Module 6: Liquid Crystal Thermography
  Lecture 36: HSI model
 

Contd...

The HSI -coordinates shown in the RGB color cube (Figure 6.2) establish the required relationship between the two set of coordinates.

Figure 6.2: HSI coordinates shown in RGB color cube.

The hue component in both color spaces is an angular measurement, analogous to angular position on a color wheel. However, such color spaces are not unique, and a number of definitions of hue can be found in the literature. The following formula is one of them and can be used to convert RGB values to HSI values (Smith, 2001). Experience shows that it yields the lowest average uncertainty while retaining computational simplicity.

 

Software packages such as MATLAB use an alternative definition



where

Saturation is defined as

It is a fraction between 0 and unity. Intensity is defined as

In camera measurements, the individual values of R , G and B would vary over 0 to 255 (8-bit resolution).

Often, intensity is expressed as a fraction between 0 and unity by dividing the value derived as above by 255. Similarly, hue can be normalized by dividing the value derived above by 360 (degrees).

In experiments involving temperature measurements, the primary quantity of interest is hue (H). Other quantities (S and I) need not be computed since they show small variation with temperature. Hence, the major advantage of using hue as a sensitizer of temperature is the one-to-one relationship obtained (as opposed to 3-to-1 in the RGB data).