🧠 Origins of Weighting and Rejection Algorithms - tinystork/zemosaic GitHub Wiki

🧠 Origins of Weighting and Rejection Algorithms

The noise variance weighting (1/σ²) and Kappa-Sigma rejection algorithms implemented in ZeMosaic are directly inspired by the work of Juan Conejero, lead developer of PixInsight and founder of Pleiades Astrophoto.

🔬 Main Inspirations:

Noise Variance Weighting (1/σ²) Weighting images based on the inverse of their background noise variance gives less influence to noisier frames and more to cleaner ones. This approach is robust and particularly useful when combining datasets with variable quality.

Kappa-Sigma Clipping

A well-established statistical rejection method that removes outlier pixels (cosmic rays, satellites, or artifacts) using an iterative standard deviation-based threshold.

These techniques are extensively described in the PixInsight documentation, particularly in the ImageIntegration section:

📖 PixInsight 1.6.1 – New ImageIntegration Features

Juan Conejero, 2010 https://pixinsight.com/forum/index.php?threads/pixinsight-1-6-1-new-imageintegration-features.2107

🙏 Acknowledgement

The ZeMosaic project respectfully acknowledges the outstanding contributions of Juan Conejero, whose pioneering work has shaped much of the modern thinking around astronomical image integration and mosaicking.

While ZeMosaic does not aim to replicate all of PixInsight’s advanced capabilities, it draws clear inspiration from its core principles — adapted for accessibility and optimized for all-in-one sensors like the Seestar S50.