Digitizer algorithms (mine worked)

After spending 9 hours digitizing a 45mm*13mm area (containing a
plastic propeller blade) with my Proxxon and home made touch probe I
have started to think about optimising the digitizing a little (tht
was over 13,000 digitized points.

Turbo CNC simply pecks a bed of nails from a Z datum height. The
problem is that my propeller for example was high and thin in the
middle and wide and lower at the tips. The area of prop in the
middle was very quick but the areas to each side (i.e. the not
propeller bits) took ages as the probe went all the way down to the
base plate. The wider lower parts took longer as the probe had to go
lower than the middle section but it did at least fill the area in x.

So to reiterate the problem is the variation in heights of the object
and the need to lift to the maximum height each time.

Enter my fast scan algorithm, I say mine but I have probably re-
invented it.

The probe lowers and until it hits, the x,y,z are recorded. The
probe lifts until the contact is broken and then a small offset
(about 0.5mm in my case) The probe then moves sideways to the next
probe position. If it contacts sideways before reaching the
destination then the probe is lifted to break contact, a smaller
offset added and it moves off again. In this way it moves straight
to the next point for flat surfaces and dipping surfaces but climbs
when the surface becomes higher. In practice because of the small
inter sample size and the 0.5mm offset it rarely needs to do the
ramping part but it does work quite well and is quick.

With bit of tweaking the algorithm should be great for propeller
scanning where there are narrow high areas and low wide areas,
especially if the offsets are chosen carefully.

So I wrote the program to do it and it works! Looks rather neat
while it runs as well. I haven't added the saving to disk bit and it
has no GUI and no jogging and no error checking or other stuff to
prevent probe destruction but it does step across surfaces briskly
displaying the x,y,z co-ordinates as it goes.

The time it took to get working (an evening) is more a testament to
step and direction control and code reuse than to my programming but
I'm pretty pleased. It was fairly scarey though running it the first
20 or so times. I was scanning rather soft plastic items just in case.

The point of my post?

1. This works

2. And, are there other algorithms or methods like this that are used
in industry that might be better?

Graham

p.s. If the offset were made small the probe could be mounted next to
a spindle but on the same z axis and the system used for
duplication. Probably rather rough duplication but it might suit
some apps.