display_svd_data_scene.py 11 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325
  1. #!/usr/bin/env python3
  2. # -*- coding: utf-8 -*-
  3. """
  4. Created on Fri Sep 14 21:02:42 2018
  5. @author: jbuisine
  6. """
  7. from __future__ import print_function
  8. import sys, os, getopt
  9. import numpy as np
  10. import random
  11. import time
  12. import json
  13. from PIL import Image
  14. from ipfml import processing, metrics, utils
  15. import ipfml.iqa.fr as fr_iqa
  16. from skimage import color
  17. import matplotlib.pyplot as plt
  18. from modules.utils.data import get_svd_data
  19. from modules.utils import config as cfg
  20. # getting configuration information
  21. config_filename = cfg.config_filename
  22. zone_folder = cfg.zone_folder
  23. min_max_filename = cfg.min_max_filename_extension
  24. # define all scenes values
  25. scenes_list = cfg.scenes_names
  26. scenes_indices = cfg.scenes_indices
  27. choices = cfg.normalization_choices
  28. path = cfg.dataset_path
  29. zones = cfg.zones_indices
  30. seuil_expe_filename = cfg.seuil_expe_filename
  31. metric_choices = cfg.metric_choices_labels
  32. max_nb_bits = 8
  33. error_data_choices = ['mae', 'mse', 'ssim', 'psnr']
  34. def get_error_distance(p_error, y_true, y_test):
  35. noise_method = None
  36. function_name = p_error
  37. try:
  38. error_method = getattr(fr_iqa, function_name)
  39. except AttributeError:
  40. raise NotImplementedError("Error `{}` not implement `{}`".format(fr_iqa.__name__, function_name))
  41. return error_method(y_true, y_test)
  42. def display_svd_values(p_scene, p_interval, p_indices, p_metric, p_mode, p_step, p_norm, p_error, p_ylim):
  43. """
  44. @brief Method which gives information about svd curves from zone of picture
  45. @param p_scene, scene expected to show svd values
  46. @param p_interval, interval [begin, end] of svd data to display
  47. @param p_interval, interval [begin, end] of samples or minutes from render generation engine
  48. @param p_metric, metric computed to show
  49. @param p_mode, normalization's mode
  50. @param p_norm, normalization or not of selected svd data
  51. @param p_error, error metric used to display
  52. @param p_ylim, ylim choice to better display of data
  53. @return nothing
  54. """
  55. max_value_svd = 0
  56. min_value_svd = sys.maxsize
  57. image_indices = []
  58. scenes = os.listdir(path)
  59. # remove min max file from scenes folder
  60. scenes = [s for s in scenes if min_max_filename not in s]
  61. begin_data, end_data = p_interval
  62. begin_index, end_index = p_indices
  63. data_min_max_filename = os.path.join(path, p_metric + min_max_filename)
  64. # go ahead each scenes
  65. for id_scene, folder_scene in enumerate(scenes):
  66. if p_scene == folder_scene:
  67. scene_path = os.path.join(path, folder_scene)
  68. config_file_path = os.path.join(scene_path, config_filename)
  69. with open(config_file_path, "r") as config_file:
  70. last_image_name = config_file.readline().strip()
  71. prefix_image_name = config_file.readline().strip()
  72. start_index_image = config_file.readline().strip()
  73. end_index_image = config_file.readline().strip()
  74. step_counter = int(config_file.readline().strip())
  75. # construct each zones folder name
  76. zones_folder = []
  77. # get zones list info
  78. for index in zones:
  79. index_str = str(index)
  80. if len(index_str) < 2:
  81. index_str = "0" + index_str
  82. current_zone = "zone"+index_str
  83. zones_folder.append(current_zone)
  84. images_data = []
  85. images_indices = []
  86. threshold_learned_zones = []
  87. for id, zone_folder in enumerate(zones_folder):
  88. # get threshold information
  89. zone_path = os.path.join(scene_path, zone_folder)
  90. path_seuil = os.path.join(zone_path, seuil_expe_filename)
  91. # open treshold path and get this information
  92. with open(path_seuil, "r") as seuil_file:
  93. threshold_learned = int(seuil_file.readline().strip())
  94. threshold_learned_zones.append(threshold_learned)
  95. current_counter_index = int(start_index_image)
  96. end_counter_index = int(end_index_image)
