Best Off-the-Shelf Sensor for Detecting Green LED vs Red or White in Industrial Setup

We're having trouble finding a color-sensitive sensor that will detect when a green LED comes on. We don't want it to trigger when either of the other 2 colors (red or white) glow. We've found color sensors, but they're designed for reflected light (for example, checking the colors on a printed carton) rather than emitted light.


This is for an industrial application. If it works, we hope to implement it in about 100 stations throughout our factory. Ideally we'd like an off-the-shelf solution, but if required, we can kludge together an assembly of parts.

I'm not sure about a color sensor, but you should be able to take something like a raspberry Pi, a web camera and the OpenCV libraries to detect the color. A quick overview I found is on a blog at https://www.pyimagesearch.com/2016/02/15/determining-object-color-with-opencv/

Kevin.... you can use a LED reverse-biased as a detector for its own colour. Google "LED as sensor" for more info.   From this article I got: (my bolding)As a photodiode, an LED is sensitive to wavelengths equal to or shorter than the predominant wavelength it emits. A green LED would be sensitive to blue light and to some green light, but not to yellow or red light. For example, a red LED will detect light emitted by a yellow LED and a yellow LED will detect light emitted by a green LED but a green LED will not detect light emitted by a red or yellow LED. All three LEDs will detect “white” light or light from a blue LED.
Unfortunately almost any LED will respond to white light as it contains all colours.  I don't think you'll get round that too easily. Maybe have 3 LEDs (R/G/B) and have a small microcontroller to decide what light is coming in, if you only get green or green/blue on then it's green light, if you get all 3 then it's white light...

Kevin there is also a camera just out called the Pixy 2 which can detect colours (and shapes) and report back to your uC on what it is seeing.  A bit of a review here.   That would certainly do what you want, but at around $80, how cost-effective it would be I don't know.

You can find sensors capable of detecting primary colors individually at https://ams.com/light-sensorsAMS includes the old TAOS sensors. One AMS color sensor has quadrants, three  of which sense separate primary colors.  They are light-to-frequency converters.  The outputs are compatible with digital GPIO.  Higher intensity results in a higher frequency.  The waveform is essentially a square wave.

 Utilize a LED switch one-sided as an indicator for its own shading.As a photodiode, a LED is touchy to wavelengths equivalent to or shorter than the transcendent wavelength it discharges. A green LED would be delicate to blue light and to some green light, however not to yellow or red light.For instance, a red LED will distinguish light discharged by a yellow LED and a yellow LED will recognize light transmitted by a green LED yet a green LED won't identify light produced by a red or yellow LED.Every one of the three LEDs will distinguish "white" light or light from a blue LED. Shockingly any LED will react to white light as it contains all colours.I don't think you'll get round that too effortlessly.Possibly have 3 LEDs (R/G/B) and have a little micro controller to choose what light is coming in, in the event that you just get green or green/blue on, it's green light, in the event that you get each of the 3, it's white light...

Adafruit offers a RGB sensor board for about eight dollars based on one of the AMS sensors.  Adafruit added a white LED so that it can be used for reflected light as well as emitted light.  The Adafruit writeup emphasizes reflected light, but it could also be used for direct light.  Sensors designed for reflected light need a light source, but the light source need not be used for the function you describe.  The sensor has an adjustable range of sensitivity from low to millions-to-one.  High sensitivity is needed for the reflected light.  If the lowest sensitivity setting doesn't work, try adding a neutral density optical filter.  Neutral density filters are optical-speak for attenuators and come in a variety of values.

Imagine looking at a printed red mark on white stock. The red mark looks dark in contrast to the white paper. Now imagine placing a red transparent filter in front of your eyes while looking at the red mark. The mark now becomes difficult, if not impossible, to see. A sensor equipped with a red LED would have the same problem and, most probably, would not register the mark.

This looks like a good solution. Pair this with an Arduino Nano or some such and a bit of software and you would have an inexpensive solution.

I'm intrigued. This sounds really interesting, I'm going to have to try this out over the Christmas holidays.