Improved predictor blending technique for lossless image coding
Grzegorz Ulacha, Ryszard Stasiński · ICSES 2010 International Conference on Signals and Electronic Circuits · 2010
A highly efficient predictor blending method is described in the paper. A set of 19 blended predictors is defined, including a context static predictor, ALCM, CoBALP, and texture context mapping one. A sophisticated error bias removal method is applied to few predictors. Prediction error is coded using an advanced adaptive context arithmetic coder. Experimental results show that the presented technique is more efficient and faster than GLICBAWLS, not talking about other state-of-the-art methods.