Efficient Directional Interpolation for Block Recovery Using Difference Values of Border Pixels

Video Communications Over Wireless Ad-Hoc Networks Using Source Coding Diversity and Multiple Paths

“…Aign & Fazel (1995) proposed to interpolate lost pixel values from the boundary pixels of the four neighboring MBs. In addition, Kwok & Sun (1993), Zeng & Liu (1999), Zhang et al (2004), Kung et al (2006), Hyun et al (2008), Luo et al (2009) used different algorithms to detect the edges within the lost MBs and directionally interpolate the lost pixels along the edges. A more complex approach called coarse-to-fine block replenishment (CFBR) (Belfiore et al, 2002) performed the interpolation by first recovering the smooth large-scale patterns, then the large-scale structures, and finally the local edges in the lost MB.…”