Spike Defect Detection and Removal on Aged Films

NCUIR > College of Electrical Engineering & Computer Science > Department of Computer Science and information Engineering > journal & Dissertation > Item 987654321/51584

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/51584

Title:	Spike Defect Detection and Removal on Aged Films
Authors:	Chang,RC;Shih,TK
Contributors:	資訊工程學系
Keywords:	ALGORITHM;VIDEO
Date:	2010
Issue Date:	2012-03-27 18:56:42 (UTC+8)
Publisher:	國立中央大學
Abstract:	Aged films contain valuable historical information. However, some films may have defects due to dirt, scratches or for other reasons. Usually, there is no efficient prediction model to precisely identify these defects. This study proposes a set of new algorithms based on the analysis of features among frames in aged films. Both spatial and temporal characteristics are used. The algorithms further use filters with adjustable thresholds to precisely identify defects, such as spikes and scratches. Finally, a temporal inpainting method and a multiple resolution image inpainting algorithm are used to repair the defects. Since no efficient automatic quantitative mechanism is available due to the loss of original films, the experiments use subjective evaluation on more than ten films. The results show that most defects are repaired with a pleasant visual quality.
Relation:	JOURNAL OF INFORMATION SCIENCE AND ENGINEERING
Appears in Collections:	[Department of Computer Science and information Engineering] journal & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	344	View/Open

社群 sharing