Bibliographic citations
Vargas, M., (2012). Image Restoration by Inpainting Methods Applied to CT Reconstruction [Trabajo de investigación, Institut National des Sciences Appliquées de Lyon]. http://renati.sunedu.gob.pe/handle/sunedu/1336616
Vargas, M., Image Restoration by Inpainting Methods Applied to CT Reconstruction [Trabajo de investigación]. FR: Institut National des Sciences Appliquées de Lyon; 2012. http://renati.sunedu.gob.pe/handle/sunedu/1336616
@mastersthesis{renati/1776,
title = "Image Restoration by Inpainting Methods Applied to CT Reconstruction",
author = "Vargas Ángeles, Marcos Christian",
publisher = "Institut National des Sciences Appliquées de Lyon",
year = "2012"
}
Digital inpainting is the art of filling in lost or deteriorated parts of images or videos in a form that is not noticeable by an ordinary observer. Artifact is any error in the information representation introduced by the involved equipment or technique. Artifact Reduction, is considered a critical topic in x-ray CT research and development. Our approach is to get a metal artifact reduction (MAR), by comparing 2D inpainting results in the projection and sinogram domains. Our CT images came from: CIVA (simulation), SkyScan2011, PerkinElmer (experimental). First, we tested eight inpainting algorithms: Total Variation (TV), Curvature Driven Diffusion (CDD), Mean Curvature Diffusion (MCD), Euler’s Elastica, Morphologic Rotation Invariant (MRI), Fast Marching Method (FMM), Anisotropic Diffusion-Based (AdB) and Exemplar-Based (ExB). Then, we validated our results using mean squared error (MSE), peak signal-to-noise ratio (PSNR) and Multi-Scale Structural Similarity Index (MS-SSIM). Results were compared among them considering: time and the three image quality assessments mentioned. Finally, our experiments demonstrated that our two best MAR quality results were achieved by CDD and ExB, for our experimental images in the projection domain. The results in the sinogram domain were only qualitative.
File | Description | Size | Format | |
---|---|---|---|---|
Vargas_Angeles_MC.pdf | Trabajo de investigación | 4.32 MB | Adobe PDF | View/Open |
Autorizacion.pdf Restricted Access | Autorización del registro | 556.11 kB | Adobe PDF | View/Open Request a copy |
This item is licensed under a Creative Commons License