Color sensors would be a cool product to have. any thoughts?
There already is a light sensor.
I tested it out about 2 years ago to see the distinction between the red and blue tile and there was one if I remember correctly, so each color would have a different value. I might test it out again and report the results.
Next year I might try to program the robot with the light sensor to know what color of tile I’m on while the robots are disabled so I don’t have to choose on the LCD. I’m surprised 1103 didn’t do this in Round Up, or did he?
yeah im planning on doing some tests with the light sensor to decide if there is a big enough difference from red to blue. if you do tests please post the results. Thanks!
From the toss up manual,
Hope that helps!
Hmm, that rule make it’s even easier! Easier than a light sensor would be just a bumper switch. Why take the easier route though
i have a specific application in which a color sensor would be beneficial. ill do some tests with light sensors and color filters
From my experience, red and blue look very similar under the light sensor.
Also, I have heard that light sensors have issues in the low lighting of the vex dome…
Anyone have experience with this?
Yes but if you use a color filter it should allow you to detect the amount of a certain color
A nice explanation: http://scienceblogs.com/dotphysics/2010/05/04/how-do-color-filters-work/
Disclaimer: I’ve never actually used color filters. I prefer just to use jumper clips to set different settings rather than rely on the robot’s calibration, which can sometimes act funny in different competition lighting.
Pick up a filter swatch book.
The CdS photocell is most sensitive to green, much like the human eye, it’s called a photopic response. Anyway, that aside, once you get the swatch book look for a filter with a cutoff above 550nm (ie. a red filter) and try it in front of the sensor, if it is not selective enough then try another. As the photocell is most sensitive to green light cutting out everything except red (or blue) will reduce it’s sensitivity to light considerably so, although you may be able to discriminate between red and blue, the output will be much lower.
Another option, though maybe not competition legal is the red tape used to repair tail lights.
I’ve tested this and it does give a reasonable difference between red and blue objects with enough ambient light. The readings scale with available light and the difference between red and not red was hard to read in dimly lit areas. Maybe a VEX flashlight could help in that case.
ive done some research on color filters and it seems like it could work. i havent yet been able to test it, ill post my results when i do the testing
Don’t the Line trackers use infrared light. If so I don’t think filtering out a color would work. The line trackers have a infrared filter already on them. Maybe if you removed it you could get it to filter another color
On the bright side infrared light might reflect off of red better than blue making a detectable difference.
Without having done tests yet (I plan to this summer), I think that the detectable difference won’t be large enough to rely on for competition. My bet will be that you could tell, but making a definite if greater than/less than statement will not suffice, esp. with lighting changes.
We’re talking about using the light sensor (http://www.vexrobotics.com/276-2158.html), not the line sensor. You are correct about the line sensor, though.
The light sensor even explicitly says “You can even give your robot color vision by putting colored filters on different light sensors!”
You would need a light source to make the readings consistent. The LED indicator pack Could suply some light but i think you would need to use multiple LEDs to get get good results. There also is The flash light, it uses up a motor port but would do the job niclely.
You might be able to make a red sensor with a red LED indicator and a light sensor, with a red filter.
Thanks for the suggestion. We’re definitely considering adding a color sensor to the VEX line, and it is on the future roadmap.
YES! A color sensor would be a great addition to the product line! A light sensor doesn’t do the same thing. I definitely think a color sensor would be awesome. This will be cool to experiment with and create cool devices with.
My high school team had actually worked on this sort of thing for a while to ensure we didn’t accidentally score the opponents’ buckies in autonomous (we’ve actually seen it happen a few times in matches and didn’t want to win auto for our opponents). We figured out a very good way to calibrate the sensor on the fly.
What we did was to bring two game objects with us to the queue (marked very clearly to indicate they are for calibration purposes and are not actual game objects) and as we were queuing and therefore in roughly the same light conditions as the field, we used an LCD menu to calibrate what values were red and which were blue. We basically just put a blue bucky in front of the light sensor, detected what value the light sensor got, put a red bucky in front of the light sensor and detected what value the light sensor got, and the program saved the value halfway between those two values to the flash memory as the cutoff point between red and blue.