🔭 astrophysics

Quantifying Symmetry: Transformation Information for Planetary Nebulae and Supernova Remnants

This paper introduces a quantitative, non-parametric pipeline based on Transformation Information (TI) to identify and classify geometric symmetries in astrophysical images, successfully applying it to reveal morphological structures in planetary nebulae and distinguish Type Ia from core-collapse supernova remnants.

Original authors: Dmitry Shishkin, Amir Michaelis

Published 2026-06-04

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Dmitry Shishkin, Amir Michaelis

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you have a messy pile of puzzle pieces, and you want to know if they actually form a picture with a hidden pattern, like a star or a flower. You could try rotating the pile or flipping it over to see if the pieces line up better. If they do, you've found a "symmetry."

This paper introduces a new, super-smart way to do that puzzle test, but instead of using your eyes, it uses math to measure "how well things match" for thousands of different angles. The authors call this method Transformation Information (TI).

Here is a breakdown of what they did and found, using simple analogies:

1. The Core Idea: The "Mirror and Spin" Test

The authors treat an image of space (like a glowing cloud of gas) as a piece of paper. They ask two questions:

The Spin Test: If I rotate this picture by 10 degrees, then 20 degrees, then 30 degrees... at what angle does it look most like the original?
The Mirror Test: If I flip this picture over a line (like a reflection in a pond), at what angle does the reflection look most like the original?

They use a mathematical tool called KL Divergence (think of it as a "mismatch meter") to measure the difference between the original and the rotated/flipped version.

High Mismatch: The picture looks very different.
Low Mismatch (The "Valley"): The picture looks almost identical.

When they find a "valley" (a low point) in their mismatch meter, they know they've found a symmetry axis. It's like finding the perfect spot to fold a piece of paper so the edges line up perfectly.

2. Testing the Tool: The "Wind Rose"

Before looking at space, they tested their tool on a fake image of a wind rose (a compass-like shape with big spikes and small spikes).

The Result: The tool perfectly found the angles where the big spikes lined up with other big spikes, and the small spikes lined up with small spikes. It even noticed that some matches were "better" (deeper valleys) than others. This proved their math works.

3. Looking at Planetary Nebulae: The "Cosmic Flowers"

Planetary Nebulae are glowing shells of gas left behind by dying stars. Many of them look like flowers or hourglasses.

The Application: The authors applied their tool to images of these "cosmic flowers."
The Result: The tool successfully found the "spine" of these shapes. For example, it found the vertical line that splits a butterfly-shaped nebula in half, or the horizontal line that matches the top and bottom. It confirmed what astronomers already suspected: these shapes often have specific, repeating patterns.

4. Looking at Supernova Remnants: The "Cosmic Shrapnel"

Supernova remnants are the messy, chaotic debris left after a star explodes. They are much less organized than the "flowers."

The Challenge: These images are like a pile of shattered glass. It's hard to see a pattern.
The "Silhouette" Trick: To handle the mess, the authors tried a new version of their tool. Instead of looking at the brightness of every pixel (the "grayscale" version), they turned the image into a simple black-and-white silhouette (like a shadow puppet).
- Analogy: Imagine trying to recognize a person's face by looking at the details of their skin (hard if they have a scar) versus just looking at the outline of their head (easier).
The Result: By looking at the "shadow" of the explosion, the tool could find symmetry axes that the detailed version missed. It highlighted the overall shape and the big "ears" or "tails" sticking out, ignoring the messy details inside.

5. Sorting the Explosions: "The Neat vs. The Messy"

The authors wanted to see if they could tell the difference between two types of stellar explosions just by looking at their symmetry:

Type Ia (The "Neat" Ones): These come from white dwarf stars. They tend to be very round and uniform.
Core-Collapse (The "Messy" Ones): These come from massive stars collapsing. They tend to be lopsided and irregular.

The Discovery:
They created a "Symmetry Score" based on how deep and narrow the "valleys" were in their mismatch meter.

Type Ia remnants had "shallow, wide valleys." This means they are generally symmetrical in many directions, but not perfectly sharp in any single one (like a slightly squashed ball).
Core-Collapse remnants had "deep, narrow valleys." This means they are very asymmetrical overall, but they have specific, sharp features (like a big protrusion) that line up perfectly at one specific angle.

The Conclusion:
Using this score, the tool successfully separated the "Neat" explosions from the "Messy" ones into two distinct groups, proving that their shapes are fundamentally different.

Summary

The paper presents a new, automated way to measure "how symmetrical" a picture of space is. It works like a super-precise protractor and mirror combined.

It works on simple shapes and complex "cosmic flowers."
It can look at the "shadow" of an object to find patterns in messy explosions.
It can mathematically prove that two different types of stellar explosions look different, even if they both look like random blobs to the naked eye.

The authors emphasize that this is a tool for classification and measurement, not a way to predict the future or change how we treat diseases. It simply helps astronomers describe the shapes of the universe more accurately.

