The authors propose a new mapping technique for volume rendering on massively parallel computer. Previous mappings of volume rendering algorithms onto array processors either required large amounts of interprocessor c...
详细信息
The authors propose a new mapping technique for volume rendering on massively parallel computer. Previous mappings of volume rendering algorithms onto array processors either required large amounts of interprocessor communication or lacked the generality needed to deal with arbitrary rotations and perspective projections. The mapping developed here is based on an enhanced ray-tracking algorithm. Instead of tracing rays into volumeric data, each processor determines the samples of a local slice through which a ray intersects. The samples are collected and interpolated in such a way that optimal systolic design is guaranteed. This mapping algorithm achieves optimal pipeline design for back-to-front projections, handles arbitrary viewing angles and allows orthographic as well as perspective projections.< >
Fragmented objects are encountered in a variety of diverse engineering and scientific fields including industrial inspection, customized medical prosthesis design, forensic science, paleontology, and archaeology. The ...
详细信息
Fragmented objects are encountered in a variety of diverse engineering and scientific fields including industrial inspection, customized medical prosthesis design, forensic science, paleontology, and archaeology. The arbitrarily broken pieces must be reassembled and new material often added to complete the process of shape reconstruction. To prevent physical damage of the pieces during reconstruction and enhance shape visualization scientists have begun to exploit 3D data acquisition and graphical modeling tools. An algorithm for enabling free-form shape reconstruction from digitized data of fragmented pieces is described in this paper. The method exploits the topological structure and learning algorithm of a 3D self-organizing feature map (SOFM). The lattice of the SOFM is a spherical mesh that maintains the relative connectivity of the neighboring nodes as it transforms under external forces. The weight nodes of the lattice represent vertices of the constituent elements in the facetted surface model. The technique is illustrated by reconstructing two clay objects with closed geometries from several fragmented parts
The conference materials contain 59 papers. The main topics covered include volume segmentation, feature analysis, multimodality registration, image processing and transforms, rendering and interpretation, visualizati...
详细信息
ISBN:
(纸本)081941008X
The conference materials contain 59 papers. The main topics covered include volume segmentation, feature analysis, multimodality registration, image processing and transforms, rendering and interpretation, visualization tools, visualization systems, surgery and treatment planning, diagnosis and image interpretation, biology and functional imaging.
Experiences during the investigation of parallel methods for faster isosurface generation on SIMD (single instruction stream, multiple data stream) machines are described. A sequential version of a well-known isosurfa...
详细信息
ISBN:
(纸本)9780818628962
Experiences during the investigation of parallel methods for faster isosurface generation on SIMD (single instruction stream, multiple data stream) machines are described. A sequential version of a well-known isosurfacing algorithm is algorithmically enhanced for a particular type of SIMD architecture. The SIMD implementation takes full advantage of the data parallel nature of the algorithm, and experiments have proven the implementation to be highly scalable. A parallel tool, which can generate 170 K polygons/s, gives scientists the means to explore large 3D scalar or vector fields interactively.< >
The proceedings contains 69 papers. The following topics are dealt with: volume visualization;biomedical applications;models of visualization;computer vision in visualization;chaos and fractals in electroencephalograp...
详细信息
ISBN:
(纸本)0818620390
The proceedings contains 69 papers. The following topics are dealt with: volume visualization;biomedical applications;models of visualization;computer vision in visualization;chaos and fractals in electroencephalography;artificial intelligence;visual perception;human-machine interfaces;systems and approaches;visualization in the neurosciences;mathematical and computational models;and general applications and visualization in medical educations.
This paper examines the role of experimental design, data acquisition equipment, and system integration in the holistic solution picture. Issues include data formats, distributed computing environments, and the need f...
详细信息
ISBN:
(纸本)9780818622458
This paper examines the role of experimental design, data acquisition equipment, and system integration in the holistic solution picture. Issues include data formats, distributed computing environments, and the need for truly interactive, even real-time systems. A major theme is reaching beyond volume visualization to 'volume comprehension' through volume segmentation, mensuration, and geometry extraction.< >
A set of volume visualization tools that are based on the use of recursive ray tracing as the primary vehicle for realistic volume imaging is presented. The tools include shadows, mirrors, specularity, and constructiv...
详细信息
ISBN:
(纸本)9780818622458
A set of volume visualization tools that are based on the use of recursive ray tracing as the primary vehicle for realistic volume imaging is presented. The tools include shadows, mirrors, specularity, and constructive solid geometry. The underlying representation for the ray tracer is a 3-D raster of voxels that holds the discrete form of the scene. Unlike traditional volume rendering techniques, the discrete recursive ray tracer models many illumination phenomena by traversing discrete rays in voxel space. The approach provides true ray tracing of sampled or computed datasets, as well as ray tracing of hybrid scenes where sampled or computed data are intermixed with geometric models and enhances the understanding of complex biomedical datasets.< >
暂无评论