(a)
(b)
(c)
Andrei Lintu, MPI Informatik
Lars Hoffmann, MPI Informatik
Sascha El-Abed, MPI Informati
Ting-Hui Lee, National Optical Astronomy Observatory, USA
Marcus Magnor, TU Braunschweig
Hendrik P. A. Lensch, MPI Informatik
Hans-Peter Seidel, MPI Informatik
Abstract
This paper addresses the problem of reconstructing the 3D structure of planetary nebulae from 2D observations. Assuming axial symmetry, our method jointly reconstructs the distribution of dust and ionized gas in the nebulae from observations at two different wavelengths. In an inverse rendering framework we optimize for the emission and absorption densities which are correlated to the gas and dust distribution present in the nebulae. First, the density distribution of the dust component is estimated based on an infrared image, which traces only the dust distribution due to its intrinsic temperature. In a second step, we optimize for the gas distribution by comparing the rendering of the nebula to the visible wavelength image. During this step, besides the emission of the ionized gas, we further include the effect of absorption and scattering due to the already estimated dust distribution. Using the same approach, we can as well start with a radio image from which the gas distribution is derived without absorption, then deriving the dust distribution from the visible wavelength image considering absorption and scattering. The intermediate steps and the final reconstruction results are visualized at real-time frame rates using a volume renderer. Using our method we recover both gas and dust density distributions present in the nebula by exploiting the distinct absorption or emission parameters at different wavelengths.
|
|
|
|
(a) |
(b) |
(c) |
(a) Reflection and Planetary Nebulae such as the M2-9, often exhibit some axial symmetry due to their creation process. (b) By exploiting this axial symmetry it is sufficient to reconstruct volume densities for one 2D slice in order to create a 3D volume. (c) Dust particles have wavelength dependent scattering and absorption coefficients. Give multi-wavelength input images this can be exploited to reconstruct both the dust distribution and the distribution of the emissive ionized gas in the nebulae. |
| 
| 
|
(a) |
(b) |
(c) |
(a) Input image for the M3-35 planetary nebula, corresponding to H_alpha wavelength. (b) Recovered gas density map. (b) Rendering of the reconstruction. Some deviations are visible because the input is not perfectly symmetric. |
Publications:
Andrei Linţu, Lars Hoffmann, Marcus Magnor, Hendrik P. A. Lensch, Hans-Peter Seidel.
Andrei Linţu, Hendrik P. A. Lensch, Marcus Magnor, Sascha El-Abed,
Hans-Peter Seidel.
3D Reconstruction of Emission and Absorption in Planetary Nebulae . In
Procedeeings of IEEE/EG International Symposium on Volume Graphics,
2007, pages 9-16.
Andrei Linţu, Hendrik P. A. Lensch, Marcus Magnor, Ting-Hui Lee, Sascha El-Abed, Hans-Peter Seidel.
Multi-wavelength-based Method to de-project Gas and Dust Distributions of several Planetary Nebulae.
In Procedeeings of Asymmetrical Planetary Nebulae IV, 2007, pages 1-6.
Video
The Nebula NGC1999
recontructed from a single image (6MB)