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

Understanding Contours and Its Appli...

Mao, Xiangyu.

Linked to FindBook

Google Book

Amazon

博客來

Understanding Contours and Its Applications.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Understanding Contours and Its Applications./
Author:	Mao, Xiangyu.
Published:	Ann Arbor : ProQuest Dissertations & Theses, : 2017,
Description:	154 p.
Notes:	Source: Dissertation Abstracts International, Volume: 79-05(E), Section: B.
Contained By:	Dissertation Abstracts International79-05B(E).
Subject:	Computer science. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10753202
ISBN:	9780355552935

Understanding Contours and Its Applications.
Mao, Xiangyu.

Understanding Contours and Its Applications. - Ann Arbor : ProQuest Dissertations & Theses, 2017 - 154 p.

Source: Dissertation Abstracts International, Volume: 79-05(E), Section: B.

Thesis (Ph.D.)--The Chinese University of Hong Kong (Hong Kong), 2017.

This item is not available from ProQuest Dissertations & Theses.

Nowadays computer vision has become more and more popular in our daily life. From cell phones to drones, the booming of electronic devices with camera greatly promotes the development of computer vision techniques. Among these techniques, contour detection is a fundamental technique which benefits many applications. Correct and clear contour detection results help a lot not only in regular prenatal care of mothers, but also in environmental analysis based on images sent back from robots on Mars. However, due to the large diversity of natural images, a general contour detection method with satisfying results in all scenarios has not been found by human yet. In this thesis, we propose novel perception based contour analysis methods specialized for different applications which detect contours for cartoon, introduce stereoscopic effect into cel animation and share visual content between color vision deficiency (CVD) audiences and normal-vision audiences.

ISBN: 9780355552935Subjects--Topical Terms:

523869
Computer science.

Understanding Contours and Its Applications.
LDR:04474nmm a2200337 4500 001 2203566
005 20190531105740.5
008 201008s2017 ||||||||||||||||| ||eng d
020 $a 9780355552935
035 $a (MiAaPQ)AAI10753202
035 $a AAI10753202
040 $a MiAaPQ $c MiAaPQ
100 1 $a Mao, Xiangyu. $3 3430366
245 1 0 $a Understanding Contours and Its Applications.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2017
300 $a 154 p.
500 $a Source: Dissertation Abstracts International, Volume: 79-05(E), Section: B.
500 $a Adviser: Tien Tsin Wong.
502 $a Thesis (Ph.D.)--The Chinese University of Hong Kong (Hong Kong), 2017.
506 $a This item is not available from ProQuest Dissertations & Theses.
520 $a Nowadays computer vision has become more and more popular in our daily life. From cell phones to drones, the booming of electronic devices with camera greatly promotes the development of computer vision techniques. Among these techniques, contour detection is a fundamental technique which benefits many applications. Correct and clear contour detection results help a lot not only in regular prenatal care of mothers, but also in environmental analysis based on images sent back from robots on Mars. However, due to the large diversity of natural images, a general contour detection method with satisfying results in all scenarios has not been found by human yet. In this thesis, we propose novel perception based contour analysis methods specialized for different applications which detect contours for cartoon, introduce stereoscopic effect into cel animation and share visual content between color vision deficiency (CVD) audiences and normal-vision audiences.
520 $a We first makes efforts in contour detection for cartoon based on perceptual analysis. Cartoon is a worldwide popular visual entertainment with a long history. Nowadays, there has been a rising need of digitizing the old classic cartoons for further editing, e.g., deformation, colorization, etc. To perform such editing, extracting the contour lines inside the cartoon images is indispensable. Traditional contour detection methods perform poorly when facing compression artifacts and spatially varying lines. Our perception based approach works in an out of the ordinary way which analyze the problem in perspective of regions.
520 $a Secondly, we tackle the problem of introducing stereoscopic effect into cel animations based on contour junction analysis. We propose a novel method to automatically synthesize stereoscopic cel animations from ordinary 2D cel animation, without 3D geometry reconstruction nor labor-intensive manual assignment of depth. We first infer the initial ordering of regions from the Tjunction of contours available in cartoons. The final ordering is determined by a graph-cut process to resolve the temporal inconsistency problem. A further depth optimization process with the ordering formulated as constraints is proposed to generate smooth-changing depth. Finally we re-render the cel animation from two new viewpoints and inpaint the disoccluded pixels with depth based contour extension method to generate convincing stereoscopic effect.
520 $a The last problem we tackled is the visual sharing between CVD and normal audiences. 4 percent of world population suffer from color vision deficiency (CVD). They can hardly share the same visual content with normal-vision audiences. We propose the first system that allows CVD and normal-vision audiences to share the same visual content simultaneously. The key that we can achieve this is because the ordinary stereoscopic display offers users two visual experiences (with and without wearing stereoscopic glasses). By allocating one experience to CVD audiences and one to normal-vision audiences, we allow them to share. The core problem is to synthesize an image pair, that when they are presented binocularly, CVD audiences can distinguish the originally indistinguishable colors; and when it is in monocular presentation, normal-vision audiences cannot distinguish its difference from the original image. We solve the image-pair recoloring problem by contour structure analysis of CVD and normal perceptions.
590 $a School code: 1307.
650 4 $a Computer science. $3 523869
650 4 $a Computer engineering. $3 621879
690 $a 0984
690 $a 0464
710 2 $a The Chinese University of Hong Kong (Hong Kong). $b Computer Science and Engineering. $3 3428136
773 0 $t Dissertation Abstracts International $g 79-05B(E).
790 $a 1307
791 $a Ph.D.
792 $a 2017
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10753202