Title :
Removing Channel Error Artifacts in JPEG 2000 Coded Images
Author :
Zargari, Farzad ; Sefidpour, Ali ; Moin, M. Shahram
Author_Institution :
Multimedia Res. Group, Iran Telecom Res. Center, Tehran, Iran
Abstract :
Segmentation symbol is an error resilience feature in JPEG 2000 image coding standard. When this feature is used, not detecting the segmentation symbol at the end of a bit plane by decoder, indicates that an error has occurred within that bit plane. However, a detected and correctly decoded segmentation symbol does not guarantee that the bit plane is error free. In this paper, an error concealment method is proposed for removing artifacts caused by the damaged bit planes that were not detected by the segmentation symbol. The proposed method provides a model for coded bit planes in the encoder and uses this model for error concealment in decoder. Simulation results show that the proposed method can effectively remove artifacts and improve the quality of corrupted images by up to 6 dB in PSNR.
Keywords :
decoding; image coding; image segmentation; JPEG 2000 coded images; PSNR; channel error artifacts removal; decoder; error concealment method; image coding standard; segmentation symbol; Code standards; Computer errors; Decoding; Discrete wavelet transforms; Error correction; Image coding; Image segmentation; Signal processing; Telecommunications; Tiles; Channel Error; Error Concealment; JPEG 2000; Segmentation Symbol;
Conference_Titel :
Signal Processing and Communications, 2007. ICSPC 2007. IEEE International Conference on
Conference_Location :
Dubai
Print_ISBN :
978-1-4244-1235-8
Electronic_ISBN :
978-1-4244-1236-5
DOI :
10.1109/ICSPC.2007.4728326