Technical Summary: Quantifying Symmetry via Transformation Information in Astrophysical Images

Problem Statement
Astrophysical systems, ranging from planetary nebulae (PNe) to supernova remnants (SNRs), exhibit diverse morphologies shaped by underlying physical mechanisms. While symmetry is a fundamental organizing principle for classifying these structures and inferring their formation dynamics (e.g., binary interactions, jet launching, or explosion mechanisms), quantitative identification of symmetry remains challenging. Existing methods often rely on subjective visual classification or neural networks that lack explicit, reproducible symmetry metrics. The authors posit a need for a non-parametric, data-driven framework to quantify symmetry (or symmetry breaking) to distinguish between different remnant populations, such as Type Ia and core-collapse supernovae.

Methodology: Transformation Information (TI)
The paper introduces a quantitative symmetry-identification pipeline based on Transformation Information (TI), an information-theoretic measure of self-similarity under geometric transformations.

Core Formalism: The method treats an image $I(x)$ as a probability distribution (after normalization) and calculates the Kullback-Leibler (KL) divergence between the original image and its transformed counterpart $T_a I(x)$ (where $T_a$ represents rotation or reflection) over their overlapping domain.
$TI(a) = \frac{1}{|\tilde{D}_\alpha|} \int_{\tilde{D}_\alpha} I(x) \ln \left[ \frac{I(x)}{T_a I(x)} \right] dA$
Local minima in the TI curve as a function of transformation parameters (rotation angle $\theta$ or reflection axis $\theta$ ) correspond to symmetry operations. The depth of the minimum quantifies the strength of the symmetry.
Implementation: The authors implemented the pipeline in Python using SciPy and OpenCV. Images are converted to grayscale to aggregate multi-wavelength emission. The system scans rotation angles ( $0^\circ$ to $360^\circ$ ) and reflection axes ( $0^\circ$ to $180^\circ$ ).
Thresholded Variant: To emphasize global morphology and outline-driven symmetries while suppressing internal intensity variations, a two-level (binary) variant is introduced. The image is thresholded into a binary silhouette, and the TI measure is applied to these silhouettes. This variant is particularly sensitive to edge features and protrusions.
Quantitative Descriptor: To compare populations, the authors define a minima prominence-to-width score ( $R_p$ ). This metric characterizes the "significance" of a detected symmetry axis, where a high score indicates a sharp, deep minimum (strong, localized symmetry) and a low score indicates a shallow, broad minimum (weak or diffuse symmetry).

Key Results

Validation: The method was first validated on a synthetic "wind-rose" image with known rotational and reflectional symmetries. The TI curves successfully recovered the correct angles, with deeper minima corresponding to exact matches (e.g., large spikes overlapping large spikes) and shallower minima corresponding to partial matches.
Planetary Nebulae (PNe): Applied to a set of PNe (including M1-37, MyCn 18, NGC 6543, NGC 5307, and NGC 7293), the pipeline successfully identified symmetry axes consistent with known bipolar, multipolar, and point-symmetric morphologies. For example, in M1-37, the method recovered axes corresponding to the split lobes and the overall bipolar structure, matching previous qualitative classifications.
Supernova Remnants (SNRs):
- Intensity-Based Analysis: Applied to SNRs (Vela, N132D, S147, N63A, G321.3-3.9), the method identified symmetry axes associated with protrusions, bright rims, and substructures. However, the authors note that for SNRs, the lack of established "ground truth" symmetry axes makes definitive physical interpretation more difficult than for PNe.
- Thresholded Analysis: The binary variant successfully identified additional symmetry axes in N63A that correlated with specific protrusions (e.g., "Ear" features) not as prominent in the intensity-based analysis, demonstrating its utility for analyzing global shape.
Population Separation (Type Ia vs. Core-Collapse):
- The authors applied the TI framework to a sample of 24 SNRs (10 Type Ia, 14 Core-Collapse) from Lopez et al. (2011).
- While the number of detected symmetry axes did not significantly differ between the two groups, the minima prominence-to-width score ( $R_p$ ) successfully separated them into distinct populations.
- Type Ia (G1): Characterized by more uniform, spherical morphologies, resulting in TI profiles with low, broad minima (low $R_p$ scores).
- Core-Collapse (G2): Characterized by irregular, less standard morphologies, resulting in TI profiles with sharper, deeper, and more localized minima (higher $R_p$ scores).
- Exponential fits to the distribution of these scores confirmed the statistical separation of the two populations.

Significance and Claims
The paper claims to provide a non-parametric, reproducible framework for symmetry identification that does not rely on pre-defined geometric models. Its primary contributions are:

Quantitative Morphology: It transforms qualitative visual assessments of symmetry into quantitative metrics (TI curves and $R_p$ scores).
Population Studies: It demonstrates that symmetry strength, quantified by the prominence-to-width of TI minima, can serve as a compact descriptor to distinguish between Type Ia and core-collapse supernova remnants, supporting the hypothesis that their formation channels produce morphologically distinct populations.
Complementary Analysis: The introduction of the thresholded binary variant offers a complementary tool to intensity-based analysis, specifically useful for isolating global outlines and edge features in complex, filamentary structures like SNRs.

The authors maintain a modest stance regarding the interpretation of individual SNR axes, acknowledging that without ground-truth symmetry definitions for many remnants, the physical link between a detected axis and a specific feature requires further feature-aware refinement. However, they assert the method's robustness for statistical population studies and its ability to reproduce known morphological classifications in PNe.