In this paper, we describe a method to create an approximate ray-traced stereoscopic pair by transforming a fully raytraced left-eye view into an inferred righteye view. Performance results from evaluating several random scenes, which indicate that the second view in a stereoscopic image can be computed with as little as 5% of the effort required to fully ray-trace the first view, are presented. We also discuss worst-case performance of our algorithm and demonstrate that our technique is always at least as efficient as two passes of a standard ray-tracer.
Currently almost all computer graphic stereoscopic images are generated by doubling the work required to create a single image. In this paper we derive and analyze algorithms for simultaneous generation of the two views necessary for a stereoscopic image. We begin with a discussion of the similarities of the two perspective views of a stereo pair. Following this, several graphics algorithms that have been optimized from known single-view methods are described and performance results obtained from testing the new stereo algorithms against the originals are presented.
One of the drawbacks of standard volume rendering techniques is that it is often difficult to comprehend the three-dimensional structure of the volume from a single frame; this is especially true in cases where there is no solid surface. Generally, several frames must be generated and viewed sequentially, using motion parallax to relay depth. Another option is to generate a single stereoscopic pair, resulting in clear and unambiguous depth information in both static and moving images.Methods have been developed which take advantage of the coherence between the two halves of a stereo pair for polygon rendering and ray-tracing, generating the second half of the pair in significantly less time than that required to completely render a single image. This paper reports the results of implementing these techniques with parallel ray-traced volume rendering. In tests with different data types, the time savings is in the range of 70 -80%.
Images AbstractRay-tracing is a well-known method for producing realistic images. If we wish to view a ray-traced image stereoscopically, we must create two distinct views of the image: a left-eye view end a right-eye view. The most straight-forward way to do this is to ray-trace both views, doubling the required work for a single perspective image. We have developed a reprojection algorithm that produces stereoscopic images efficiently with little degradation in image quality. In this paper, we derive the necessary stereoscopic separation technique required to transform a ray-traced left-eye view into an inferred right-eye view. Minor changes in the technique ne~J_ed to expand from single to multiple rays per pixel are noted. Finally, results from evaluating several random scenes are given, which indicate that the second view in a stereo image can be computed with less than 16 percent of the effort of computing the first.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.