A Modified BTC Using Quincunx Subsampling and Pattern Fitting for Very Low BPP.
Bibhas Chandra Dhara, Bhabatosh Chanda · Indian Conference on Computer Vision, Graphics and Image Processing · 2004
In conventional block truncation coding images are divided into blocks of size and then each block is encoded by two gray levels and a bit-pattern. Here instead of blocks, image is divided into blocks. From each block two sub-blocks are generated by quincunx sampling method. In this work a modified scheme of BTC for the block have been proposed where the computed representative gray levels are the bias and the contrast in each block. Secondly, instead of determining bit-pattern for each block, an optimum bit-pattern is selected from a patternbook.Thirdly, if the contrast is low the block is assumed to be smooth and bit-pattern is not required to reconstruct the block. Again to reduce the bit rate the contrast component and the predictive residual of the bias component are entropy coded to achieve lower bit rate. Further to reduce the bit rate, preserving same quality, indices of best fit patterns are coded using two different index graph. Finally, block is reconstructed using interpolation technique.