I have a small program that I'd like to use two level of parallelization. I'm doing some image analysis where I'd like to have the image analsys being parallized and the search through a collection of images.
Currently I have a custum C++ class that does the image manipulation on each file I'm looking at. so.. basic pseudo code main: get src image process image compare src to dst (and repeat for 1 of 1000 different dst objects) the search seems easy enough to parallelize, but I would need to do the image analysis split among processors. Would there be any problems with having MPI initiated and finalized within a class? sorry if this is confusing, just wanted to run my ideas by some far more capable individuals, making sure my approach makes some sense. Thanks again, -- Samir
