Regularization-Based Error Concealment in JPEG 2000 Coding Scheme Export

@article{citeulike:2373699, abstract = {This letter offers a solution to bit-error recovery for high-frequency subbands at high decomposition levels of JPEG 2000 bitstreams, with a novel perspective. Being inspired by the regularization approach, corrupted coefficients are estimated by imposing a smoothness constraint on the gradient of the reconstructed data. The approach can be generalized to other compression engines as well. Experimental results show that the proposed method achieves the best reported results so far, with up to 6 dB of PSNR improvement. The method does not require any parameter tuning, however, it does require numerical minimization.}, author = {Oktem, Rusen}, booktitle = {Signal Processing Letters, IEEE}, citeulike-article-id = {2373699}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/LSP.2007.906623}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4358026}, doi = {10.1109/LSP.2007.906623}, journal = {IEEE Signal Processing Letters}, keywords = {2000, bit-error, coding, coefficients, compression, concealment, constraint, corrupted, engines, error, image, jpeg, minimization, numerical, parameter, reconstruction, recovery, regularization, regularization-based, scheme, smoothness, statistics, tuning}, month = {December}, number = {12}, pages = {956--959}, posted-at = {2008-02-14 13:02:17}, priority = {2}, title = {Regularization-Based Error Concealment in JPEG 2000 Coding Scheme}, url = {http://dx.doi.org/10.1109/LSP.2007.906623}, volume = {14}, year = {2007} }

TY - JOUR ID - citeulike:2373699 L3 - citeulike-article-id:2373699 N2 - This letter offers a solution to bit-error recovery for high-frequency subbands at high decomposition levels of JPEG 2000 bitstreams, with a novel perspective. Being inspired by the regularization approach, corrupted coefficients are estimated by imposing a smoothness constraint on the gradient of the reconstructed data. The approach can be generalized to other compression engines as well. Experimental results show that the proposed method achieves the best reported results so far, with up to 6 dB of PSNR improvement. The method does not require any parameter tuning, however, it does require numerical minimization. IS - 12 JF - IEEE Signal Processing Letters EP - 959 TI - Regularization-Based Error Concealment in JPEG 2000 Coding Scheme VL - 14 T2 - Signal Processing Letters, IEEE SP - 956 KW - 2000 KW - bit-error KW - coding KW - coefficients KW - compression KW - concealment KW - constraint KW - corrupted KW - engines KW - error KW - image KW - jpeg KW - minimization KW - numerical KW - parameter KW - reconstruction KW - recovery KW - regularization KW - regularization-based KW - scheme KW - smoothness KW - statistics KW - tuning AU - Oktem, Rusen PY - 2007/12// UR - http://dx.doi.org/10.1109/LSP.2007.906623 ER -