Segmentation Assisted Object Distinction For Direct Volume Rendering

Ray casting is a direct volume rendering technique for visualizing 3D arrays of sampled data. It has vital applications in medical and biological imaging. Nevertheless, it is inherently open to cluttered classification results. It suffers from overlapping transfer function values and lacks a suff...

Full description

Saved in:

Bibliographic Details
Main Author:	Irani, Arash Azim Zadeh
Format:	Thesis
Language:	English
Published:	2013
Subjects:	QA75.5-76.95 Electronic computers Computer science
Online Access:	http://eprints.usm.my/43239/1/ARASH%20AZIM%20ZADEH%20IRANI24.pdf
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my-usm-ep.43239
record_format	uketd_dc
spelling	my-usm-ep.432392019-04-12T05:26:13Z Segmentation Assisted Object Distinction For Direct Volume Rendering 2013-06 Irani, Arash Azim Zadeh QA75.5-76.95 Electronic computers. Computer science Ray casting is a direct volume rendering technique for visualizing 3D arrays of sampled data. It has vital applications in medical and biological imaging. Nevertheless, it is inherently open to cluttered classification results. It suffers from overlapping transfer function values and lacks a sufficiently powerful voxel parsing mechanism for object distinction. In this research work, we are proposing an image processing based approach towards enhancing ray casting technique’s object distinction process. The ray casting architecture is modified to accommodate object membership information generated by a K-means based hybrid segmentation algorithm. Object membership information is assigned to cubical vertices in the form of ID tags. An intra-object buffer is devised and coordinated with inter-object buffer, allowing the otherwise global rendering module to embed multiple local (secondary) rendering processes. A local rendering process adds two advantageous aspects to global rendering module. First, depth oriented manipulation of interpolation and composition operations that lead to freedom of interpolation method choice based on the number of available objects in various volumetric depths, improvement of LOD (level of details) for desired objects and reduced number of required mathematical computations. Second, localization of transfer function design that enables the utilization of binary (non-overlapping) transfer functions for color and opacity assignment. A set of image processing techniques are creatively employed in the design of K-means based hybrid segmentation algorithm. 2013-06 Thesis http://eprints.usm.my/43239/ http://eprints.usm.my/43239/1/ARASH%20AZIM%20ZADEH%20IRANI24.pdf application/pdf en public phd doctoral Universiti Sains Malaysia Pusat Pengajian Sains Komputer
institution	Universiti Sains Malaysia
collection	USM Institutional Repository
language	English
topic	QA75.5-76.95 Electronic computers Computer science
spellingShingle	QA75.5-76.95 Electronic computers Computer science Irani, Arash Azim Zadeh Segmentation Assisted Object Distinction For Direct Volume Rendering
description	Ray casting is a direct volume rendering technique for visualizing 3D arrays of sampled data. It has vital applications in medical and biological imaging. Nevertheless, it is inherently open to cluttered classification results. It suffers from overlapping transfer function values and lacks a sufficiently powerful voxel parsing mechanism for object distinction. In this research work, we are proposing an image processing based approach towards enhancing ray casting technique’s object distinction process. The ray casting architecture is modified to accommodate object membership information generated by a K-means based hybrid segmentation algorithm. Object membership information is assigned to cubical vertices in the form of ID tags. An intra-object buffer is devised and coordinated with inter-object buffer, allowing the otherwise global rendering module to embed multiple local (secondary) rendering processes. A local rendering process adds two advantageous aspects to global rendering module. First, depth oriented manipulation of interpolation and composition operations that lead to freedom of interpolation method choice based on the number of available objects in various volumetric depths, improvement of LOD (level of details) for desired objects and reduced number of required mathematical computations. Second, localization of transfer function design that enables the utilization of binary (non-overlapping) transfer functions for color and opacity assignment. A set of image processing techniques are creatively employed in the design of K-means based hybrid segmentation algorithm.
format	Thesis
qualification_name	Doctor of Philosophy (PhD.)
qualification_level	Doctorate
author	Irani, Arash Azim Zadeh
author_facet	Irani, Arash Azim Zadeh
author_sort	Irani, Arash Azim Zadeh
title	Segmentation Assisted Object Distinction For Direct Volume Rendering
title_short	Segmentation Assisted Object Distinction For Direct Volume Rendering
title_full	Segmentation Assisted Object Distinction For Direct Volume Rendering
title_fullStr	Segmentation Assisted Object Distinction For Direct Volume Rendering
title_full_unstemmed	Segmentation Assisted Object Distinction For Direct Volume Rendering
title_sort	segmentation assisted object distinction for direct volume rendering
granting_institution	Universiti Sains Malaysia
granting_department	Pusat Pengajian Sains Komputer
publishDate	2013
url	http://eprints.usm.my/43239/1/ARASH%20AZIM%20ZADEH%20IRANI24.pdf
_version_	1747821184644284416

Segmentation Assisted Object Distinction For Direct Volume Rendering

Similar Items