WSEAS Transactions on Signal Processing
Print ISSN: 1790-5052, E-ISSN: 2224-3488
Volume 12, 2016
Quality Assessment on Digital Images Using Parallel Non-Linear De-Blocking Model
Authors: D. Ferlin Deva Shahila, S. H. Krishnaveni
Abstract: To achieve a non-linear de-blocking on digital images, a framework called Parallel Non-linear De-Blocking (PN-DB) is developed in this paper. The PN-DN framework is developed to reduce the blocking artifacts and therefore to increase the image quality. In PN-DB framework, Adaptive Structures Directional Lifting (ASDL) with Discrete Wavelet Transform (DWT) performs multi scale histogram representation on digital images. The ASDL changes the sampling matrix into sub-regions of digital images and improves the performance of the lossy-to-lossless image coding application. In the ASDL scheme, a Sinc interpolation filter with constant coefficient is adopted to interpolate both straight and perpendicular direction of the digital image aiming to minimize prediction errors during coding results. Finally, lossy and lossless digital image coding results of the PN-DB framework are shown to validate the advantage of the proposed structure. The performance of the proposed framework is evaluated using The Manuscripts and Archives Digital Images Database (MADID). Simulations conducted with MADID show the performance improvement in image quality by minimizing prediction error during coding results that further reduced the blocking artifacts in an extensive manner.
Keywords: Multimedia, Non-linear de-blocking, Artifacts, Adaptive Structures, Directional Lifting, Discrete Wavelet Transform, Sinc interpolation filter
Pages: 118-129WSEAS Transactions on Signal Processing, ISSN / E-ISSN: 1790-5052 / 2224-3488, Volume 12, 2016, Art. #15