A no reference texture granularity index and application to visual media compression

Abstract

Texture granularity is an important attribute to quantify the level of details present in the image. This work presents a no-reference texture granularity index and shows using subjective experiments that the proposed granularity index correlates well with the perceived granularity of textures. In addition, a subjective study is conducted to assess the effect of compression on textures with varying degrees of granularity. It is shown that a measure of texture granularity can predict the compression quality.

Predicting video rate-distortion curves using textural features

“…There exists some previous work that relates textural features to video compression efficiency [9,13,14]. In [9], Subedar et al define a no-reference metric of granularity in static textured images and discuss its relation to compression efficiency, but with no clear association to RD curves.…”

“…Textural features are conventionally defined with the purpose of facilitating similarity, browsing, retrieval and classification applications [2,[6][7][8][9][10][11][12]. Additionally, most works have only considered static textures, namely images [2,[6][7][8][9]. Hence, most textural features do not capture the dynamic characteristics that texture obtains in videos.…”

“…In [9], Subedar et al define a no-reference metric of granularity in static textured images and discuss its relation to compression efficiency, but with no clear association to RD curves. In [14], elementary statistics of prediction error for texture and motion vectors for motion are obtained from H.264/AVC encoder and used to build variability-distortion models.…”

Synthesized Texture Quality Assessment via Multi-scale Spatial and Statistical Texture Attributes of Image and Gradient Magnitude Coefficients

“…The HH and VH subbands at the j th scale are denoted by HH j and V H j , respectively. For computing the granularity, G j , and regularity, R j , features at the j th scale, local peaks are detected by locating (as in [34]) the local maxima of the wavelet coefficients' magnitude along the rows and columns of the HH j and V H j subbands, respectively. Distances between adjacent located peaks are computed for every row (column) in the HH j ( V H j ) subband.…”

Video texture analysis based on HEVC encoding statistics

“…A texture with smaller size primitives has a high granularity level, while a texture with a larger size primitives has a low granularity level [15].…”