Highly noise resistant Beltrami-based texture maps with window derivative

A Novel Joint Blocking and Ringing Artifact Reduction using Advanced Beltrami Based Texture Maps

“…Therefore, neighbours of strong edge pixels may be suffered from ringing artifacts. The strong edge map [10] is chosen to construct the ringing map. Figure 3(d) shows a ringing map with search window = 8, where green color pixels are ringing pixels and red color pixels are strong edge pixels.…”

“…With the fact that details such as texture are lost during compression, texture estimation based on the compressed images is thus a more difficult task. Most of the previous texture estimation methods focus on estimation on original images in high quality ( [1,9,10]). In this paper, the compressed images are used while original images are also used for accuracy comparison.…”

“…But texture maps based on patches are not highly accurate since the classification error is high. In [10], the texture feature F(x, y) based on window derivative is introduced to obtain the texture map with high accuracy as in (7). Let W(x, y) be a window of (2R + 1) × (2R + 1) pixels.…”

“…As can be seen in Figure 4(c) and Figure 4(d), there are many isolated texture pixels which do not correctly represent the texture area because they are generally located in groups. So isolated pixels in the texture map should be removed since textures are geometric structures while noise is not ( [10]). As can be seen in Figure 4 5 Proposed Deblocking and Deringing Using Texture Map Figure 5 shows the flow chart of the proposed deblocking and deringing method.…”

“…Construction of these object maps thus play an essential role in order to control the filter's strength. The authors 1859-378X-2017-1203 c 2017 REV in [8], [9], [10] and [11] use edge maps to adapt the fuzzy filter artifact reduction. In these methods, the variance and standard deviation value ( [12], [13]) are used to construct the edge map.…”