Use this URL to cite or link to this record in EThOS:
Title: Resolution improvement and effect of blurring in medical image analysis and computer vision
Author: Ho, Edward Yu Tat
ISNI:       0000 0004 2676 5124
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2008
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
With the increasing ability of computational workstations, there are similarly increasing demands for better quality and higher resolving power in reconstructed images. We are particular interested to investigate how image resolution can be improved and the effect of blurring in medical image analysis and computer vision can be reduced and predicted. One of the main factors in image degradation is blurring. If we can make predictions on how the effect of blurring has on local shape boundaries, this will be advantageous for better image analysis and understanding of the edge localisation properties due to blurring. Part of this thesis investigates how to make predictions on the effect of blurring on shape boundaries. Generalised Kaiser-Bessel window functions are first used for image filtering. They have also been used as the basis functions (blobs) for image reconstruction from projections. However, the ability of blobs to replace pixels in 2-D or voxels in 3-D as the basis functions are not limited to medical image reconstructions. In this thesis, we introduce our novel approach of incorporating blob-based basis functions into the super-resolution reconstruction from multiple low- resolution image datasets. Super-resolution imaging and reconstruction itself is a rather new research framework for medical image reconstructions. We are interested in investigating how the use of blob-based basis functions can help in further improving the quality and resolving power of reconstructed images. We show in the thesis that the quality of the reconstructed images is increased by many folds when using our novel approach. We are also interested to investigate how blobs can be used to regularise iterative algorithms for image reconstructions. Blob-based basis functions can be used to stabilize iterative image reconstructions as we have also shown in this thesis. Finally, we give proposal and extension of the applications of blob-based super-resolution reconstructions for various medical imaging and reconstruction frameworks.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available