Solving the Region Growing Image Segmentation Problem .....

This paper sddresses a well known and difficult parallel
computing problem. Its practical value is not so clear but
undoubtedly it is an intellectually challenging area.
The paper is well enough written as far as it goes but I am doubtful
about publication for two reasons
1) There was no discussion of previous work and all the references are
 quite old.The authors should be more dilgent and use citation searchs
 and such techniques. They could start with Copty's Syracuse thesis which
 extends the cited paper [1].
2) The authors only consider one image and that is an essentially trivial one.
 Any problem for which region finding was important would be far messier
 and far more complex. It would probably perform much worse on a parallel
 machine (which is one reason why this area has not been a forefront of research.
 Another is that in practical cases one is usually processing very many images
 and parallelism over these image (not regions in one image) is trivial and
 very efficient). The authors should consider several representative images.