ,
L
d i
k
and avg (d ) i 1
k
i
L
The segmentation procedure is implemented as follows. For each Sk , k 1, M Rk is calculated. If Rk = THRESHOLD, then the selected fragment of the image is
subdivided into the subdomains by the "watershed" algorithm [11,12]. In Fig. 4 the results of this algorithm is shown.
98
Proceedings of Information Technologies, Telecommunications and Control Systems (ITTCS) - 2017
Figure 4. Formation of ore particle images using the "watershed" algorithm
At the next stage, for each selected area, its main axis is calculated, the length of which E2a is determined by the
formula
p 2 2 p 2 2
E2 a (8)
2 2 A 2 2 A
where p is the perimeter of the selected area, and A is its area in pixels. In Fig. 5 main axes are represented by black
lines. In the crosswise direction the maximum width E2b of the ore particle region is determined, the value of which is
calculated in this way
p 2 2 p 2 2
E2b (9)
2 2 A 2 2 A
In Fig. 5 lines corresponding to the maximum width of the ore particle region are drawn in white. These axes will
characterize the dimensions of bulk material pieces.
Figure 5. The results of estimating the dimensions of crushed ore pieces
99
Proceedings of Information Technologies, Telecommunications and Control Systems (ITTCS) - 2017
3 Testing and Results
Industrial tests of the GRANICS complex were carried out, in particular, at the Erdenet Mining Corporation, Mongolia.
According to the program of industrial tests, the central research laboratory of the plant (CRL) sampled the ore after fine
crushing from the conveyor No. 18 in order to determine the condition and nominal grain-size, with parallel fixation of the
same parameters of the crushed ore according to the GRANICS output. The grain-size distribution of the ore is shown in
Tables 1 and 2.
Table 1. Grain-size distribution according to CRL analysis
Grain-size
Exp №1 Exp №2 Exp №3 Exp №4 Exp №5 Exp №6 Exp №7
class, mm
20 0.09 0.08 0.05 0.14 0.08 0.06 0.13
18 0.46 0.61 0.53 0.73 0.48 0.50 0.52
15 2.32 2.46 2.45 3.76 2.80 2.74 2.37
12.5 11.48 11.72 11.46 14.02 13.38 10.85 12.08
6.3 32.91 33.67 35.47 32.94 35.02 31.43 36.66
-6.3 52.74 51.46 50.04 48.41 48.24 54.42 48.24
TOTAL 100.00 100.00 100.00 100.00 100.00 100.00 100.00
Table 2. Grain-size distribution according to GRANICS output result
Grain-size
Exp №1 Exp №2 Exp №3 Exp №4 Exp №5 Exp №6 Exp №7
class, mm
20 0.47 0.62 0.46 0.44 0.67 0.43 0.66
15 2.34 3.13 2.66 2.17 2.93 2.15 2.95
12 7.82 9.65 8.40 7.39 8.79 7.30 8.76
10 9.73 11.32 10.49 9.37 10.34 9.44 10.34
6 20.16 22.49 21.29 19.05 19.80 19.60 20.87
-6 59.48 52.79 56.70 61.58 57.47 61.08 56.42
TOTAL 100.00 100.00 100.00 100.00 100.00 100.00 100.00
To perform a comparative analysis of the obtained data, the results of determining the nominal grain-size of the crushed
ore and the content of the "+15 mm" grain-size class are presented in Table 3.
Table 3. The nominal grain-size of crushed ore and the content of the grain-size class "+15 mm"
nominal grain-size, мм content of grain-size class "+15 mm"
Experiment №
CRL GRANICS CRL GRANICS
1 14.30 13.81 2.87 2.81
2 14.40 14.37 3.15 3.75
3 14.40 14.03 3.03 3.12
4 14.80 13.71 4.63 2.61
5 14.50 14.26 3.36 3.60
6 14.40 13.71 3.30 2.58
7 14.40 14.26 3.02 3.61
average value 14.46 14.02 3.34 3.15
relative error 3.01% 5.48%
100
Proceedings of Information Technologies, Telecommunications and Control Systems (ITTCS) - 2017
4 Conclusion
System GRANICS and CRL have different principles of evaluation of the same technological process. However, the above
results show that the relative error of the GRANICS complex in determining of the nominal grain-size of the crushed ore
was only 3.01% compared to the CRL, and in the evaluation of the controlled class "+ 15mm" the error is 5.48%.
Thus, the performed industrial tests illustrate that the complex accurately determines the grain-size distribution of the
crushed particles.
References
1. C. Shanthi , R. Kingsley Porpatham, N. Pappa. Image Analysis for Particle Size Distribution / International Journal
of Engineering and Technology (IJET). 2014, Vol. 6 No 3, pp. 1340-1345.
2. M. Mangold. Use of a Kalman filter to reconstruct particle size distributions from FBRM measurements / Journal of
Chemical Engineering Science. 2011, 70, pp. 99-108.
3. Z. Zhang, J. Yang, L. Ding and Y. Zhao. An improved estimation of coal particle mass using image analysis /
Journal of Powder Technology. 2012, vol. 229. Pp. 178-184.
4. C. Igathinathane, U. Ulusoy and L.O. Pordesimo. Comparison of particle size distribution of celestite mineral by
machine vision ΣVolume approach and mechanical sieving, / Journal of Powder Technology. 2011, vol. 215, pp.
177-278.
5. Z. Huang, J. Zhang, X. Li and H. Zhang. Remote sensing image segmentation based on dynamic statistical region
merging. / Optik-International Journal for Light and Electron Optics. 2014, 125(2), pp. 870–875.
6. S. Al-Thyabat, N.J. Miles, T.S. Koh,, Estimation of the size distribution of particles moving on a conveyor belt /
Minerals Engineering. 2007. 20, pp. 72-83.
7. S. Arasan, S. Akbulut, A.S. Hasiloglu. Effect of particle size and shape on the grainsize distribution using image
analysis / International Journal of Civil and Structural Engineering. 2011. Vol. 1, No 4., pp. 968-985.
8. Beth Junker. Measurement of bubble and pellet size distributions: past and current image analysis technology //
Bioprocess Biosyst Eng (2006), vol. 29, p. 185-206.
9. Simphiwe Mkwelo. A machine vision-based approach to measuring the size distribution of rocks on a conveyor belt.
Dissertation for the degree of Master of Science in Engineering. Submitted to the Department of Electrical
Engineering. University of Cape Town. May 2004. 147 p.
10. U. Prett, Digital image processing in 2 books: / Translated from English. M.: Mir, 1982; vol. 2:, 480 p. (In Russian)..
11. Sonka M., Hlavac V., Boyle R. Image processing, analysis, and machine vision / Milan Sonka - Brooks/Cole
Publishing Company, 1999,. 770 p.
12. R.C. Gonzalez, R.E. Woods, Digital Image Processing, 3th edition / Prentice-Hall, Inc, USA, 2007, 976 p.
101