RE: [CAD_CAM_EDM_DRO] some more milk digitizing ideas

As long as you know how much fluid you are adding each time, you (meaning the program, not you personally) can calculate how thick each "slice" is. As long as there are no sudden changed in cross-section (i.e. no surfaces parallel to the surface of the milk), you can approximate the slice thickness by observing that for a small amount of milk added, (milk volume added) = (background area) X (slice thickness) where (background area) is the total surface of the container less the area of the last slice. This is just the calculation of volume as the integral of (slice Area) X (infinitesimal slice thickness) approximated for a small (but not infinitesimal) slice thickness.

The background area is easy to calculate. Just make sure the entire container area is visible to the camera and count the number of pixels that are milk-coloured.

The inaccuracies involved are the error in calculating the area (poor focus, image distortion, and quantization due to pixel boundaries) and the error introduced by using non-infinitesimal slice thicknesses (which is worst when object surfaces are parallel to the milk surface).

Of course there will also be a scaling factor which depends on the resolution of the camera, the units used to measure the milk volume, and the units you want the slice thickness to be measured in.
-Kevin Martin

-----Original Message-----
From: CAD_CAM_EDM_DRO@yahoogroups.com [mailto:CAD_CAM_EDM_DRO@yahoogroups.com] On Behalf Of Abby Katt

[...]
The other issue of course is how to measure the Z-value, as due to the volume of the object adding a known amount of fluid will move the actual level in a "profile".
I guess my thinking is more down the lines of how to automate it, you could just glue a ruler to the side and add water by hand otherwise - but that wouldn't be CNC or on topic anymore! :)