Institute of Information Theory and Automation

Publication details

Forensics image and video restoration

Abstract

Zitová Barbara, Šroubek Filip, Kamenický Jan, Kotera Jan, Šorel Michal, Bartoš Michal, Flusser Jan, Šíma Z., Švarc P., Hořínek J.


serial: Abstract book of 7th European Academy of Forensic Science Conference

action: European Academy of Forensic Science Conference 2015 /7./, (Praha, CZ, 06.09.2015-11.09.2015)

project(s): VG20102013064, GA MV

keywords: image restoration, image forensics

preview: Download

abstract (eng):

We have addressed image and video restoration in situations when the quality of acquired images or videos in terms of their noise level and sharpness is not sufficient and their improvement is strongly required. Nowadays forensic analysis is often based on the analysis of digital photographs and videos (segmentation, object detection, change analysis, to name a few), but not all the time their quality is sufficient for their efficient processing. Analyzed images and video can be blurred, noisy, with low resolution, or degraded in some other unknown way. The aim is to be able to combine all available information in order to produce restored image/video, which is sharper, with less noise and blur, and possibly with higher resolution, then the original data file. A methodology based on the blind deconvolution methods and the super-resolution forms a basis of proposed image quality enhancement and restoration methods.

RIV: JD

Responsible for information: admin
Last modification: 21.12.2012
Institute of Information Theory and Automation