Skip to Main Content
Summary form only given. The application of vector quantization (VQ) has, to a great deal, been constrained because its encoding process is very heavy. The paper presents a fast VQ encoding algorithm called the multiple feature-ordered partial codebook search (MFPS) algorithm. The proposed algorithm uses the Walsh-Hadamard transform for energy compaction and a look-up table for a fast reference. Simulation results show that the proposed MFPS algorithm is faster than the existing search algorithms. Compared with the exhaustive full search (EFS) algorithm, the MFPS algorithm reduces the computational complexity by 99% for a codebook size of 1024, while maintaining the same encoding quality as that of the EFS algorithm.