In this paper, we present an approach to increase the amount of Ve cod
ebook patterns to improve image quality without a large overhead on th
e codebook memory and search time. Basically, our approach characteriz
es a pattern into a key value, which is used in codebook search; ''Ref
lection operation'' was performed on codewords using the key values to
get more patterns to improve image quality. Our basic Ve system is a
2-step ye where Ve is used to compress prediction errors resulting fro
m simple scalar quantization of pixel blocks. The codebook search time
for our technique is O(logN+K) for each input vector, where N is the
codebook size and K is the Voronoi's range. Experimental results are p
resented in the paper.