One way of detecting symmetric objects is convolution. Bubbles are pretty symmetric, so if you do a convolution of the image with itself, you should get peaks where the bubbles are. On Mon, May 17, 2010 at 8:34 AM, Sarin Sukumar A wrote: > i tried these but the edge detection algorithms are not able to detect the > edges well like closed figures, it will detect some bright portion of the > object, what i am saying is impossible or not developed yet? > YOurs SaRIn.... > > > On Sun, May 16, 2010 at 10:15 PM, Jake Anderson wrote: > >> Sarin Sukumar A wrote: >> > I have a project where i want to detect and count bubbles in a test tube. >> > I am using opencv; i used hough circle detecting algorithm. But its of no >> > use for this purpose >> > Any one know any kind of algorithm to count specific objects or any shape >> > analyzing techniques? >> > any kind of information or ideas are welcomed >> > thanks in advance >> > YOurs SaRIn.... >> > >> I don't know specifically about bubble counting but my first blush at >> how to do this would be to run an edge detection filter, smooth it >> somewhat with a simple kernel, perhaps one slightly weighted to make >> circles of whatever the average diameter of your bubbles are to fill in >> any 1 pixel gap type issues. Then go through and count the "holes" in >> the resultant image (look for closed polygons). >> -- >> http://www.piclist.com PIC/SX FAQ & list archive >> View/change your membership options at >> http://mailman.mit.edu/mailman/listinfo/piclist >> > -- > http://www.piclist.com PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist > -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist