Median ellipse parameterization for robust measurement of fuel droplet size

Michael Greminger, Alison Hoxie

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The combustion properties of blended fuel combinations can be characterized by performing single droplet fuel combustion experiments. These combustion experiments are visualized using high speed image acquisition. Once the high speed images are obtained, the burn rate and other characteristics of combustion, such as the occurrence of microexplosions, can be characterized. Currently these quantities are either measured manually or are measured using automated software. However, the current software packages used for this task are limited in that they can only measure droplets that are elliptical in shape and manual corrections often have to be made to avoid significant errors in the measurement. An automated droplet tracking algorithm is presented that can automatically track droplet size without manual intervention due to its robustness to the presence of missing or extra edges in the images. In addition, the proposed method can track shapes more general than ellipses, which is required in order to track the droplet during microexplosions. The proposed algorithm starts by fitting ellipses to numerous five point subsets from the droplet edge data. The closed contour is parameterized by determining the median perimeter of the set of ellipses. The resulting curve is not an ellipse, allowing arbitrary closed contours to be parameterized. The performance of the proposed algorithm and the performance of existing algorithms are compared to a ground truth segmentation of the fuel droplet images. This comparison demonstrates that the median ellipse parameterization algorithm has improved performance for both area quantification and edge deviation.

Original languageEnglish (US)
Article number025304
JournalMeasurement Science and Technology
Volume27
Issue number2
DOIs
StatePublished - Dec 31 2015

Fingerprint

Ellipse
ellipses
Parameterization
parameterization
Droplet
Combustion
high speed
High Speed
computer programs
fuel combustion
ground truth
Closed
Image Acquisition
set theory
Perimeter
acquisition
Image acquisition
Software Package
Quantification
Experiment

Keywords

  • combustion
  • contour fitting
  • droplet size measurement
  • image processing
  • robust ellipse fitting

Cite this

Median ellipse parameterization for robust measurement of fuel droplet size. / Greminger, Michael; Hoxie, Alison.

In: Measurement Science and Technology, Vol. 27, No. 2, 025304, 31.12.2015.

Research output: Contribution to journalArticle

@article{bd3f3bd281c74062a27724a41501d7b3,
title = "Median ellipse parameterization for robust measurement of fuel droplet size",
abstract = "The combustion properties of blended fuel combinations can be characterized by performing single droplet fuel combustion experiments. These combustion experiments are visualized using high speed image acquisition. Once the high speed images are obtained, the burn rate and other characteristics of combustion, such as the occurrence of microexplosions, can be characterized. Currently these quantities are either measured manually or are measured using automated software. However, the current software packages used for this task are limited in that they can only measure droplets that are elliptical in shape and manual corrections often have to be made to avoid significant errors in the measurement. An automated droplet tracking algorithm is presented that can automatically track droplet size without manual intervention due to its robustness to the presence of missing or extra edges in the images. In addition, the proposed method can track shapes more general than ellipses, which is required in order to track the droplet during microexplosions. The proposed algorithm starts by fitting ellipses to numerous five point subsets from the droplet edge data. The closed contour is parameterized by determining the median perimeter of the set of ellipses. The resulting curve is not an ellipse, allowing arbitrary closed contours to be parameterized. The performance of the proposed algorithm and the performance of existing algorithms are compared to a ground truth segmentation of the fuel droplet images. This comparison demonstrates that the median ellipse parameterization algorithm has improved performance for both area quantification and edge deviation.",
keywords = "combustion, contour fitting, droplet size measurement, image processing, robust ellipse fitting",
author = "Michael Greminger and Alison Hoxie",
year = "2015",
month = "12",
day = "31",
doi = "10.1088/0957-0233/27/2/025304",
language = "English (US)",
volume = "27",
journal = "Measurement Science and Technology",
issn = "0957-0233",
publisher = "IOP Publishing Ltd.",
number = "2",

}

TY - JOUR

T1 - Median ellipse parameterization for robust measurement of fuel droplet size

AU - Greminger, Michael

AU - Hoxie, Alison

PY - 2015/12/31

Y1 - 2015/12/31

N2 - The combustion properties of blended fuel combinations can be characterized by performing single droplet fuel combustion experiments. These combustion experiments are visualized using high speed image acquisition. Once the high speed images are obtained, the burn rate and other characteristics of combustion, such as the occurrence of microexplosions, can be characterized. Currently these quantities are either measured manually or are measured using automated software. However, the current software packages used for this task are limited in that they can only measure droplets that are elliptical in shape and manual corrections often have to be made to avoid significant errors in the measurement. An automated droplet tracking algorithm is presented that can automatically track droplet size without manual intervention due to its robustness to the presence of missing or extra edges in the images. In addition, the proposed method can track shapes more general than ellipses, which is required in order to track the droplet during microexplosions. The proposed algorithm starts by fitting ellipses to numerous five point subsets from the droplet edge data. The closed contour is parameterized by determining the median perimeter of the set of ellipses. The resulting curve is not an ellipse, allowing arbitrary closed contours to be parameterized. The performance of the proposed algorithm and the performance of existing algorithms are compared to a ground truth segmentation of the fuel droplet images. This comparison demonstrates that the median ellipse parameterization algorithm has improved performance for both area quantification and edge deviation.

AB - The combustion properties of blended fuel combinations can be characterized by performing single droplet fuel combustion experiments. These combustion experiments are visualized using high speed image acquisition. Once the high speed images are obtained, the burn rate and other characteristics of combustion, such as the occurrence of microexplosions, can be characterized. Currently these quantities are either measured manually or are measured using automated software. However, the current software packages used for this task are limited in that they can only measure droplets that are elliptical in shape and manual corrections often have to be made to avoid significant errors in the measurement. An automated droplet tracking algorithm is presented that can automatically track droplet size without manual intervention due to its robustness to the presence of missing or extra edges in the images. In addition, the proposed method can track shapes more general than ellipses, which is required in order to track the droplet during microexplosions. The proposed algorithm starts by fitting ellipses to numerous five point subsets from the droplet edge data. The closed contour is parameterized by determining the median perimeter of the set of ellipses. The resulting curve is not an ellipse, allowing arbitrary closed contours to be parameterized. The performance of the proposed algorithm and the performance of existing algorithms are compared to a ground truth segmentation of the fuel droplet images. This comparison demonstrates that the median ellipse parameterization algorithm has improved performance for both area quantification and edge deviation.

KW - combustion

KW - contour fitting

KW - droplet size measurement

KW - image processing

KW - robust ellipse fitting

UR - http://www.scopus.com/inward/record.url?scp=84955444115&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84955444115&partnerID=8YFLogxK

U2 - 10.1088/0957-0233/27/2/025304

DO - 10.1088/0957-0233/27/2/025304

M3 - Article

AN - SCOPUS:84955444115

VL - 27

JO - Measurement Science and Technology

JF - Measurement Science and Technology

SN - 0957-0233

IS - 2

M1 - 025304

ER -