i was using color sensor 34725 with arduino to detect colors

but it seems it is working a bit strange

i have uploaded the picture of the data i gather from my experiment

here are the confusing things

1.why green value is also high when i place sensor on blue paper?

2.why green and red color are significantly more in yellow as compared when they are place on their own color.

and so on, i mean the color sensor does not seem to work properly

 

here is the pic

 

Virtual reality (computers) use combinations of red, green & blue to make other colors.

In actual reality, red, yellow & blue are primary colors. Yellow and blue make green.
Maybe you just need to think about it differently.

https://graf1x.com/product/color-mixing-guide-poster-chart-tool-pdf/

 

This surprises me that in RGB colors, red and green make yellow.

 

This surprises me that in RGB colors, red and green make yellow.

RGB is additive.

http://www.differencebetween.net/science/difference-between-additive-colors-and-subtractive-colors/

 

i was using color sensor 34725 with arduino to detect colors

but it seems it is working a bit strange

i have uploaded the picture of the data i gather from my experiment

here are the confusing things

1.why green value is also high when i place sensor on blue paper?

2.why green and red color are significantly more in yellow as compared when they are place on their own color.

and so on, i mean the color sensor does not seem to work properly

You have to remember how that sensor is working. You are basically using ADC values from different red, green, blue, and clear light photo-diode sensors. I haven't looked real carefully, but what it looks like you are getting is not unusual at all.

One way of using the information to identify colors is to use the values to calculate x and y chromaticity coordinates. I took a stab at doing this with the TCS34725 and I was actually pretty amazed at the results that I got - see here if you think it might help.

I would add that if I were going to do it again, to build a color sorter for example, I would choose a more advanced chip that actually gives you the x,y,z coordinates - something like this one or this one.

 

the color sensor does not seem to work properly

If you are using the sensor to check the colour of a printed item, a lot will depend on the reflectance of the inks used and also on the illuminating light source. It is doubtful that a supposedly red printed item would reflect pure red, for example.

 

Maybe look at the ratios instead of the absolute amounts. That may tell you more.