Fast Golomb coding parameter estimation using partial data and its application in hyperspectral image compression
Hongda Shen, W. David Pan, Dongsheng Wu, Maliha Lubna · 2016
The efficiency of a Golomb code depends on how well its coding parameter reflects the input data to be compressed. Typically, such a coding parameter is estimated from the entire dataset. We propose a fast processing method, which estimates the coding parameter based on a very sparse sampling of the dataset. Simulations on both synthetic data and hyperspectral image data showed that the fast method achieves compression efficiency closely resembling the full-data estimation method. By involving only a small portion of the original data, the proposed method tends to incur less processing delays and memory usage.