earticle

영어

Transmission of block-coded images over wireless channels results in lost blocks. These losses may occur in region-of-interest (ROI) of images which contain important information of images. In this paper, we propose a new error concealment method for covering up the high packet losses of an original image after its transmission through an error-prone channel. In this scheme, Discrete Wavelet Transform (DWT) is applied to each block of the original image in order to produce a lower resolution copy of the block. Then, we choose approximation coefficients of each block as replica of the block and embed it into a remote block (called masking block) of the image in the spatial domain. We use a new algorithm to find the masking block for embedding the replica. The advantage of this algorithm is that it avoids losing simultaneously the information of the block and its masking block. It is shown that the proposed scheme provides significant improvement over existing algorithms in terms of both subjective and objective evaluations. This technique can be implemented for wireless channels to combat degradations in a backward-compatible scheme. It is also shown that this scheme can efficiently restore the lost blocks in the ROI of images.