A fast algorithm for DEM lossless compression based on embedded wavelet coding
Jingjing Zheng,Jinyun Fang,Chengde Han
DOI: https://doi.org/10.3772/j.issn.1002-0470.2009.12.006
2009-01-01
Abstract:To decrease the digital elevation model (DEM) data volume for storage and transmission in a net-geographic information system (GIS), a fast embedded compression (FEC) algorithm for DEM lossless compression is proposed. It has characteristics of resolution embedding, region of interest (ROI) embedding, and fidelity embedding, and its computing complexity is low. Every resolution of 2/6 integer wavelet transform coefficients is partitioned into many precincts according to the area. In each sub-band of each precinct, only the spatio-temporal neighborhood relationship is used to remove redundancies, and the bits of every bit-plane are modeled and reordered to form three sub-processes and run-length encoded only in one pass. The adaptive Golomb_Rice coding is used to entropy code effectively. The experiments show that, compared with JPEG2000, FEC's encoding and decoding time decrease 36.04% and 44.49% respectively, while the compression ratio decreases only 8.05%;compared with SPIHT, FEC's encoding and decoding time can decrease 32.28% and 37.49% respectively, while the compression ratio decreases only 4.58%;compared with GZIP, FEC's compression ratio increases 79.63%;compared with the n neighbors' prediction algorithm, FEC's compression ratio increases 9.23%. Thus, FEC can decrease the coding and decoding time largely, while maintains favorable compression performance.