Anti-aliasing on deformed area using adaptive super sampling during volume ray-casting

Abstract

Purpose In modern GPUs, deformation has been possible to perform direct space warping in texture-based rendering and GPU-based ray-casting. The problem of direct space warping is lead to provide aliasing artifacts on deformed area. The variation of sampling interval in original space and deformed space is cause of the aliasing. Methods To reduce this artifact, we apply the adaptive super sampling method to deformed area. We divide view plane into several areas and applied different super sampling kernel to each detected aliased region. Firstly, we decide the deformed area using deformation mask which includes transformed position. Secondly, we select a boundary region in the deformed area. Aliasing artifact mostly appears in a boundary of an object. Then we execute super sampling on the intersection of deformed and boundary regions. Results We reduce rendering time which is cause of additional super sampling operation by selecting the target area of view plane. Our method has been about 20% faster than the fully super sampled method since we reduced the super sampling region in the view plane. Conclusions Volume deformation is commonly used in visualization techniques especially in medical volume simulation field. In this paper, we propose an adaptive super sampling method on deformed area. It is possible to visualize enhanced result image only as spending a little additional rendering time.

Efficient Massive Computing for Deformable Volume Data Using Revised Parallel Resampling

“…This issue has been addressed in the literature. When only a portion of the entire volume data was deformed, the screen area corresponding to the deformed portion was calculated in advance, and supersampling was applied to that area during rendering in [8]. In [9], an efficient approach to decompose the entire volume unequally was proposed.…”