How to compare a matrix of values to all values within an array without loops

I'm looking to compare a gray-scale image to an array of values and if a value does not exist within the image, increment a seperate matrix as a counter. I'd like to do this without for loops because it can take several of seconds per frame.

As an example, if i have the following fake image:

A = [1 2 4; 5 6 9;12 13 15];

B then is an array defined by min(A(:)):max(A(:)) to cover all possible values.

The following code does what i need and the output is the following:

grayLevelZeros = setdiff(B,unique(A))-1;
grayLevelZerosFix = zeros(size(A));
for i = 1:length(grayLevelZeros)
    grayLevelZerosFix(A>grayLevelZeros(i)) = i;
end

grayLevelZerosFix =

0 0 1

1 1 3

5 5 6

How could I do this without using a for loop to speed up the processing?

[Stephen23]

"How could I do this without using a for loop to speed up the processing?"

Did you PROFILE the code? SETDIFF is likely the slowest part of that code, not the loop.

Answer 1

[Matt J]

A = [1 2 4; 5 6 9;12 13 15];

grayLevelZeros = setdiff(min(A(:)):max(A(:)),unique(A))-1;

grayLevelZerosFix  = discretize(A,[-inf,grayLevelZeros,inf],'inc','right')-1

grayLevelZerosFix = 3×3

0 0 1 1 1 3 5 5 6

[Scott Greenwald]

Thank you so much!