Right now I am building a system to recognize blood cell, i.e. red cell, white cell and the element inside it. Does anybody know what kind of Artificial Algoritm should I use ? Thank you Xastro

Difficult to say without more detailed information but neural networks are commonly used for image recognition and classification.





Stimulate

I think you''ll find it''s harder than you think...

You''ll probably need some sort of image processing software, assuming you have images of some sort. At a minimum you might need to remove noise from the picture, using something like a median or blur filter. Most image recognition and computer vision stuff will also run equalisation, thresholding, pseudo coloring, or any of a number of other processes on the image to make it more clear and distinctive.

Then you can process this data with your AI technique. As fup says, neural networks are commonly used for this sort of thing. You might get some milage out of some sort of expert-system, but that would require some good image-processing to generate some high level facts from the images in the first place. Alternatively you might find fuzzy clustering could suit your needs, as it performs similarly to some sorts of neural networks. Take a look at http://www.fuzzy-clustering.de/ for a few more details.

I am not an expert on any of the above, so no doubt Someone Who Knows What They''re Talking About will come and correct me soon, but there''s still a few points to be researching while you wait.

[ MSVC Fixes | STL | SDL | Game AI | Sockets | C++ Faq Lite | Boost | Asking Questions ]