platesolving

Plate solving, also referred to as platesolving or plate solving, is a computational technique used in astronomy to determine the exact orientation and pointing of a digital astronomical image by matching observed stars to a celestial catalog.

The process typically begins with source extraction to detect star-like objects in the image, followed by astrometric solving that converts pixel coordinates to sky coordinates. A solver compares the measured star pattern against a large index of cataloged star patterns, often using robust geometric hashing or triangulation, to find a unique transformation that maps image coordinates to right ascension and declination and derives the plate scale, rotation, and, if needed, optical distortions.

There are two main modes: blind solving, which requires no initial pointing information, and approximate-solved approaches that use a rough pointing estimate to narrow the search. Open-source and commercial solvers, such as astrometry.net, are widely used and rely on precomputed pattern indices to achieve fast, reliable solutions.

Plate solving is used to add world coordinate system information to images, enabling automatic telescope pointing during acquisitions, accurate cross-matching with catalogs, and improved data reduction for imaging surveys.

Limitations include dependence on star density, image quality, field-of-view, and catalog completeness. In crowded fields or with poor SNR, solving may fail or yield ambiguous results. Solutions that produce a WCS enable subsequent analysis and interoperability.