How to improve Deep Learning results with image processing - Isaamarod/Matlab_processing_image GitHub Wiki

STEPS AFTER DOWNLOAD THE CODE

Input images atbs:

Execute in matlab prompt:

  • h = imshow('blur_20_RBG_-100_test_con-018.jpg');
  • info = imfinfo('blur_20_RBG_-100_test_con-018.jpg');
  • imageinfo(h,info)

It's important because in some functions in matlab you'll have to convert your image class. For example in this case:

  1. input image class: uint8 256x256x3 --> color image
  2. Before colormap processing we have to convert to 256x256 --> convert to gray --> color2gray.m dependencie

image_info_github

Code general dependencies in your path to add

Download export_fig.m : export_fig.m

Dependencies for Colormap part

Convert color2gray.m: color2gray.m

colomap_blur_20_rbg_-100_test_con-018

Dependencies for Fuzzy c-means part

Repository: Fuzzy c-means

Install License Image Processing Toolbox Add this function : initfcm.m

Important code observations: m_color.m --> Number of clusters are relevant. For example:

  • cluster=9 fuzzy_cluster9_colomap_blur_20_rbg_-100_test_con-018
  • cluster=8 fuzzy_cluster8_colomap_blur_20_rbg_-100_test_con-018
  • cluster=7 fuzzy_cluster7_colomap_blur_20_rbg_-100_test_con-018
  • cluster=5 fuzzy_cluster5_colomap_blur_20_rbg_-100_test_con-018
  • cluster=2 fuzzy_cluster2_colomap_blur_20_rbg_-100_test_con-018