Subblock sum matching algorithm for block-based interframe coding
作者:
Highlights:
•
摘要
A fast block-matching algorithm for motion estimation is described for interframe image coding. Utilizing the fact that the motion is searched by a macroblock basis while the compression is performed by a block basis in many video compression standards, the proposed algorithm produces high PSNR and fast computation while keeping the bit-rate as low as those of conventional fast algorithms. The proposed algorithm uses the subblock sum as a matching parameter and the reversed square sum as a matching criterion. The new matching parameter and criterion reduce the total number of computations for block matching and also keep the quality of the decompressed image. The computational complexity and the compression performance of the proposed algorithm are compared with those of other block-matching algorithms.
论文关键词:Block-matching algorithm,Subblock sum (SBS),Bit-rate,Macroblock,Reverse squared sum (RSS)
论文评审过程:Received 28 August 1997, Available online 12 January 2000.
论文官网地址:https://doi.org/10.1016/S0923-5965(98)00054-X