Implement Algorithm 16.2 as a Matlab function:

function s = CV_E_step(params,support,imf)
%
%CV_E_step - compute expected values of classes
%On input:
%     params (gx1 array): parameters of g classes defined as alpha,mu,sigma fields
%     support (mxnxg array): In band g, likelihood that pixel belongs to class g
%     imf (mxnxp array): Feature image
%On output:
%     s (gx1 array): updated support array
%Call:
%     s_next =  CV_E_step(params,s_prev);
%Author:
%     Your name
%     UU
%     Fall 2012
%