1096Views16Replies

Author Options:

Pin-Pressions/ Rotation Sensor Answered

I'm trying to build a sensitive pinpression device: one that can see 3d digital images. The device would be able to read, record and organize the depth of each pin. I think the first problem to overcome is figuring out a cheap way to measure pin depth. I think it will be best to do this with rotation sensors. Does anyone know how to build or where to buy cheap rotation sensors?

Discussions

0
None
burningsuntech

10 years ago

One possibility is using a laser with either a moving galvanometer mirror or a spinning multi-faceted mirror and bounce the beam off the pinheads then reading the reflection with a photodiode. But if you're going to extremes, why not just bounce the laser off the surface your recording with the pins in the first place? Just a thot

0
None
Patrick Pending

10 years ago

The complexity of a device to measure each individual pin with its own individual sensor would be extremely high. The cheapest, simplest, method would be to record the pin data individually with a digital depth gauge. Is there a particular reason why you are trying to construct a face from a pin impression rather than use some other method? Cheers, Pat. Pending

0
None
yesPatrick Pending

Reply 10 years ago

Well, I'm actually interested in mapping the whole body: I'm interested in pin impressions specifically, because I think they formalize an essential? way of thinking about sight. This device would "see" through "touch". I think that's nice. Also, full disclosure, this is the beginning of a larger project that may make sense to pursue with light based 3d scanners: eventually I want to build clothing from "skinning" and cutting the resultant digital model with digital tools. Does this make sense?

0
None
Patrick Pendingyes

Reply 10 years ago

The practicalities of making a homemade device with that density of electronic sensors would be somewhere between extremely daunting to nigh-on impossible. If you managed to get your sensor device built you would then face the extremely complex task of interfacing all those sensors and collecting and processing the data.

However, you can map a face with just 6 points (or even less*):

1 & 2) Assume head is an ovoid. Measure chin and left side width

3) Measure left eye center point

4) Measure tip of nose (you could skip this if you were modeling M.J.**)

5) Measure center of mouth

6) Measure left ear

Now reflect points for other side of face and you have a very rough (albeit symmetrical) starting point. This procedure can be refined by adding the required level of detail (e.g., nose width, cheek height, eye width, etc.). Most faces are put together the same way, it's just the proportions that vary. By measuring only the key points you would save a lot of data processing and complexity.

  • - One I mapped earlier >> ;-)

** - M.J.>> :)

Cheers,

Pat. Pending
0
None
yesPatrick Pending

Reply 10 years ago

Also, is this the appropriate section of this site to post questions of this nature? Thnx

0
None
Patrick Pendingyes

Reply 10 years ago

Yes, I think so (it will also appear in the All forum where most people will see it). It would have been less confusing if you had provided more detail of what you are trying to achieve, but I see you have rectified that now. Cheers, Pat. Pending

0
None
gmoon

10 years ago

Several different types for sure.

All (mechanical) computer mice have two optical rotation sensors--could you mod one of those? They sure have cheap goin' for them...

0
None
zachninmegmoon

Reply 10 years ago

I agree, those two would be the best bet.
Heres how they work:
You have the 2 sensors, when you rotate them, they'll go in this order.
(1 = on, 0 = off)
0  01  01  10  1
Or... in reverse:
0  00  11  11  0
You'll have to figure out which is which, but from there you're good.
0
None
Patrick Pendingzachninme

Reply 10 years ago

Are you proposing one sensor for each pin? It's just that I count in excess of 4000 pins on the one in the photograph! Maybe the Topic poster could suggest how many points are required to map the face. Cheers, Pat. Pending

0
None
yesPatrick Pending

Reply 10 years ago

Yes, proposing one sensor per pin. The resolution doesn't have to be fantastic. Somewhere on the order of 1000 pins. Give or take half?

0
None
gmoonPatrick Pending

Reply 10 years ago

I'm missing something--what face? what photograph? Not much in the OP other than measure pin depth..

0
None
Patrick Pendinggmoon

Reply 10 years ago

This was posted in forums - PinPressions and there is a photograph of a pin-impression face - PHOTO

Cheers,

Pat. Pending

0
None
gmoonPatrick Pending

Reply 10 years ago

Ah, thanks, Pat. P. That looks like a real challenge...

0
None
NachoMahma

10 years ago

. I'd use a linear potentiometer.

0
None
Goodhart

10 years ago

Hmm, my first though was "stepper motor mod" but that could get complicated. The old stators (rotors) in automobile distributors, had a sensor in it of this sort, but it was cam driven.