Re: anybody know?
hi,sorry guyz...
there are three different locations
for a run-length coder depending on the sorts of runs it aims
to code: front (Run-Length-Front: RLF), middle (Run-
Length-Internal: RLI), back (Run-Length-Back: RLB).
the RLI coder is located between the
modelling and the coding stage. The purpose of the RLI
coder is to code not repetitions of symbols as in the previous
case but repetitions of dictionary locations. This means that
the RLI coder monitors the output of the modelling stage
and, when the same match location is output more than
once, a single RLI code is used to replace the stream of
identical dictionary indexes. The advantage of this mechanism
is that, without adding information to the flag, RLI
coding is not limited to a particular symbol repetition but it
can code any repeating symbol, since a set of identical
symbols will generate an identical string of matches at the
same dictionary location.