Abstract
Image and video compression standards such as JPEG, MPEG, H.263 are severely sensitive to errors. Among typical error propagation mechanisms in video compression schemes, loss of block synchronization causes the worst result. Even one bit error in block synchronization may result in data to be placed in wrong positions which causes spatial shifts, the worst error one can see with naked eye. Traditional Forward Error Correction methods add significant redundancy. In this thesis, an improved method over the Error Resilient Entropy Code (EREC) algorithm, originally designed by Kingsbury and Redmill, is proposed. This thesis shows that the new method increases computational efficiency immensely such that the new method performs about 200 times faster than EREC for encoding and about 300 times for decoding, based on H.263 CIF picture format. The new method also achieves a higher resolution of reconstructed images statistically without redundancy. Also, SOT (Spatial-Orientation Tree) code of wavelet coefficients combining with the proposed method will enhance compression compared to DCT-based methods.
Min, Jungki (1999). A new and efficient error resilient entropy code for image and video compression. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1999 -THESIS -M56.