BLOCK ARTIFACT REDUCTION USING BINARY EDGE RETAINING MAPS
Wasim Akram Khan, Tabrez Khan · International Journal of Advanced Research in Computer Science · 2018
in this paper we propose a new edge-retaining algorithm for block artifact reduction. We have classified the pixels to categorize every pixel into one amongst two categories, particularly edge region and smooth region and then Edge-Retaining maps are used to describe these regions. Finally, A Two-step adaptive filter, based on these maps and incorporating offset filtering and edge retaining filtering is used for removing the block artifacts. Experimental Results prove that our proposed edge retaining algorithm adequately reduces block artifacts while conserving the details and the Peak to Signal Noise Ratio (PSNR) performance is better than that of various previous algorithms which use block classification.