Anti-aliasing method based on rotated spatial filtering
Lázaro Ribeiro Monteiro Júnior,Leandro Carlos de Souza,Sílvio Roberto Fernandes
DOI: https://doi.org/10.1007/s11042-024-18632-y
IF: 2.577
2024-02-29
Multimedia Tools and Applications
Abstract:Due to the evolution of graphics processors over recent decades it has now become possible to produce high-quality, realistic three-dimensional scenes. However, aliasing occurs during the sampling process performed during rasterization, which causes a serrated effect around the edges of the objects presented in the scene, highlighting an unreal aspect of the image and causing displeasure to the viewer. This paper aims to develop an anti-aliasing treatment based on rotated spatial filtering. It undertakes edge detection by applying spatial filtering with a simple linear regression technique. A smoothing spatial filter is then rotated to match the direction of the inspected edge and is applied to the affected regions. Testing was performed on an OpenGL application, processing the rendered image from the framebuffer, and Blender, a 3D modeling software that enables scenes with more complex graphics. The results have demonstrated the effectiveness of the proposed method in smoothing aliasing with good quality while preserving the details of the scene. Hence, the problem was managed effectively with a post-filtering approach and without oversampling. The running time of the algorithm is O(n), and the memory used is O(n).
computer science, information systems, theory & methods,engineering, electrical & electronic, software engineering