Compression and decompression of reference frames in a video decoder
AssigneeSharp Laboratories of America, Inc.
InventorShaw-Min Lei
Life of this patent
Estimating the term…
Loading patent…
AssigneeSharp Laboratories of America, Inc.
InventorShaw-Min Lei
Estimating the term…
Methods and apparatus for compressing and decompressing reference frames in a video decoder present a low-complexity approach to conserving reference frame memory. At the same time, this approach provides PSNR and drift-reduction performance comparable to competing methods of relatively high-complexity, allowing typical compression rates of 2:1 to 8:1 to be used with good performance. In a preferred embodiment, reference frames are blocked, and each block is Haar wavelet transformed. The subbands of the wavelet transform with similar expected bit-level run-length statistics are grouped together, and each group is embedded coded using Golomb-Rice codes with a parameter look-up adaption method. The embedded code is stored in a compressed frame memory, indexed by block, for later retrieval and decompression when the reference block is needed by the video decoder.
A method for compressing a digital image, the method comprising the steps of:
A method for storing and retrieving reference frames in a digital video decoder, the method comprising the steps of:
A digital video decoder having a reference frame storage and retrieval subsystem comprising:
Dependencies are inferred from the claim text.