Write a function to add "randn" Gaussian noise to an specific percentage of images inside imageData store in matlab

Answer 1

[Nitin Kapgate]

Hi Mohamed,

I can understand that you want to add noise to a percentage of images in an ImageDataStore which are chosen at random.

Here is a demo code which will help you to solve your problem.

% Load images from a folder as ImageDataStore objects
  
imds = imageDatastore('C:\Users\XYZ\MyProject'); % Assuming your images are in “MyProject” folder
  
percentImages = 0.2; % Fraction of Images you want to add Noise
  
numTotalImg = numel(imds.Files) % Total number of Images in the Folder
  
numNoisyImg = round(percentImages * numTotalImg) % Number of images you want to add the Noise
  
selectedNoisyImages = randperm(numTotalImg,numNoisyImg); % Generate random indices of images to which noise is to be added
  
for k = 1:numNoisyImg
  
    fileName = imds.Files{selectedNoisyImages(k)} % read the filename of the image with random index, to add noise
  
    I = imread(fileName); % Read the image using imread function  
  
    % Insert Code to add noise to the image I here
    
    % You may use "randn" or "imnoise" function to add the noise of your choice   
  
    % If needed, write back noisy image to a different folder.
    
end