2005, Vol.8, No.3, pp.302-307
Three kinds of transforms for pattern recognition in
morphology are studied. The recognition processes for these
transforms are computer simulated. The conditions applied for all
transforms are established. From all of theese transforms, ROHMT
is found most effective. Pattern recognition is implemented using
an optoelectronic image processing system. The advantages of ROHMT
transform are demonstrated.
Key words:
mathematical morphology, pattern recognition, hit-miss
transform
Full text:
Acrobat PDF (294KB)
PostScript (5759KB)
PostScript.gz (596KB)
Copyright © Nonlinear Phenomena in Complex Systems. Last updated: September 8, 2005