Dear Mr. Cornwall and Ms. Murthy, thank you very much for your helpful advices. Actually, I am looking for the temperature range from 273K to 373K. As Ms. Murthy sheared, all the places I have looked there are references for higher temperatures like 1000K but I have not found any references that supports my requirements (273K to 373K). 

I have noticed an interesting fact from Ms. Murthy's reference and another reference that I have linked below, it is mainly a plot of T Vs RGB. At first I want to say that it is only my opinion, I do not know the reality but below 1000K, the value of "R" is always "255" or very close to "255" that is very difficult to differentiate and when the spectrum comes close to 373K the only component left is  "R" (G=0, B=0).

Please correct me if I am mistaken.

Reference:

http://www.tannerhelland.com/4435/convert-temperature-rgb-algorithm-code/

@Aparna Murthy: Ma'am,

  'Red is cooler, blue is warmer' - really informative.

I too yet have believed in the reverse!

Yes, I agree with you. So is it a redundant idea trying to find the color below 373K?

Actually, I got interested in that spectrum range because I have seen different thermal cameras performing in that range. But a brief research shows me that most of the time they use a reference spectrum of higher temperature and map it to show the color difference between different temperatures.Their temperature sensor detects the highest temperature and map other lower temperatures accordingly. So the Emissivity is applicable only to the temperature sensor or value, but not the graphical display.

Please do correct me if I am wrong.

The displays in thermal cameras use a false color concept.  They arbitrarily map a temperature to a color.  A simplified form of this is to use a gray scale mapping so that (usually) white is the hot end of the range and black is the cold end of the range.  Since the mapping is arbitrary it can also be white is cold and black is hot.  The latter is sometimes preferred for visual presentation of some scenes.  

There is no RGB equals T or equals wavelength direct, physical, immutable correspondence.  If you look at the various displays used by IR cameras to present temperature maps, you will find various palettes used with names like Rainbow, Ironbow, Medical, etc.  Each maps a different portion of the detected temperature to a different color.  Some, like Ironbow, are fairly continuous and rely on the colors displayed by metal as it is heated for a guide.  However, this is only a convenience.  Others are selected to maximize the visibility of T differences or of a particular portion of the T range. 

You already have temperature.  You indicate that the range is 100K.  You can decide how many colors to use for your display and arbitrarily assign them to a T range.  

Dear Mr. Kleinfeld,

Thank you very much for your valuable information. 

It's just a vizualisation. The same thing is color of relief in geographical maps.

For the monochromic vizualisation you can nomalize T to (0;255) and use RGB(T;T;T).

The color vizualisation (T) -> (R, G, B) is the mapping from 1d to 3d space. The single right way just doesn't exist - you can use a lot of algorithms.

Good luck!   

You're welcome!