Subband Coded Image Reconstruction for Lossy Packet Networks
An algorithm is presented for reconstructing image subband coefficients lost or delayed during transmission. The algorithm consists of two parts, allowing reconstruction of both low-frequency and high-frequency coefficients. Low-frequency coefficients are reconstructed using bicubic interpolation in which the interpolation grid is adapted to achieve accurate edge placement in the synthesized image. The adaptation is based on subband analysis of an edge model, which demonstrates that edges can be identified using the decimated high-frequency subbands without requiring edge detection on the low-frequency band itself. High-frequency coefficients are reconstructed using one-dimensional linear interpolation, which provides good visual performance as well as maintaining properties required for edge placement in the low-frequency reconstruction. The algorithm performs well on losses of single coefficients, l-d vectors, and small blocks, and is therefore applicable to a variety of coding techniques.
KeywordsForward Error Correction Decomposition Level Error Concealment Edge Center Frequency Coefficient
Unable to display preview. Download preview PDF.
- Shacham, N. and McKenney, P. Packet recovery in high-speed networks using coding and buffer management. Proceedings IEEE Infocom 90, Los Alamitos, CA, 1990, Vol. 1, pp. 124–31.Google Scholar
- MacDonald, N. Transmission of compressed video over radio links BT Technology Journal, Vol. 11, No. 2, April 1993, pp. 182–5.Google Scholar
- Johnsen, O., Shentov, O. V. and Mitra, S. K. A Technique for the Efficient Coding of the Upper Bands in Subband Coding of Images Proc. ICASSP 90, Vol. 4, pp. 2097–2100, April 1990.Google Scholar
- Woods, J. W. Subband Image Coding Kluwer Academic Publishers, Boston 1991.Google Scholar
- Johnston, J. D. A Filter Family Designed for Use in Quadrature Mirror Filter Banks Proc. IEEE ICASSP 80, vol. 1, pp. 291–4, Denver, CO, April 1980.Google Scholar
- Kay, S. M. Fundamental of Statistical’ Signal Processing: Estimation Theory Prentice Hall, Englewood Cliffs, New Jersey, 1993.Google Scholar
- Tsern, E. K. and Meng, T. H.-Y. Image Coding Using Pyramid Vector Quantization of Subband Coefficients Proc. IEEE ICASSP 94, vol. 5, pp. 601–4, Adelaide, Australia, April 1994.Google Scholar
- Hung, A. C. and Meng, T. H.-Y. Error Resilient Pyramid Vector Quantization for Image Compression Proc. International Conference on Image Processing, vol. 1, pp. 583–7, Austin, TX, November 1994.Google Scholar