Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Shells and Spheres: A Novel Framework for Variable Scale Statistical Image Analysis

Cois, C. Aaron (2006) Shells and Spheres: A Novel Framework for Variable Scale Statistical Image Analysis. Master's Thesis, University of Pittsburgh. (Unpublished)

Primary Text

Download (3MB) | Preview


A framework for analyzing images, called emph{Shells and Spheres},has been developed based on a set of spheres with adjustable radii,with exactly one sphere centered at each image pixel. This set ofspheres, known as a emph{sphere map}, is considered optimized wheneach sphere reaches, but does not cross, the nearest boundary.Calculations denoted as emph{Variable-Scale Statistics} (VSS) areperformed on populations of pixels within spheres, as well aspopulations of adjacent and overlapping spheres, in order to deducethe proper radius of each sphere. Spheres grow or shrink by addingor deleting an outer shell one pixel thick . Unlike conventionalfixed-scale kernels, our spherical operators consider as many pixelsas possible to differentiate between objects and accuratelydelineate boundaries. The term ``sphere" is used for brevity, thoughthe approach is not limited to 3D and is valid in $n$-dimensions.The approach is illustrated using both real images and noiselesssynthetic images containing objects with uniform intensity, and moreclosely examined and validated using various synthetic images withadded white noise and multiple contrast enhanced CT scans of theaortic arch. A particular algorithm using Shells and Spheres isdescribed and demonstrated on segmentation of the aortic arch in acontrast-enhanced CT scan, both in 2D and 3D.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Cois, C.
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee CoChairStetten,
Committee CoChairHauskrecht, Milosmilos@cs.pitt.eduMILOS
Committee MemberLi, C. C.ccl@engr.pitt.eduCCL
Date: 28 September 2006
Date Type: Completion
Defense Date: 9 May 2006
Approval Date: 28 September 2006
Submission Date: 16 May 2006
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Institution: University of Pittsburgh
Schools and Programs: Dietrich School of Arts and Sciences > Computer Science
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: computer vision; image processing; image segmentation; medical image analysis; medical imaging
Other ID:, etd-05162006-155257
Date Deposited: 10 Nov 2011 19:44
Last Modified: 15 Nov 2016 13:43


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item