This CNN model could be used to process large (infinity image number) dataset Split images file read original images and combined with labels. The labels are starting from group 0. Images are rotated by 90, 180, 270 degrees to amplify dataset. Data were splitted into training and testing datasets with 0.8 and 0.2 percentage.
- Training data and testing data, training label and testing label were uploaded to google drive.
- Link: https://drive.google.com/open?id=1V9rIgpgWcP1NnmiM10p6w8WfQGGMk2om
- Example image.