How to replace all nan values in an image to 255 different colors

How to replace all nan values in an image to 255 different colors distributed almost equally

Answer 1

[David Hill]

idx=yourImage==nan;
N=nnz(idx);
yourImage(idx)=uint8(randi(255,1,N));

[ammu v]

thanku for response, but using randi will make it change everytime i suppose.is there a way to fix that

Answer 2

[DGM]

Depends what your needs are. If I assume that your image is an RGB color image to begin with, and that you want colors picked sequentially from a given color table:

% specify a color table
ncolors = 256;
ct = parula(ncolors);

% get an image
A = im2double(imread('peppers.png'));
A = imresize(A,0.5);
s = size(A);

% put some NaNs in the image
idx = randi([1 numel(A)],1024,1);
A(idx) = NaN;

% reshape the image
A = reshape(A,[],3);
% get a map of the NaNs
nanmap = any(isnan(A),2);

% generate a list of replacement pixels
fillcolors = ct(mod(0:nnz(nanmap)-1,ncolors)+1,:);
% fill and reshape
A(nanmap,:) = fillcolors;
A = reshape(A,s(1),s(2),3);

imshow(A)

Note that there are 1024 NaNs added to the image, but the color table has only 256 entries. Note that the color table visibly repeats 4 times across the image.