How to store images in a single array or matrix

Hello, i have a number of jpeg images stored in a folder; how can i read all the images and store them as a single matrix or array (reserving the image format).

Answer 1

[Image Analyst]

Straight out of the FAQ:

myFolder = 'C:\Documents and Settings\yourUserName\My Documents\My Pictures';
if ~isdir(myFolder)
  errorMessage = sprintf('Error: The following folder does not exist:\n%s', myFolder);
  uiwait(warndlg(errorMessage));
  return;
end
filePattern = fullfile(myFolder, '*.jpg');
jpegFiles = dir(filePattern);
for k = 1:length(jpegFiles)
  baseFileName = jpegFiles(k).name;
  fullFileName = fullfile(myFolder, baseFileName);
  fprintf(1, 'Now reading %s\n', fullFileName);
  imageArray = imread(fullFileName);
  imshow(imageArray);  % Display image.
  drawnow; % Force display to update immediately.
end

Obviously, you can adapt as needed, because I'm not sure what you mean when you say "store them as a single matrix". There is a montage() function that will do that if that's what you want. It will stitch together all the image files into one giant image.

[Bashir]

Thank you very much, this is really helpful. what i mean is that i want to store the images in a single matrix(as jpeg images not uint8) so that i can read them later for further processing. if this is not possible, how do i extract the uint8 images stored in a cell array to be processed(as normal jpeg images) for feature extraction. thank you once more.

[David Sanchez]

Do you mean to store all your images in a single matrix or each matrix in each own matrix?

[Image Analyst]

Bashir - that doesn't make sense. Once imported into MATLAB, your image will be uint8 - and whatever format it had while stored on disk doesn't matter anymore. Likewise, in MATLAB, your image is simply a uint8 array, and it doesn't matter how it's stored. No matter what format it's stored in, you can recall it back into MATLAB as a uint8 image. Though I recommend PNG rather than JPEG so that you don't lose any information due to JPEG's lossy compression process.

To extract anything from a cell array's cell, use braces. You will find the FAQ on cell arrays very informative.

[Bashir]

it is now clear to me, thank you for this useful information. hope to get your response in my subsequent questions.

[shikha mangal]

How can i save all the images stored in a cell array in the matlab directory path,so that i can use them further.What is the syntax of writing multiple images from cell array to the system in jpg format?please help.

for i = 1:numel(myCellArr)
    fileName = sprintf('image_%04d.jpg');
    imwrite( myCellArr{i}, 'fileName.jpg' );
end

I have used this code but it is not saving all of my images stored in cell but only the last image.'please help me to obtain the desired result.

[Image Analyst]

They're all being saved in the file called 'fileName.jpg' because you hard coded in the string instead of passing in the variable. You need to do this:

for i = 1:numel(myCellArr)
    fileName = sprintf('image_%04d.jpg');
    imwrite(myCellArr{i}, fileName);
end

[shikha mangal]

Thanks for correcting me image analyst but now it is giving me the error as 'unable to determine the file format from fileName'. how can i remove this error?And my cell array is 'g'.

[Image Analyst]

You need to put i into the sprintf, and replace myCellArr by (the poorly-named) "g":

for i = 1:numel(g)
    fileName = sprintf('image_%04d.jpg', i);
    imwrite(g{i}, fileName);
end

[shikha mangal]

Yes its working. Thsnk you very much sir.

[Vikas Dhiman]

Hi. I have used the same code for reading images..Saved it in imageArray The dimensions of imageArray is 243x320 but my images are 165 only. How can i use this array for clustering. I mean shouldn't be that one image represents one row of an array ? please provide some insight Thanks

[Image Analyst]

What are the features that you want to identify clusters for? In other words, what are you measuring in the images? If you plot a scatterplot, what features go along the axes?

[Parth Sane]

Excellent answer! Thank you!

[moahaimen talib]

@Image Analyst i have the same question i need to input 20 images and extract features for all of them such as entropy energy and then store these features to `use it train a classifier but when i store them i have bad results in the soms nn

[Image Analyst]

If you have the same question, see my answer to the same question. And I don't know much if anything about self organizing map neural networks.

[Ganesh Gajavelli]

One thing to look out for with montage() is that it requires all of the pictures to be of the same dimensions. But thanks anyways!

[Ganesh Gajavelli]

Another thing that could be useful is imageSet() Maybe this could work.

Answer 2

[Azzi Abdelmalek]

f=dir('*.jpg')
files={f.name}
for k=1:numel(files)
  Im{k}=imread(files{k})
end

[Bashir]

I very much appreciate your fast answer,but i am having an error msg: Error in ==> multireadIm at 5

    Im{k} = imread(files{k}); pls what is wrong with this line code.

Thank you

[Azzi Abdelmalek]

Check if your files are in the current folder, if not change the code to:

f=dir('folder\*.jpg')  %'folder' is the folder containing your jpg files
files={f.name}
for k=1:numel(files)
Im{k}=imread(files{k})
end

[Image Analyst]

You still need to construct the full filename (as shown in the FAQ) since files{k} is only the base file name and you'll have the same problem.

folder = 'D:\My Pictures\whatever'
filePattern = fullfile(folder, '*.jpg');
f=dir(filePattern)
files={f.name}
for k=1:numel(files)
  fullFileName = fullfile(folder, files{k})
  cellArrayOfImages{k}=imread(fullFileName)
end

[Karen Bugeja]

how can I output the images stored in cellArrayOfImages? I am using imshow(cellArrayOfImages); and I'm getting an error. thanks!

