A Lossy Compression Algorithm for Bayer Pattern Color Filter Array Data

An efficient algorithm for lossy compression and decompression of Bayer pattern Color Filter Array (CFA) data is presented. The compression is carried out in the camera while the decompression followed by demoasicing and other ancillary image processing tasks are performed in an external device. The proposed method provides better quality reconstructed color image at a given bitrate, reduces computational complexity and prolongs battery life by reducing the processing steps performed in camera. The algorithms combines modified color space transformation and prediction. Simulations demonstrate that the results of proposed algorithm are superior to some previously proposed approaches.