Ray casting sparse level sets

Abstract

This paper presents an interactive rendering method for sparse level sets. A GPU-CPU hybrid ray casting approach is developed to perform a surface extraction of a level set embedded in a signed distance field. The sparseness of the data structure is exploited to allow object-order empty-space skipping. The signed distance nature of the dataset is used to naturally provide adaptive sampling which leads to high quality surface reconstruction. Ray surface intersections are rendered as oriented points with smooth normals computed directly from the dataset. Alternative shading methods are possible after the intersections are found, however, we explicitly use only normals as the main purpose is to have this method work interactively with other similarly shaded objects in the same scene.

A Shader-Based Ray Tracing Engine

“…Ray casting is the fundamental steps for performing the ray-object interactions. It is actually a procedure of ray collision checking in the virtual space [21,22]. The simplest form of ray casting is to compare all the primitives (or graphics objects) against the ray, taking the time complexity of O(n) against n objects in the scene.…”

Efficacy of Images Versus Data Buffers: Optimizing Interactive Applications Utilizing OpenCL for Scientific Visualization

“…Slightly more recent work includes a summary of techniques usable in real time volume rendering [6] and advanced illumination methods for volume rendering [7]. Recent work in ray casting techniques (most of the early volume rendering was back projected ray casting) includes Lux and Fröhlich [8] and Zhu et al [9]. Early works use texture GPU memory in a manner that presages the algorithm described here, though our work uses general purpose image buffers; our data is also streamed.…”