The RGB values for the grey scale background are the key to linearizing and standardizing your model. The point is firstly, the RGBs for each point on the grey scale represent and define a balanced visual equivalence scale; and secondly that they can be mapped onto the well-known CIE standard Observer X,Y,Z visual effect model. The mapping includes an RGB to XYZ matrix cross dependency derivation and a separate Linearization function for each dimension. See my RG post:-

Computer Systems for Specifying Visualizing and Communicating Colour.

Since you say MATLAB !!! Check my RG page I have code using detectHarrisFeatures() - code to distinguish the foreground vs background OR alternates for differentiating the two ...

Thank you David and Aparna for your valuable suggestions.

First convert the gray image into RGB image using Matlab command refer weblink:

https://www.mathworks.com/matlabcentral/answers/377407-how-to-convert-grayscale-to-rgb

Then extract the R,G and b matrices as follows:

I = imread('imagename');

R = I(:,:,1)

G = I(:,:,2)

B = I(:,:,3)