東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Topics in variational PDE image segm...

Song, Bing.

Linked to FindBook

Google Book

Amazon

博客來

Topics in variational PDE image segmentation, inpainting and denoising.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Topics in variational PDE image segmentation, inpainting and denoising./
Author:	Song, Bing.
Description:	78 p.
Notes:	Source: Dissertation Abstracts International, Volume: 64-08, Section: B, page: 3857.
Contained By:	Dissertation Abstracts International64-08B.
Subject:	Mathematics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3100721

Topics in variational PDE image segmentation, inpainting and denoising.
Song, Bing.

Topics in variational PDE image segmentation, inpainting and denoising. - 78 p.

Source: Dissertation Abstracts International, Volume: 64-08, Section: B, page: 3857.

Thesis (Ph.D.)--University of California, Los Angeles, 2003.

Variational methods have been extensively used and studied in image processing in the past decade because of their flexibility in modeling and various advantages in the numerical implementation. Examples of this include image segmentation, object tracking, texture synthesis, vector field visualization, etc. This dissertation contains the study of variational methods in image segmentation, inpainting and denoising. We propose a fast algorithm for level set based optimization. The main advantage of this method is that the gradient of the functional is not needed. This enables it to be applied to broader range of optimization problems. When applying this algorithm to Chan-Vese image segmentation model, it converges extremely fast. For image inpainting, we formalize a low-level global deterministic solution for texture synthesis and image inpainting. A global energy is defined as a function of correspondence map, which is defined as linking each blank or missing pixel to the pixel where its value is taken from, in the seed image. We demonstrate why they should not be seen as procedures to sample a probability distribution on the correspondence maps. Finally, we discuss the adaptive total variation denoising. This method is a modification of the original total variation denoising method. We prove that, after this modification, the numerical scheme for the Euler-Lagrange equation of the adaptive total variation model is stable. We show numerical experiment that this model can keep the edge of object as well as that of the TV model.Subjects--Topical Terms:

515831
Mathematics.

Topics in variational PDE image segmentation, inpainting and denoising.
LDR:02389nmm 2200253 4500 001 1856261
005 20040621101134.5
008 130614s2003 eng d
035 $a (UnM)AAI3100721
035 $a AAI3100721
040 $a UnM $c UnM
100 1 $a Song, Bing. $3 1944044
245 1 0 $a Topics in variational PDE image segmentation, inpainting and denoising.
300 $a 78 p.
500 $a Source: Dissertation Abstracts International, Volume: 64-08, Section: B, page: 3857.
500 $a Chair: Tony F. Chan.
502 $a Thesis (Ph.D.)--University of California, Los Angeles, 2003.
520 $a Variational methods have been extensively used and studied in image processing in the past decade because of their flexibility in modeling and various advantages in the numerical implementation. Examples of this include image segmentation, object tracking, texture synthesis, vector field visualization, etc. This dissertation contains the study of variational methods in image segmentation, inpainting and denoising. We propose a fast algorithm for level set based optimization. The main advantage of this method is that the gradient of the functional is not needed. This enables it to be applied to broader range of optimization problems. When applying this algorithm to Chan-Vese image segmentation model, it converges extremely fast. For image inpainting, we formalize a low-level global deterministic solution for texture synthesis and image inpainting. A global energy is defined as a function of correspondence map, which is defined as linking each blank or missing pixel to the pixel where its value is taken from, in the seed image. We demonstrate why they should not be seen as procedures to sample a probability distribution on the correspondence maps. Finally, we discuss the adaptive total variation denoising. This method is a modification of the original total variation denoising method. We prove that, after this modification, the numerical scheme for the Euler-Lagrange equation of the adaptive total variation model is stable. We show numerical experiment that this model can keep the edge of object as well as that of the TV model.
590 $a School code: 0031.
650 4 $a Mathematics. $3 515831
690 $a 0405
710 2 0 $a University of California, Los Angeles. $3 626622
773 0 $t Dissertation Abstracts International $g 64-08B.
790 1 0 $a Chan, Tony F., $e advisor
790 $a 0031
791 $a Ph.D.
792 $a 2003
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3100721