Abstract: In order to improve the compression efficiency of the image data compression algorithmadopted by consulative committee for space data system (CCSDS), a two dimensional prediction model was proposed. By utilizing the spatial correlation of pixels existing in the image to improve and optimize the prediction of direct current (DC) coefficients and the entropy coding, the compression performance of DC coefficients as well as enhance the ability for error resiliency could be improved. Experimental results show that the algorithm can obtain better quality of image encoding at low bitrate image and improve (S/N)_P by 0.13~0.3dB compared with the original algorithm in CCSDS. Moreover, the proposed algorithm significantly meet the need of the low bit rate transmission of the onboard image compression data.

Key words: Differential pulse code modulation , Entropy coding , Two-dimensional prediction model , Direct current coefficient, Remote sensing image