how to correct this error?

for x=1:rows
    for y=1:cols
      wt(x,y)=t1(x,y)/max(t1(x,y));
    end
end
c=norm(double(wt.*(eimg1(x,y)-eimg(x,y))))/norm(double(wt.*eimg(x,y)));

For the last line,

Error using  .* 
Integers can only be combined with integers of the same class, or scalar doubles.

I want this norm value to use as a condition in a loop eimg eimg1 are matrices of an image. wt is the weight function. t1 is the suppression term at the first iteration.

Answer 1

[Walter Roberson]

You have

      wt(x,y)=t1(x,y)/max(t1(x,y));

Your x and y are scalars, so your t1(x,y) is going to be a scalar, and max() of a scalar is the scalar itself. Your wt() calculation is always going to result in 1 for every location. If t1 is floating point then your wt is going to be floating point; if your t1 is integer, your wt would be integer of the same class.

My guess is that you should replace your double loop with the single line

wt = double(t1) ./ double( max(t1) );

This would give you an array of floating point.

If your eimg1 and eimg are your images, they are probably integer class. In that case you would need to replace

c=norm(double(wt.*(eimg1(x,y)-eimg(x,y))))/norm(double(wt.*eimg(x,y)));

with

c = norm(wt .* (double(eimg1) - double(eimg))) / norm(wt .* double(eimg));

Notice here that I did not subscript with x and y: you are outside the loops at this point so x and y are nearly meaningless. Notice also that I used double() before subtracting: when you subtract two items of integer class, the result might not be what you expect.

[Divya Praisy]

Error using .* Matrix dimensions must agree.

Error in mainw>pushbutton3_Callback (line 196)

c = norm(wt .* (double(eimg1) - double(eimg))) / norm(wt .* double(eimg));

[Image Analyst]

You need ./ not just /

c = norm(wt .* (double(eimg1) - double(eimg))) ./ norm(wt .* double(eimg));

[Walter Roberson]

What is size(eimg) and size(wt) ? What size of output are you expecting for c ?

[Divya Praisy]

eimg and wt are the size the image loaded

[Walter Roberson]

The norm() calls will return scalars so ./ should not be needed. And the error message was about .* not about /

Please show the exact values of

rows
cols
size(t1)
size(wt)
size(eimg1)
size(eimg)

And also please say what size you are expecting "c" to be when it is calculated.

[Divya Praisy]

the sizes vary with the size of any image inputted the condition for the loop is c > 7 * 10 ^ -6

[Walter Roberson]

You posted above that you received an error with .* in executing my code. For that particular combination of inputs, I need you to show the values I requested above. If you do not show the exact values associated with an occurrence of the error message, I will not be able to assist you further.