Society of Robots - Robot Forum

Electronics => Electronics => Topic started by: airman00 on November 25, 2007, 04:38:19 PM

Title: Green LED and IR
Post by: airman00 on November 25, 2007, 04:38:19 PM
I just had an idea


If you use an IR transmitter / reciever for line detection , you could shine a green LED or maybe red LED on the floor. Then the IR  would detect different shades of green or red , perhaps lowering sensor noise from sunlight  and probably giving better readings!


I'll do some experimenting this week ( or maybe even tonight) .

Has anyone done or heard of anything like this before?


Title: Re: Green LED and IR
Post by: Admin on November 25, 2007, 04:50:35 PM
but . . . green light isn't IR light . . .

look into something called modulation, where the sensor looks for pulses of light at a certain frequency only, subtracting the light that isn't pulsing (environmental interference).
Title: Re: Green LED and IR
Post by: airman00 on November 25, 2007, 05:18:01 PM
I know that green light is not IR light


I was thinking - if IR can detect colors , then if I were to shine a red light or something then the IR would detect different shades of red, allowing better detection


whatever, ill experiment ,see what happens
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 06:58:16 PM
Quote
I was thinking - if IR can detect colors

Quote
but . . . green light isn't IR light . . .

 :) :) :) :) :)

IR can only detect ir type 'colors' not green, yellow blue etc.. you might get lucky and have the red led change the detection slightly since visible red and infrared overlap slightly on the spectrum
 
Title: Re: Green LED and IR
Post by: airman00 on November 25, 2007, 08:05:57 PM
But IR can detect colors!

With my line following sensor I saw very very different number ranges for different colors!
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 08:20:48 PM
sometimes, infrared light will also be emitted with other light colours. This is a by product of a light emitting devices inefficiency. Any use of this type of system will be inefficient and will vary a lot from product to product. so 1 red led might produce 2% ir light but try with a different red led and the results will change.

ir emitters / detectors generally work on a wavelength of around 800 - 950 nm a green led has a wavelength of around 565nm and so it is not within the range of the ir detector, only some kind of by product would be and that by product is an inefficiency and cant be trusted to provide accurate data
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 08:32:57 PM
if you really want t0 try and improve ir detection, you could maybe investigate polarizing filters for ir light, I think this would make a great supplementary tutorial....

(especially if you can explain the physics behind it)
Title: Re: Green LED and IR
Post by: airman00 on November 25, 2007, 08:37:30 PM
By colors I meant green colored materials, not green light

It can give a consisted range of values for black , for white ,and for green. ( thats which colors i tested)
Title: Re: Green LED and IR
Post by: Admin on November 25, 2007, 08:52:22 PM
have you read this yet?

http://www.societyofrobots.com/sensors_color.shtml
Title: Re: Green LED and IR
Post by: airman00 on November 25, 2007, 09:06:12 PM
Yes , now my problem is that I want to use the line following sensors to see black , white ,and green.


Is it better to do that or just to add a photoresistor to the bottom with a green LED for green detection?
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 09:08:31 PM
the problem is that different materials will produce different ir reflections. it doesnt matter about the colour. green paper and green plastic will produce 2 different ir readings, you are probably best with a phtoresistor
Title: Re: Green LED and IR
Post by: bens on November 25, 2007, 09:10:14 PM
Your IR detector isn't detecting "green", it's detecting the IR absorption/reflectance of the particular green material being used (which is why you cannot use a green LED as your light source, because otherwise your IR detector will have virtually no signal to detect).  If you want to be sure you're seeing green, you should use a photoresistor.
Title: Re: Green LED and IR
Post by: Admin on November 25, 2007, 09:10:47 PM
I've said this a million times before . . . but . . .

I really don't like using IR for the simple reason that it's hard to debug what you can't see . . .
Title: Re: Green LED and IR
Post by: bens on November 25, 2007, 09:13:06 PM
I really don't like using IR for the simple reason that it's hard to debug what you can't see . . .

It can be annoying, though you can often use cheap digital camera displays to see IR (also, you can buy some paper that glows when IR light hits it--I used this stuff to align arms of some IR laser interferometers).
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 09:15:19 PM
sorry for ganging up about the IR, but it can only see IR colours... :)
Title: Re: Green LED and IR
Post by: airman00 on November 25, 2007, 09:16:36 PM
  If you want to be sure you're seeing green, you should use a photoresistor.
Thank You, photoresistor it is .


One more question  :P

IF I need to detect aluminum foil tape , what color LED would I use , white?
(http://www.aircraftspruce.com/catalog/thumbnails/aluminumfoiltape.jpg)

For detecting green AND white I would need one photoresistor with a Green LED and a ???? led.
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 09:18:56 PM
you could always use normal colour led's to also detect colours. It isnt a much used thing with led's but it does actually wok on an a2d port
Title: Re: Green LED and IR
Post by: airman00 on November 25, 2007, 09:21:25 PM
you could always use normal colour led's to also detect colours. It isnt a much used thing with led's but it does actually wok on an a2d port

Please be a bit more specific...

Meaning use only one green LED and a photoresistor to detect two colors?
Title: Re: Green LED and IR
Post by: bens on November 25, 2007, 09:25:18 PM
paulstreats means you can use LEDs as color-specific photodetectors by measuring the currents they generate as they absorb photons.  A green LED will generate current when hit by green photons, a red led when hit by red photons, etc. 
Title: Re: Green LED and IR
Post by: airman00 on November 25, 2007, 09:27:23 PM
Ahhh OK


but what color LED would I use to detect aluminum foil?
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 09:27:31 PM
meaning that you can use a green led to emit light and also a green led to detect light within the green leds specified wavelengths. so if you use a white light to emit, then a green, red and blue led to detect you can see any visible colour with a bit of programming . ps. yellow is a mix of green and red
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 09:29:45 PM
the only way to detect a specific substance is to use the rgb system, and get calibration data fiirst. then store the calibration data, and when your bot detects the same data it means it is detecting aluminium foil
Title: Re: Green LED and IR
Post by: bens on November 25, 2007, 09:31:26 PM
but what color LED would I use to detect aluminum foil?

Conductors reflect light.  How well they reflect depends on their conductivity and their thickness.  Aluminum tape will, for the most part, very effectively reflect all wavelengths of light in the visible spectrum, which means it doesn't matter what color LED you use as your light source.

For the most part, aluminum tape will look a lot like a white surface, except that the reflection angle off of aluminum will depend on the angle of incidence of the incoming light.  White surfaces absorb incoming light and then reradiate it in all directions, so there isn't any one direction of strong reflection.
Title: Re: Green LED and IR
Post by: airman00 on November 25, 2007, 09:36:06 PM
I understand now

but still avoiding my original question:


I want to detect when there is green tape on the ground and when there is aluminum tape. Any other color I do not care about.

What would I need to do this ?

From the previous posts I'm guessing one green LED ,one photoresistor, and one red LED.
Correct?


I want to thank you guys for your help.  ;D  I now understand the whole color thing , and I just want to make sure that I'm doing it right.

I am working on a robot for a competition and the line sensors are not detecting the green tape and aluminum tape properly. Everything else is basically finished. Stupid colors ( which are just a figment of my imagination ,lol)
Title: Re: Green LED and IR
Post by: paulstreats on November 25, 2007, 09:40:36 PM
i would use 1 green led, 1 red led and 1 blue led.
if you calibrate  your system with the aluminium tape, you will get a reading from all 3 leds, this reading will be around the same all of the time for aluminium foil.
To detect something green is easy, if yu have a white light shining on a green material, the green led will produce a greater output than the red and blue leds