Michael Knapp
Memory Allocation Strategies for Large Volumetric Data-Sets
[Thesis]

Information

Abstract

Since the development of medical three dimensional imaging devices in the 1970s, volumetric data processing has tremendously gained in importance. With the growing size of the data-sets, exhausting the capabilities of the hardware of its time, methods for efficient volumetric data processing have been always a hot topic. In this diploma thesis two approaches for processing large volumetric data-sets are presented. Both approaches utilize a block volume for storing the data. Further data compression and out-of-core processing are incorporated. Efficiency is achieved by processing only the required portion of data while omitting the non-related data having no effect on the intended result of the algorithm. This is supported by utilization of the knowledge about access patterns of the algorithm. Also methods for optimizing the efficiency by exploiting architectural properties of the computer hardware are presented.

Additional Files and Images

Additional images and videos

image
image

Additional files

Thesis
Thesis

Weblinks

No further information available.

BibTeX

@mastersthesis{knapp_michael_2004_MAS,
  title =      "Memory Allocation Strategies for Large Volumetric Data-Sets",
  author =     "Michael Knapp",
  year =       "2004",
  abstract =   "Since the development of medical three dimensional imaging
               devices in the 1970s, volumetric data processing has
               tremendously gained in importance. With the growing size of
               the data-sets, exhausting the capabilities of the hardware
               of its time, methods for efficient volumetric data
               processing have been always a hot topic. In this diploma
               thesis two approaches for processing large volumetric
               data-sets are presented. Both approaches utilize a block
               volume for storing the data. Further data compression and
               out-of-core processing are incorporated. Efficiency is
               achieved by processing only the required portion of data
               while omitting the non-related data having no effect on the
               intended result of the algorithm. This is supported by
               utilization of the knowledge about access patterns of the
               algorithm. Also methods for optimizing the efficiency by
               exploiting architectural properties of the computer hardware
               are presented.",
  month =      dec,
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Institute of Computer Graphics and Algorithms, Vienna
               University of Technology ",
  keywords =   "volume processing, large data",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2004/knapp_michael_2004_MAS/",
}