Publication Details

Comparison of bubble detectors and size distribution estimators

KÄLVIÄINEN Heikki, EEROLA Tuomas, LENSU Lasse, ILONEN Jarmo, ZEMČÍK Pavel, DUBSKÁ Markéta and JURÁNEK Roman. Comparison of bubble detectors and size distribution estimators. Pattern Recognition Letters, vol. 101, no. 1, 2018, pp. 60-66. ISSN 0167-8655. Available from: https://www.sciencedirect.com/science/article/pii/S0167865517304282
Czech title
Srovnání postupů detekce bublin a zjištění velikostí i distribuce velikostí
Type
journal article
Language
english
Authors
Kälviäinen Heikki, prof. (LUT)
Eerola Tuomas, doc., Ph.D. (LUT)
Lensu Lasse (TUT)
Ilonen Jarmo (LUT)
Zemčík Pavel, prof. Dr. Ing. (DCGM FIT BUT)
Dubská Markéta, Ing., Ph.D. (DCGM FIT BUT)
Juránek Roman, Ing., Ph.D. (DCGM FIT BUT)
URL
Keywords

Bubble detection
Size distribution estimation
Boosting-based detection
Convolutional neural networks (CNN)
Pulping

Abstract

Detection, counting and characterization of bubbles, that is, transparent objects in a liquid, is important in many industrial applications. These applications include monitoring of pulp delignification and multiphase dispersion processes common in the chemical, pharmaceutical, and food industries. Typically the aim is to measure the bubble size distribution. In this paper, we present a comprehensive comparison of bubble detection methods for challenging industrial image data. Moreover, we compare the detection-based methods to a direct bubble size distribution estimation method that does not require the detection of individual bubbles. The experiments showed that the approach based on a convolutional neural network (CNN) outperforms the other methods in detection accuracy. However, the boosting-based approaches were remarkably faster to compute. The power spectrum approach for direct bubble size distribution estimation produced accurate distributions and it is fast to compute, but it does not provide the spatial locations of the bubbles. Selecting the most suitable method depends on the specific application.

Published
2018
Pages
60-66
Journal
Pattern Recognition Letters, vol. 101, no. 1, ISSN 0167-8655
Publisher
Elsevier Science
DOI
10.1016/j.patrec.2017.11.014
UT WoS
000418101400009
EID Scopus
2-s2.0-85035027807
BibTeX 
@ARTICLE{FITPUB12245,
   author = "Heikki K{\"{a}}lvi{\"{a}}inen and Tuomas Eerola and Lasse Lensu and Jarmo Ilonen and Pavel Zem\v{c}\'{i}k and Mark\'{e}ta Dubsk\'{a} and Roman Jur\'{a}nek",
   title = "Comparison of bubble detectors and size distribution estimators",
   pages = "60--66",
   journal = "Pattern Recognition Letters",
   volume = 101,
   number = 1,
   year = 2018,
   ISSN = "0167-8655",
   doi = "10.1016/j.patrec.2017.11.014",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/12245"
}
Back to top