Skip to main content
. 2021 Jan 29;210:104256. doi: 10.1016/j.chemolab.2021.104256

Fig. 6.

Fig. 6

The fuzzy tree-based image generation.
  • Step 3: Resize the generated 15 images to 256 ​× ​256 sized images.
  • Step 4: Extract 768 features from each image by using MKLBP.
  • Step 5: Divide images into 128 ​× ​128 sized exemplars.
  • Step 6: Extract 768 features from each exemplar.
  • Step 7: Concatenate the features and obtain a final feature vector with a size of 768 ​× ​75 ​= ​57,600.
  • Step 8: Apply INCA to extracted 57,600 features and select 616 most valuable features.