> > It sounds similar to stuff available for astronomers except there they > KNOW stars are really point sources. > you're talking about SGBNR? another nice program .. No, but it is a nice program for removing random noise. The general problem for any deconvolution function is determining the function to be deconvoluted. With a star field it is relatively straight forward: at a simple level if the stars are like bars you apply corrections that make them the best fit to points. Other non-point-source features in the image (nebulae) subject to the same pattern of camera shake are also improved. Photgraphic images are nowhere near so well defined. Not all features are at the same distance (so not subject to same degree of blur). Superimposed is a (probably) non-linear motion blur (camera shake) applied to the whole image. The other problem is the features in real photos are seldom well defined: they might not have sharp edges. If there is any feature that *should* be a point or a straight line, it becomes a lot easier to unravel the unknowns.. Basically the less the program has to assume the more accurate can be the result ... Now where is that reference ... Bob __________________________________________________________________________ Freeserve AnyTime - Go online whenever you want for just £6.99 a month for your first 3 months, that's HALF PRICE! And then it's just £13.99 a month after that. For more information visit http://www.freeserve.com/time/ or call free on 0800 970 8890
