Flicker Reduction and Visual Quality Improvement for High Definition Compressed Videos

2016 Fiscal Year Final Research Report

• Project/Area Number: 26330128
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Multimedia database
• Research Institution: The University of Electro-Communications
• Principal Investigator: ZHANG Xi 電気通信大学, 大学院情報理工学研究科, 教授 (40251706)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 動画像符号化 / ノイズ除去 / フリッカー低減 / ウェーブレット変換 / 閾値処理 / 高精細動画像 / ディジタルフィルタ / 画質改善

Outline of Final Research Achievements

Flicker artifact can appear in compressed videos even at medium bit rate. It is perceptible as a small variation in pixel value, and is extremely visible in high definition videos. This visual artifact impairs significantly the visual quality of compressed videos. Therefore, it is important to suppress the flicker artifact and to improve the visual quality of compressed videos.
In this research, we introduce a new approach to flicker reduction of compressed videos by using wavelet thresholding. The flicker artifact can be thought of as Gaussian noise in the temporal domain. We view this flicker reduction as a denoising problem in the temporal domain, and apply the wavelet thresholding algorithm in solving this flicker reduction problem. The wavelet thresholding algorithm has exhibited the excellent denoising property. It is shown from our experimental results that our wavelet-based method suppresses the flicker artifact and achieve an improvement, compared with the conventional methods.

Free Research Field

画像処理