Algorithm 5. Automated data augmentation and centering for RBC image dataset

Auto_Data_Augmentation(data_list, image_path)   for each folder in data_list do     define folder_path     if folder_path exists then       for each image_file in folder_path do         load image         for each scale_factor in [0.98, 0.99, 1.00, 1.01] do           resize image while embedding onto black background           save augmented image           for each rotation angle based on num_rotations do             rotate resized image             for each flip_code in [0, 1, −1] do               flip rotated image (vertical, horizontal, both)     else: Print warning (folder not found)   end for   for each folder in data_list do     define folder_aug     for each image_file in folder_aug do       load augmented image   → image       apply mean-shift filtering  → shifted       convert to grayscale and apply Otsu thresholding → thresh       detect contours → cnts       for each contour c in cnts do         extract ROI with small padding         generate binary mask and apply bitwise extraction         if extracted cell size < 128×128:           embed cell into black 128×128 background, centered           save centered image   end for generate augmentation report summarizing transformations applied   return augmented_dataset_summary