Channel noise filtering for subband image coding using vector quantization

In this paper, we design a post-filter for the subband coding of images using vector quantization. The purpose of this filter is to remove the inevitable channel noise after the compressed image is transmitted across a channel. We first proposed two simple schemes to detect the edges and channel noises in the subbands based on the image and subband characteristics. Then the noise corrupted subbands at the uniform, edge, and texture region are then adaptively low-pass filtered. After noise filtering, the image is then reconstructed from all its subbands. Since the proposed filter does not assume any knowledge about the quantization, it can be used for any subband coding schemes. The experimental results show that the proposed filter always improves the coding performance when the channel bit error rate is greater than 0.01%. The largest gain (5.7 dB) occurs when the bit error rate is 1%.