[Image Analyst]

For example, to display the 23rd image, do

imshow(cellArrayOfImages{23});

the braces mean "Contents of". Please see the FAQ: http://matlab.wikia.com/wiki/FAQ#What_is_a_cell_array.3F

[preeti]

thanks its so helpful :):)

[Venkatesh chari]

folder = 'D:\My Pictures\whatever'
filePattern = fullfile(folder, '*.jpg');
f=dir(filePattern)
files={f.name}
for k=1:numel(files)
  fullFileName = fullfile(folder, files{k})
  cellArrayOfImages{k}=imread(fullFileName)
end

After this, I need to process a specific image. Let's say for e.g. image 23. How can I crop it, convert it to grayscale & double, add median filter to it, and subtract from some other image?

Can you give me syntax for these operations?

[Image Analyst]

Try this:

thisImage = cellArrayOfImages{23};
thisImage = double(rgb2gray(thisImage));
medFiltered = medfilt2(thisImage, [3,3]);
diffImage = someOtherImage - medFiltered;

[Anjali S]

I am facing the same problem. I tried above code so my all images are now stored in matrix having size 1*6. Every cell is storing size of image (1*48000 due to resize)But I want every row should store actual bits of image and not the size. That is I want to store each image as a one row for creating training dataset. Ex my image is of 200*240 so I resized to 1*48000 . First row of the matrix should have 48000 columns like that matrix size will be 6 * 48000. I REQUEST KINDLY HELP ME ..

[habtamu miheretie]

hey man, i have exactly the same problem. I have many images and want to make a matrix out of it. Have you solved your problem? if so can you share what you did? thanks in advance!

[Jan]

[MOVED from flags] UKM wrote:

Very simple and Very perfect>> Thanks alot.

@UKM: Please use flags only to inform admins and editors about inappropriate content like sopam or rudeness.

[Abhishek Singh]

@Image Analyst After I have the array for the images can I compare those images without using loop? I need to compare two images using ssim() which was suggested by you only but I think loop is very slow in Matlab I was looking for something in arrays and vectors which could help me.

Answer 3

[Shelly Smith]

Hey

I have a problem and will be grateful if someone help me. I am reading 200 images through this code:

folder = 'C:\Users\whatever'
filePattern = fullfile(folder, '*.pgm');
f=dir(filePattern)
files={f.name}
for k=1:numel(files)
  fullFileName = fullfile(folder, files{k})
  cellArrayOfImages{k}=imread(fullFileName)
      imshow(cellArrayOfImages{k});
end

I get a row vector with 200 length. In each cell there's a 50x50 uint8 image. I need to have the elements value of 50x50 image as well. So I want to basically read all the images (200 images) and stored the value of all images in a matrix. Can anyone tell me how can I do that? Thanks in advance.

[Image Analyst]

I don't understand the question. You are reading in 200 images, which are 50 x 50 uint8 images, and storing them. It seems to be doing what you are asking for, so what's the problem? Do you want a 2-D array of cells instead of a 1-D row vector of cells? If so, why? And how many rows and columns do you want?

Answer 4

[Lav Palve]

Store images in cell array

clc;
clear all;
close all;
arr = cell(25,77); % Create Cell array 
k = 1;
% get 'zero' 25 image cell array 
img_folder = ('E:\4th year\Project\Images\Chardatabase\Sample0'); % Enter name folder and its path
filenames = dir(fullfile(img_folder,'*.jpg')); % Read all image with specified extantion
Total_image = numel(filenames); % Count total image
for i=1:Total_image
    j = 1;
    for j=1:77
    f = fullfile(img_folder,filenames(i).name); % Stroe ith image path
    Output = imread(f); % read image
    Output = imresize(Output,[11 7]);
    Output = im2bw(Output);
    Output = reshape(Output,[],77); % cell array divide by '77'
    Output = im2double(Output);
    arr{k,j} = Output(1,j); % get all pixel value of 'Output' image
    end
    k = k+1;
end

Answer 5

[Vojtech Raska]

Hi, i´m using this code to make 3-D matrix of 2-D MRI images, but if I load more than 3 images that every image besides firts three and the last one is just black and all values of that matrices are zeros. The images are uint8 format. I tried to trasfer them to double but it didn´t work. Is there some solution or the image matrices can be only three for each other?

clc; clear all; close all;

[Name, Path] = uigetfile('*.*',  'All Files (*.*)','MultiSelect','on');
if ~iscell(Name)  
    Name = {Name};
end

%% matrix
for i = 1:1:length(Name)
    Nazev = Name(1,i);
    Nazev = char(Nazev);
    Img_info = [Path Nazev];
    Img = imread(Img_info);
    Img(:,:,i) = rgb2gray(Img);
end

Answer 6

[Shreyas S]

im = {'image1.png','image2.png','image3.png'};
for k = 1:length(im)
    image1 = im2double(imread(im{k}));
    % ......
end