TStar
View on GitHub
TStar stands out as a powerful image processing platform based on C++17 and Qt6 architecture, explicitly designed to meet the needs of modern astrophotography and produce scientific-grade results. The software offers an extremely flexible MDI (Multi-Document Interface) environment that ensures full compatibility with 8, 16, and 32-bit integer and floating-point FITS/XISF/TIFF files, allowing total control over linear raw data.
The heart of TStar's workflow lies in its ability to transform linear signals into visible images with extreme precision: thanks to cutting-edge tools such as Generalized Hyperbolic Stretch (GHS) and ArcSinh Stretch, it is possible to independently manage shadows, midtones, and highlights, preserving saturation and fine details. Color fidelity is ensured by Photometric Color Calibration (PCC), which astrometrically solves the image and calibrates colors based on Gaia and APASS star catalogs, supported by effective tools for removing light gradients and background casts.
A distinctive aspect of TStar is the native integration of AI-based technologies for image restoration. The Cosmic Clarity module uses Deep Learning for noise reduction and sharpening, while compatibility with StarNet++ and GraXpert facilitates complex operations such as star removal and advanced gradient extraction. Completing the technical picture are advanced features for narrowband composition (with customizable artistic palettes), a powerful PixelMath engine for arithmetic operations between images, and a sophisticated masking system that allows isolation of luminance, chrominance, or specific stellar features for high-precision localized interventions.