Computer Science
8 years ago

Noise-free image relighting set to transform digital media experiences

IET Image Processing
Noise-resistant hemispherical basis for image-based relighting
Ping-Man Lam, Chi-Sing Leung, Tien‐Tsin Wong

Paper Summary

Related papers

Theme background image
Physics and Astronomy
8 years ago

New basis functions revolutionize analysis of scattering from large targets!

Electromagnetic Scattering
Metasurfaces
Finite-Difference Time-Domain
Theme background image
Physics and Astronomy
8 years ago

New triangular patch basis speeds up moment method convergence for better simulations.

Electromagnetic Scattering
Finite-Difference Time-Domain
Fracture Mechanics
Theme background image
Computer Science
8 years ago

New hemispherical basis revolutionizes rendering for accurate global illumination.

Computer Graphics
Stereo Vision
Shape Analysis

A new method called eigen hemispherical harmonic (EHSH) basis has been developed to improve image relighting. This method helps control the size of coefficients, reducing noise and artifacts in relit images. The EHSH basis has the same accuracy as the commonly used spherical harmonic (SH) basis. The researchers also discuss how to convert SH coefficients to EHSH coefficients and demonstrate how to relight images using the EHSH basis.