Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Flame front tracking by laser induced fluorescence spectroscopy and advanced image analysis

Open Access Open Access

Abstract

This paper presents advanced image analysis methods for extracting information from high speed Planar Laser Induced Fluorescence (PLIF) data obtained from turbulent flames. The application of non-linear anisotropic diffusion filtering and of Active Contour Models (Snakes) is described to isolate flame boundaries. In a subsequent step, the detected flame boundaries are tracked in time using a frequency domain contour interpolation scheme. The implementations of the methods are described and possible applications of the techniques are discussed.

©2001 Optical Society of America

Full Article  |  PDF Article
More Like This
Nonlinear diffusion filtering of images obtained by planar laser-induced fluorescence spectroscopy

Henrik Malm, Gunnar Sparr, Johan Hult, and Clemens F. Kaminski
J. Opt. Soc. Am. A 17(12) 2148-2156 (2000)

Autonomous extraction of optimal flame fronts in OH planar laser-induced fluorescence images

Mark Sweeney and Simone Hochgreb
Appl. Opt. 48(19) 3866-3877 (2009)

Burst-mode OH/CH2O planar laser-induced fluorescence imaging of the heat release zone in an unsteady flame

Ulrich Retzer, Rongchao Pan, Thomas Werblinski, Franz J. T. Huber, Mikhail N. Slipchenko, Terrence R. Meyer, Lars Zigan, and Stefan Will
Opt. Express 26(14) 18105-18114 (2018)

Supplementary Material (3)

Media 1: MOV (202 KB)     
Media 2: MOV (1019 KB)     
Media 3: MOV (144 KB)     

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (7)

Fig. 1.
Fig. 1. Schematic setup for time resolved PLIF of turbulent spark ignition.
Fig. 2.
Fig. 2. An example showing a typical experimental image sequence. (a)-(d) Four images (381×291 pixels, 0.1408 mm/pixel) captured with 1.7 ms time increments respectively.
Fig. 3.
Fig. 3. Sample image corresponding to a single shot PLIF image of OH. To the left the raw data is shown. The same data is shown to the right after 30 iterations using the non-linear diffusion algorithm described in the text. In the lower section of the figure, cross-sectional profiles corresponding to the horizontal green line in the images are shown. Contours are clearly enhanced and local noise is efficiently filtered out. An animation of the diffusion process is presented in the related animation file nldf.mov (202 KB).
Fig. 4.
Fig. 4. An example illustrating the progress of the snake iterations: (a) The original raw image. (b) The initial snake (in red) applied on the non-linear diffusion filtered image. (c)-(e) The snake after 1, 25, and 85 iterations, respectively. (f) The final result overlaid on the original raw image. See the related animation file snake.mov (1 MB).
Fig. 5.
Fig. 5. Temporal interpolation: Three validation tests (a)-(c) using synthetic examples. Top: Original synthetic sequences comprising F=16 shapes (generated by evolving a shape in time using predetermined controlled deformations). Centre: Four original shapes extracted from the synthetic sequence to be input into the interpolation algorithm. These 4 curves can be seen overlaid on the top figures in magenta, red, blue and cyan. Bottom: The interpolation result (reconstruction of 16 frames from 4 only). (a) Elliptical shapes: Error=3.53%. (b) Star shapes: Error=5.25%. (c) Shapes based on deforming a real flame boundary: Error=0.75%.
Fig. 6.
Fig. 6. Temporal interpolation: (a-f) Different validation tests on synthetic examples. The original 4 contours are displayed in black and the interpolated ones in color.
Fig. 7.
Fig. 7. An example of temporal interpolation on PLIF data: (a) The flame contours in an image sequence. The four original contours are shown in thick black and the interpolated ones in between are shown in random colors. (b) The calculated velocity vectors overlaid on the original contours. See related animation file interp.mov (144 KB). (c) The color map representation of the flame velocity values. The image shows the magnitude (in m/s) of the flame front velocity at each boundary point in each frame using the colormap shown to the right.

Equations (11)

Equations on this page are rendered with MathJax. Learn more.

t I = div ( g ( ( G σ * I ) ) I )
v i ( n ) = v i 1 ( n ) + w 1 F i tens ( n ) + w 2 F i flex ( n ) + w 3 F i ext ( n ) + w 4 F i inf ( n )
F i flex ( n ) = 2 v i ( n ) v i ( n 1 ) v i ( n + 1 )
F i flex ( n ) = 2 F i tens ( n ) F i tens ( n 1 ) F i tens ( n + 1 )
F i inf ( n ) = F ( I s ( x i ( n ) , y i ( n ) ) ) n i ( n )
F ( I ( x , y ) ) = { + 1 , I ( x , y ) T 1 , otherwise
F i ext ( n ) = P ( x i ( n ) , y i ( n ) )
X ( k , j ) = w ( k ) n = 1 N x ( n , j ) cos π ( 2 n 1 ) ( k 1 ) 2 N
w ( k ) = { 1 N , k = 1 2 N , 2 k N
x ( n , j ) = w ( k ) k = 1 N X ( k , j ) cos π ( 2 n 1 ) ( k 1 ) 2 N
ε = A i A o A i A o A o
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All Rights Reserved