Octahedron-based Projections as Intermediate Representations for Computer Imaging: TOAST, TEA, and More.

Area of Science:

• Astronomy and Astrophysics
• Computer Science
• Image Processing

Background:

• All-sky projections often contain singular points, complicating data handling.
• Dynamic transformation of imaging data requires flexible intermediate representations.

Purpose of the Study:

• To define and discuss novel rectangular all-sky projections without singular points.
• To introduce the Tesselated Octahedral Adaptive Spherical Transformation (TOAST) for efficient all-sky imaging.
• To highlight TOAST's suitability for hierarchical data and GPU computation.

Main Methods:

• Development of rectangular all-sky projections with no singular points.
• Utilizing the Tesselated Octahedral Adaptive Spherical Transformation (TOAST) as a key intermediate representation.
• Leveraging recursive computation for simple algorithm design.

Main Results:

• TOAST projections effectively represent all-sky imaging data without singular points.
• The method facilitates dynamic transformation between standardized and user-requested formats.
• TOAST demonstrates strong adaptability for hierarchical image structures and GPU acceleration.

Conclusions:

• TOAST provides a robust and efficient solution for all-sky data representation and transformation.
• Its hierarchical nature and simple recursive algorithm make it well-suited for modern imaging applications.
• The projection method is particularly advantageous for high-resolution, GPU-accelerated astronomical imaging.