Automatic Attack Identification and Frame Recovery Using Versatile Video-based Watermarking
Year: 2009
Type: conference-paper
Abstract: — Aim at attack identification and frame recovery, we propose a versatile video-based watermarking scheme. This scheme integrates halftoning and watermarking techniques, and the watermark is particularly composed of two types of data, namely analogue data and index data. To examine the extracted watermark is not only to authenticate the video content, but it is capable of attack identification and corrupted frame recovery. Furthermore, a big breakthrough of the proposed scheme is that original data is unavailable, we merely use the extracted watermark compared with the covered frame. The experiment results demonstrate that our scheme actually diagnoses the frame/video statement, detects the tampered areas and then recovers them. I.
Author: Yi‐Chong Zeng
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W56886187