Please use this identifier to cite or link to this item:
|Title:||Image restoration using sparse dictionary||Authors:||Dai, Shi||Keywords:||DRNTU::Engineering::Electrical and electronic engineering||Issue Date:||2018||Abstract:||Sparse theory has been applied widely to the field of image processing since the idea of sparse representation of images was first proposed by Dr. Stephen Mallat. Image restoration is the process of estimating the corrupt and unknown pixels in an image from its known information, making repaired image close to or achieve the visual effect of the original image. In the past decade, sparse theory applied to image denoising and inpainting has become a popular research topic in the field of image processing. This project aims to research on sparse representation theory and the concept of dictionary training and implement them to images to solve image denoising and image inpainting problems. The main research works of this project are as follow: 1. Introduce the basic concepts of sparse representation, discuss the main algorithms used to solve the problem of sparse approximation and the main dictionary algorithms in sparse representation. 2. Introduce image recovery (denoising and inpainting) problems based on sparse representation, research on K-SVD dictionary. 3. Illustrate the application of trained dictionary in image recovery (denoising and inpainting) 4. Assess the effectiveness of the training dictionary used.||URI:||http://hdl.handle.net/10356/75450||Rights:||Nanyang Technological University||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Student Reports (FYP/IA/PA/PI)|
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.