Vector quantisation: a transputer implementation
S. Omarouayache, J. Ngwa-Ndifor, Tim J. Ellis · Research Repository (Kingston University London) · 1992
The paper considers the problem of implementing image compression algorithms based on vector quantisation. A transputer based multi-processing approach is proposed which enables images to be encoded and decoded at fast rates. Results are presented for the performance of the algorithm on various processor topologies, and the near linear speed-ups, indicate that further improvements in computation rates are achievable.