generate_reconstructed_data.py 7.1 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190
  1. #!/usr/bin/env python3
  2. # -*- coding: utf-8 -*-
  3. """
  4. Created on Wed Jun 19 11:47:42 2019
  5. @author: jbuisine
  6. """
  7. # main imports
  8. import sys, os, argparse
  9. import numpy as np
  10. # images processing imports
  11. from PIL import Image
  12. from ipfml.processing.segmentation import divide_in_blocks
  13. # modules imports
  14. sys.path.insert(0, '') # trick to enable import of main folder module
  15. import custom_config as cfg
  16. from modules.utils.data import get_scene_image_quality
  17. from modules.classes.Transformation import Transformation
  18. # getting configuration information
  19. zone_folder = cfg.zone_folder
  20. min_max_filename = cfg.min_max_filename_extension
  21. # define all scenes values
  22. scenes_list = cfg.scenes_names
  23. scenes_indexes = cfg.scenes_indices
  24. path = cfg.dataset_path
  25. zones = cfg.zones_indices
  26. seuil_expe_filename = cfg.seuil_expe_filename
  27. features_choices = cfg.features_choices_labels
  28. output_data_folder = cfg.output_data_folder
  29. generic_output_file_svd = '_random.csv'
  30. def generate_data(transformation):
  31. """
  32. @brief Method which generates all .csv files from scenes
  33. @return nothing
  34. """
  35. scenes = os.listdir(path)
  36. # remove min max file from scenes folder
  37. scenes = [s for s in scenes if min_max_filename not in s]
  38. # go ahead each scenes
  39. for id_scene, folder_scene in enumerate(scenes):
  40. print(folder_scene)
  41. scene_path = os.path.join(path, folder_scene)
  42. # construct each zones folder name
  43. zones_folder = []
  44. features_folder = []
  45. zones_threshold = []
  46. # get zones list info
  47. for index in zones:
  48. index_str = str(index)
  49. if len(index_str) < 2:
  50. index_str = "0" + index_str
  51. current_zone = "zone"+index_str
  52. zones_folder.append(current_zone)
  53. zone_path = os.path.join(scene_path, current_zone)
  54. with open(os.path.join(zone_path, cfg.seuil_expe_filename)) as f:
  55. zones_threshold.append(int(f.readline()))
  56. # custom path for feature
  57. feature_path = os.path.join(zone_path, transformation.getName())
  58. if not os.path.exists(feature_path):
  59. os.makedirs(feature_path)
  60. # custom path for interval of reconstruction and feature
  61. feature_interval_path = os.path.join(zone_path, transformation.getTransformationPath())
  62. features_folder.append(feature_interval_path)
  63. if not os.path.exists(feature_interval_path):
  64. os.makedirs(feature_interval_path)
  65. # create for each zone the labels folder
  66. labels = [cfg.not_noisy_folder, cfg.noisy_folder]
  67. for label in labels:
  68. label_folder = os.path.join(feature_interval_path, label)
  69. if not os.path.exists(label_folder):
  70. os.makedirs(label_folder)
  71. # get all images of folder
  72. scene_images = sorted([os.path.join(scene_path, img) for img in os.listdir(scene_path) if cfg.scene_image_extension in img])
  73. number_scene_image = len(scene_images)
  74. # for each images
  75. for id_img, img_path in enumerate(scene_images):
  76. current_img = Image.open(img_path)
  77. img_blocks = divide_in_blocks(current_img, cfg.keras_img_size)
  78. current_quality_index = int(get_scene_image_quality(img_path))
  79. for id_block, block in enumerate(img_blocks):
  80. ##########################
  81. # Image computation part #
  82. ##########################
  83. # pass block to grey level
  84. output_block = transformation.getTransformedImage(block)
  85. output_block = np.array(output_block, 'uint8')
  86. # current output image
  87. output_block_img = Image.fromarray(output_block)
  88. label_path = features_folder[id_block]
  89. # get label folder for block
  90. if current_quality_index > zones_threshold[id_block]:
  91. label_path = os.path.join(label_path, cfg.not_noisy_folder)
  92. else:
  93. label_path = os.path.join(label_path, cfg.noisy_folder)
  94. # Data augmentation!
  95. rotations = [0, 90, 180, 270]
  96. img_flip_labels = ['original', 'horizontal', 'vertical', 'both']
  97. horizontal_img = output_block_img.transpose(Image.FLIP_LEFT_RIGHT)
  98. vertical_img = output_block_img.transpose(Image.FLIP_TOP_BOTTOM)
  99. both_img = output_block_img.transpose(Image.TRANSPOSE)
  100. flip_images = [output_block_img, horizontal_img, vertical_img, both_img]
  101. # rotate and flip image to increase dataset size
  102. for id, flip in enumerate(flip_images):
  103. for rotation in rotations:
  104. rotated_output_img = flip.rotate(rotation)
  105. output_reconstructed_filename = img_path.split('/')[-1].replace('.png', '') + '_' + zones_folder[id_block] + cfg.post_image_name_separator
  106. output_reconstructed_filename = output_reconstructed_filename + img_flip_labels[id] + '_' + str(rotation) + '.png'
  107. output_reconstructed_path = os.path.join(label_path, output_reconstructed_filename)
  108. rotated_output_img.save(output_reconstructed_path)
  109. print(transformation.getName() + "_" + folder_scene + " - " + "{0:.2f}".format(((id_img + 1) / number_scene_image)* 100.) + "%")
  110. sys.stdout.write("\033[F")
  111. print('\n')
  112. print("%s_%s : end of data generation\n" % (transformation.getName(), transformation.getParam()))
  113. def main():
  114. parser = argparse.ArgumentParser(description="Compute and prepare data of feature of all scenes using specific interval if necessary")
  115. parser.add_argument('--features', type=str,
  116. help="list of features choice in order to compute data",
  117. default='svd_reconstruction, ipca_reconstruction',
  118. required=True)
  119. parser.add_argument('--params', type=str,
  120. help="list of specific param for each feature choice (See README.md for further information in 3D mode)",
  121. default='100, 200 :: 50, 25',
  122. required=True)
  123. args = parser.parse_args()
  124. p_features = list(map(str.strip, args.features.split(',')))
  125. p_params = list(map(str.strip, args.params.split('::')))
  126. transformations = []
  127. for id, feature in enumerate(p_features):
  128. if feature not in features_choices or feature == 'static':
  129. raise ValueError("Unknown feature, please select a correct feature (`static` excluded) : ", features_choices)
  130. transformations.append(Transformation(feature, p_params[id]))
  131. # generate all or specific feature data
  132. for transformation in transformations:
  133. generate_data(transformation)
  134. if __name__== "__main__":
  135. main()