propeller based spirit detector--no soldering required

rjo_

Hi guys.

Actually... it is supposed to be a one pixel camera... the "optic" is a 35mm slide viewer, with the photoresistor from the education kit placed in the image plane. Pixel intensity is measured by hooking the photoresistor up... as per the set up in the RC Decay Parts and Circuits in the PE-kit Counters lab.

The "camera" is hot-glued to a Parallax servo... which is hot-glued to the working end of another servo.

The acquisition is performed by loading the program to the EEPROM and pushing reset... once the terminal program is ready.
In the terminal ... you choose "capture text" from the "Transfer" menu... give it a name and the data is saved to that file... no size limit.

Once your image is saved... used ImageJ... a free download from Wayne Rasband at NIH...choose "import Text Image" from the "File" menu.
Then choose "invert" from the "Edit" menu.... and as if by magic... you can see the spirit world.

The attached image is of the corner of a room... you can clearly see spirit objects descending from the ceiling... either that or there was a light leak in my camera... (duct tape isn't opaque... for opaque you need to use electrical tape[noparse]:)[/noparse].





Rich

QuattroRS4

Rich,
I have to hand it to you - your posts are always entertaining but definiely obscure.

Regards,
QuattroRS4

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Necessity is the mother of invention'

Harley

Rich,

Obscure was a good word.

What was the 'camera' actually viewing? Looks like a blurry picture of some cable connectors, maybe. Didn't see anything resembling a 'corner of a room'.

What will be the purpose of a 'one pixel camera'? Does sound like a simple project.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Harley Shanko
h.a.s. designn

Forrest

Maybe it's time to upgrade from a 1 pixel camera to a optical mouse scanner www.engadget.com/2006/01/08/turn-your-optical-mouse-into-a-crappy-scanner/

Christof Eb.

Hmm,

this looks like an interesting experiment. Good idea!
Could you post a photo of the setup?

Are you going to use something with a smaller pixel area?
I could imagine, that the structure comes from the structure of the photoresistor itself?
Perhaps it is possible to cover the photoresistor and open just a litte hole of the cover to get a small pixel?

Christof

rjo_

Paranormal imaging[noparse]:)[/noparse]

I tried everything... except the optical scanner. But like the good scientist that I am, I generally changed more than one thing at a time. The problem is I didn't understand exactly how the photoresistor, the serial communications or the servo's functioned before I started... I also couldn't read electrical diagrams very well.

After studying the Propeller for a bit... this project took about an hour to put together... but a couple of days to test. Depending on the configuration... each pixel took about a second to acquire!!! That means that I had to wait for a cloud covered day to take this "picture." The "picture" is of two windows in the corner of a room.

There is a little problem in the servo logic...the first few moves look like my little camera hiccups between two positions... and then settles down.

By the way... I also don't read "specs" very well. I think it would help us non-engineers if Parallax would publish spec sheets for all of their components... for purely heuristic reasons.

I love my Propeller.

Rich

rjo_

OH... one more thing.

The value that the photo-resistor produces... depends upon the last value measured... so the spec sheet for this would require a formula!!!

Rich

Graham Stabler

cool but time to treat yourself to a small photodiode [noparse]:)[/noparse]

Graham

Graham Stabler

Also finding a neat way to scan the image plane would be good keeping the optic still.

Graham

Graham Stabler

Here is a way to get 2D motion from 2 servos with very limited hardware, you just need to ensure that the bearings prevent the "scissors" from rocking back and forth.

This parallel kinematic solution offers ultralow intertia movement with potentially very high resolution (especially if travel is small (keep arms short)).

Some trig will provide the servo angles required for a given position, actually it will give two solutions but you can hard code to select the correct one as you know your arms are pointing upwards. To go to a positon you set the servos to the correct angles, to follow a straight line you must divide that line into intermediate postions which you move too or the line will not be straight, for scanning its easy you just plug in the co-ordinates for each position.

Here is the 3-axis equivalent:

www.youtube.com/watch?v=Du2f-EUDqio&mode=related&search=

Graham

Post Edited (Graham Stabler) : 7/25/2007 9:33:41 AM GMT