recognize 10 color.....

Go To Last Post
24 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hi.
i want to ratio 10 color to 0-9 something like this:
0=white
1=green
2=red
and..........

is there any sensor or way to recognize 10 color?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

i suppose you can use RGB LED, photo-resistor/transistor/diode and ADC - different colours made different reflection intensity for each color

Computers don't make errors - What they do they do on purpose.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

is there any sensor or way to recognize 10 color?

You can get full RGB with a Taos sensor.

http://www.taosinc.com/productfa...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

thanks for answers.....
is TAOS recognize only red-green-blue?!
because i want recognize 10 colour not 3.....if it can recognize combination of this 3 color my problem solved:)...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

When I went to school, colours other than R G or B could be made by mixing together different amounts of each (but in my experiments I always got brown :()

So surely your other colours are just different proportions of RGB that you get from the sensor - the different Vor, Vog and Vob voltage outputs that it shows in the datasheet (which you have looked at?)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Will you settle for 8 colors, 0 to 7? That makes it easier. RGB color 000 is black and 111 is white and there are of course, red, green and blue, and 3 more 2 color combinations.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The Taos chips are pretty cool, IMO, when you just want to add color recognition/processing to an app. Reasonably priced and saves a lot of research and design work to develop a from-scratch solution.

I see they have a new generation now. Stock at Mouser seems to be of most models on that page.

I used it in a test fixture for an app that had red/green LEDs for indication. With some light pipes it worked great for that.

They were in my "toy box" for prototyping an app to determine the concentration of a chemical in water, and the chemical added a blue tint to the water. The prototype seemed to work OK but the project died. ;) (Or since we are discussing color, the project "dyed"? groan)

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
When I went to school, colours other than R G or B could be made by mixing together different amounts of each (but in my experiments I always got brown Sad)

:D:D:D........

i see this:
[img]
http://up.vatandownload.com/imag...
[/img]
what is mfp?!what am i doing with this pin?

and it has vg-vr-vb if same as bobgardner said:
output is 3 bit each vout means 1 bit we have only 8 option 000-111 is this correct?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

is there any other sensor cheapest?! i see TCS3103FN for 2.49$ !!!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

RGB are not single frequency points but overlapping peaks along a continuum. By adjusting for peak sensitivities and overlap you get a reasonable analysis of the entire color spectrum. With the Taos (and others) sensors you can recognize 10 or 10,000 colors depending on your ADC and the overall accuracy of the sensors.

Smiley

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

thanks smiley...so if i use this http://www.taosinc.com/ProductDe... because that output is Frequency i have not any problem for 10 color..is it correct?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

With the TAOS sensors, you can probably recognize hundreds of colors, if not more. Any given color is simply defined by the ratio of the red, green, and blue components (plus maybe a measure of gross intensity). So, connect your micro to one of those nifty sensors with I2C/TWI, write the software to retrieve values, and categorize the results into your 10 color "bins". You can label the bins "A,B,C,..." or "1,2,3,.." or anything your heart desires (maybe even labels like mauve, chartreuse, magenta, ...). Its up to you. Report the bin name to the outside world.

By the way, 2.495 is a bargain for such a sensor!

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

and categorize the results into your 10 color "bins".

Remember that in practice, you are unlikely to get any "pure" colors--e.g. even when looking at a "red" card with good pure-white lighting, you aren't going to get e.g. max R and no counts on G & B.

Hmmm--10 colors, eh? the rainbow, plus white and black?

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Color is a subjective term that describes the typical human eye response to an input stimulus. Yellow for example could be a single-wavelength spectral yellow, a combination of several wavelengths such as red and green, or a black body spectrum from an incandescent source heated to ~2000K. Without characterizing the source of color your device might decide yellow when the input is actually red and green (are simultaneous inputs allowed?)

A general solution would be a prism focused on a linear CCD detector, backed up by some processing power.

Narrowband filters in front of phototransistors would work for discrete wavelength sources.

If you just want to characterize a single input as a color, three sensors with overlapping response can be calibrated into a 3x3 matrix which will approximate chromaticity http://en.wikipedia.org/wiki/Color_temperature. Again, color is a subjective sensation and if the sensor response differs from that of the human eye the results will be off in places.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Agree on lack of "pure" colors in practice. That is where the categorization comes in.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

thanks jim...

Quote:

By the way, 2.495 is a bargain for such a sensor!

yes i compare 2.49 with 49$ in this linkhttp://www.parallax.com/Store/Se... ...
my Purpose was cheapest than 49$ not 3$..2.49$ is really good for this sensor..

i dont understand,,what exactly send this sensor via TWI or i2c?
thanks for guide me

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

i dont understand,,what exactly send this sensor via TWI or i2c?

They have models with voltage output, frequency output, and digital I2C output. As with >>any<< other peripheral chip, >>you<< decide what best fits your needs.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
is TAOS recognize only red-green-blue?!
And the monitor you are looking at only produces red-green-blue.

Regards,
Steve A.

The Board helps those that help themselves.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A cheap alternative is the already mentioned RGB LED (or 3-4 separate LEDs) and a phototransistor / photodiode with a suitable amplifier and the µC to do the AD conversion and math. Just record the reading from the photodiode without LED and one of the LEDs at a time. To suppress noise and background light it may be a good idea to do many of these measurements. The LEDs spectra here essentially replace to color filters of the dedicated color sensor. As a reference one usually needs a white sample and may be a second black too.

The taos sensor may also need a light source, and here a white LED is not a good idea, because of its spectrum.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

The taos sensor may also need a light source, and here a white LED is not a good idea, because of its spectrum.

IIRC when I was doing my blue-tint project, we obtained some white LEDs that were made for this type of illumination and had the "right stuff". As I recall they were a little more costly than "plain", but not way more expensive. In my app just looking for blue tint it didn't make much difference--taking readings from the outside of a section of glass tube filled with moving/stationary fluid and backlit isn't going to give you "pure" results but worked well enough to get "how blue is it compared to the reading of the same setup with clear water moving through it".

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The normal white LEDs give very little green light. So they are not good to recognize colors, both by eye and a sensor. There are a few white LEDs with a more even spectrum.

Using just a RGB LED can actually have a similar problem - here its more yellow that may be missing. If its just a "teach in" problem to recognize a given set of about 10 different color samples all this does not really matter.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

thanks all.....
i bought a TSC230.... how can i read frq from TSC230?
i write my program by C....

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Don't see TSC230 at TAOS. There is TSL230. Texas Instruments makes a TSC230 which is a touch screen manager.

The TSL230 is just a "generic" light sensor. It does not have the color filters you need for color detection.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!