<title>Enhancement of decompressed images by optimal shift-invariant wavelet packet basis</title>

Wei Dong, Alan C. BovikProceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE 1996 · DOI

Abstract

A new post-processing method based on wavelet thresholding is proposed to enhance transform coded images. At the encoder, the optimal (in the mean square error sense) shift- invariant wavelet packet basis is searched using a fast optimization algorithm and the location of the best basis in the entire shift-invariant wavelet packet tree is transmitted as overhead information with the bit stream to the decoder. After the decoder reconstructs the decompressed image, the post-processor performs thresholding using the optimal basis. This method can achieve comparable enhancement performance to the method using the undecimated discrete wavelet transform, and requires the same low post-processing complexity as the method using the orthonormal discrete wavelet transform.
BibTeX
@inproceedings{dong1996lt,
  author = {Wei Dong and Alan C. Bovik},
  title = {{\<title\>Enhancement of decompressed images by optimal shift-invariant wavelet packet basis\</title\>}},
  booktitle = {Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE},
  year = {1996},
  doi = {10.1117/12.234761},
}

Loading related content...

Figure