Background Star Elimination

Author: Tianjiang Says

Website: https://cislunarspace.cn

Definition

Background star elimination is the process of removing background stellar signals from astronomical image frames during image processing. Sun et al. (2026) state that "background stars are removed from the frames, as their presence affects the SAA (stacking search algorithm) efficiency."

Core Principles

How Stellar Signals Interfere with Detection

In astronomical survey images, background stars occupy large pixel regions. When using the stacking search algorithm (SAA) to detect moving objects, stellar signals cause the following problems:

  1. Spurious correlation peaks: high-brightness stellar pixels may produce false signals under specific velocity assumptions
  2. Dynamic range compression: strong stellar signals may mask faint moving targets in the same region
  3. Wasted computation: the algorithm must process large numbers of stellar pixels unrelated to moving targets

Segmentation Map Masking

The core method for background star elimination uses a segmentation map generated by SExtractor as a mask. The specific steps are:

  1. Perform source detection on the image using SExtractor, generating a segmentation map
  2. Mark all nonzero pixels in the segmentation map (i.e., pixels belonging to stars) as mask regions
  3. "All pixel values of the detected stars are set to zero"

Criticality of Threshold Selection

The effectiveness of background star elimination is highly dependent on threshold selection, which requires careful balancing:

  • Risk of high thresholds: "a high threshold may fail to effectively remove the connected pixels," leaving residual stellar signals at the edges that affect subsequent processing
  • Risk of low thresholds: "a low threshold may inadvertently exclude faint sources," meaning genuine moving object signals may be mistakenly classified as stars and eliminated

The ideal threshold should maximally remove stellar signals while preserving the signal integrity of moving objects to the greatest extent possible.

Applications in Cislunar Observation

In optical surveys of cislunar moving objects, background star elimination is a critical linking step in the image processing pipeline. Sun et al. (2026) place it after image registration and before the stacking search algorithm, forming the following processing chain:

Raw Images → Image Registration → Background Star Elimination → Stacking Search Algorithm (SAA) → Target Detection

For cislunar observation, background star elimination faces several unique challenges:

  • Dense star fields: when the observation direction is near the galactic plane, stellar density is extremely high, increasing elimination difficulty
  • Lunar scattered light: within the lunar glare zone, the diffuse background from lunar surface scattering complicates star detection
  • Faint target protection: cislunar moving objects are extremely faint, requiring extremely fine threshold control to avoid accidental elimination

References

  • Sun, R., Zhang, Q., Yu, S., et al. Optical Survey for Cislunar Moving Objects Using Image Stacking. AJ, 2026.