An edge-preserving image compression model is presented based on subband co
ding and iterative constrained least square regularisation. The idea is to
incorporate the technique of image restoration into the current lossy image
compression schemes. The model utilises the edge information extracted fro
m the source image as a priori knowledge for the subsequent reconstruction.
Generally, the extracted edge information has a limited range of magnitude
s and it can be lossily conveyed. Subband coding, one of the outstanding lo
ssy image compression schemes, is incorporated to compress the source image
. Vector quantisation, a block-based lossy compression technique, is employ
ed to compromise the bit rate incurred by the additional edge information a
nd the target bit rate. Experiments show that the approach could significan
tly improve both the objective and subjective quality of the reconstructed
image by preserving more edge details. Specifically, the model incorporated
with SPIHT (set partitioning in hierarchical trees) outperformed the origi
nal SPIHT with the 'Baboon' continuous-tone test image. In general, the mod
el may be applied to any lossy image compression systems.