YOurs SaRIn....


On Wed, May 19, 2010 at 6:01 PM, Olin Lathrop
>
>
> Of course, what did you expect?  What you did was a special kind of
> convolution called a "auto-correlation".  You always get the maximum value
> in the center.  When you do auto-correlations for the right reasons, you
> are
> usually looking for peaks in other places to get a idea of repetitive
> signals.
>
yes i tried the method today, Yes i do get okey results , But i am just
thinking to
achieve better results. i am going to write a code for that with the help of
opencv, i read that harr training method exists, by which i can train the
algorithm to detect and count bubbles i will try this also. but by a logical
thinking we can easily count them using our eyes, then why cant i try to
implement that logic ? any way i decided to give a try i will keep on
publishing the progress....

>
> In any case now that you've shown us a picture, I don't think anything as
> heavyweight as a bunch of convolutions is necessary.  I would first play
> with the lighting to get maximum contrast between the background and any
> bubbles.  It looks like OK results could be obtained even with your current
> setup, but a little tweaking should decrease errors.
>
>
> ********************************************************************
> Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
> (978) 742-9014.  Gold level PIC consultants since 2000.
> --
> 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