Image resizing using augmentedImageDatastore

i tried to resize the image stored in the following path. C:\Users\manjurama\Desktop\study\PG\PROJECT\finger vein database\database_10

imagefolder = 'C:\Users\manjurama\Desktop\study\PG\PROJECT\finger vein database\database_10';

imds = imageDatastore(imagefolder,'IncludeSubfolders',true,'LabelSource','foldernames');

veinimage = augmentedImageDatastore([227 227],imds)

numTrainFiles = 0.75;

[TrainImages, TestImages] = splitEachLabel(veinimage, numTrainFiles,'randomize'); % this is the line 14 error shown here.

But i got this error

Check for missing argument or incorrect argument data type in call to function 'splitEachLabel'.

Error in Classification (line 14)

[TrainImages, TestImages] = splitEachLabel(veinimage, numTrainFiles,'randomize');

Could you please let me know about the problem in my code. Thanks.

Answer 1

[Ibrahim A]

Hi Rd,

I had same problem and fixed by splitting the data before I do the augmentedImageDatastore then have two augmentedImageDatastores one for the training portion and one for testing. So the code will be like.

imagefolder = 'C:\Users\manjurama\Desktop\study\PG\PROJECT\finger vein database\database_10';
imds = imageDatastore(imagefolder,'IncludeSubfolders',true,'LabelSource','foldernames');
numTrainFiles = 0.75;
imageSize = [227 227];

[TrainImages, TestImages] = splitEachLabel(imds, numTrainFiles,'randomize');
TrainImages = augmentedImageDatastore(imageSize,TrainImages)
TestImages = augmentedImageDatastore(imageSize,TestImages)

[Rd]

Your answer is really helpful. Thanks a lot sir.

Answer 2

[Mohammad Sami]

The last argument to split each label is not spelled correctly. It's

splitEachLabel(___,'randomized').

[Rd]

I changed it as randomized. Still the same error is displaying.