Lossless Coding of Bayer Color Filter Array Images Using Block-Adaptive Prediction

A novel lossless coding scheme for images captured with the Bayer color filter array (CFA), which is widely employed for color digital cameras having a single image sensor, is proposed. The scheme uses a block-adaptive prediction method to exploit spatial and spectral correlations in local areas containing different color signals. To allow adaptive prediction suitable for the respective color signals, four kinds of liner predictors that correspond to 2 × 2 samples of the Bayer CFA are simultaneously switched block-by-block. Experimental results show that the proposed scheme attains 9--15% better coding efficiency than the JPEG 2000-based lossless coding scheme.