  97. threshold_mean = np.mean(np.asarray(threshold_learned_zones))
  98. threshold_image_found = False
  99. file_path = os.path.join(scene_path, prefix_image_name + "{}.png")
  100. svd_data = []
  101. while(current_counter_index <= end_counter_index):
  102. current_counter_index_str = str(current_counter_index)
  103. while len(start_index_image) > len(current_counter_index_str):
  104. current_counter_index_str = "0" + current_counter_index_str
  105. if current_counter_index % p_step == 0:
  106. if current_counter_index >= begin_index and current_counter_index <= end_index:
  107. images_indices.append(current_counter_index_str)
  108. if threshold_mean < int(current_counter_index) and not threshold_image_found:
  109. threshold_image_found = True
  110. threshold_image_zone = current_counter_index_str
  111. image_path = file_path.format(str(current_counter_index_str))
  112. img = Image.open(image_path)
  113. svd_values = get_svd_data(p_metric, img)
  114. if p_norm:
  115. svd_values = svd_values[begin:end]
  116. # update min max values
  117. min_value = svd_values.min()
  118. max_value = svd_values.max()
  119. if min_value < min_value_svd:
  120. min_value_svd = min_value
  121. if max_value > min_value_svd:
  122. max_value_svd = max_value
  123. # keep in memory used data
  124. if current_counter_index % p_step == 0:
  125. if current_counter_index >= begin_index and current_counter_index <= end_index:
  126. images_indices.append(current_counter_index_str)
  127. svd_data.append(svd_values)
  128. if threshold_mean < int(current_counter_index) and not threshold_image_found:
  129. threshold_image_found = True
  130. threshold_image_zone = current_counter_index_str
  131. current_counter_index += step_counter
  132. print('%.2f%%' % (current_counter_index / end_counter_index * 100))
  133. sys.stdout.write("\033[F")
  134. # all indices of picture to plot
  135. print(images_indices)
  136. previous_data = []
  137. error_data = [0.]
  138. for id, data in enumerate(svd_data):
  139. current_data = data
  140. if p_mode == 'svdn':
  141. current_data = utils.normalize_arr(current_data)
  142. if p_mode == 'svdne':
  143. current_data = utils.normalize_arr_with_range(current_data, min_value_svd, max_value_svd)
  144. images_data.append(current_data)
  145. # use of whole image data for computation of ssim or psnr
  146. if p_error == 'ssim' or p_error == 'psnr':
  147. image_path = file_path.format(str(current_id))
  148. current_data = np.asarray(Image.open(image_path))
  149. if len(previous_data) > 0:
  150. current_error = get_error_distance(p_error, previous_data, current_data)
  151. error_data.append(current_error)
  152. if len(previous_data) == 0:
  153. previous_data = current_data
  154. # display all data using matplotlib (configure plt)
  155. gridsize = (3, 2)
  156. # fig, (ax1, ax2) = plt.subplots(nrows=2, ncols=1, figsize=(30, 22))
  157. fig = plt.figure(figsize=(30, 22))
  158. ax1 = plt.subplot2grid(gridsize, (0, 0), colspan=2, rowspan=2)
  159. ax2 = plt.subplot2grid(gridsize, (2, 0), colspan=2)
  160. ax1.set_title(p_scene + ' scene interval information SVD['+ str(begin_data) +', '+ str(end_data) +'], from scenes indices [' + str(begin_index) + ', '+ str(end_index) + ']' + p_metric + ' metric, ' + p_mode + ', with step of ' + str(p_step) + ', svd norm ' + str(p_norm), fontsize=20)
  161. ax1.set_ylabel('Image samples or time (minutes) generation', fontsize=14)
  162. ax1.set_xlabel('Vector features', fontsize=16)
  163. for id, data in enumerate(images_data):
  164. p_label = p_scene + '_' + str(images_indices[id]) + " | " + p_error + ": " + str(error_data[id])
  165. if images_indices[id] == threshold_image_zone:
  166. ax1.plot(data, label=p_label, lw=4, color='red')
  167. else:
  168. ax1.plot(data, label=p_label)
  169. ax1.legend(bbox_to_anchor=(0.8, 1), loc=2, borderaxespad=0.2, fontsize=14)
  170. start_ylim, end_ylim = p_ylim
  171. ax1.set_ylim(start_ylim, end_ylim)
  172. ax2.set_title(p_error + " information for whole step images")
  173. ax2.set_ylabel(p_error + ' error')
  174. ax2.set_xlabel('Number of samples per pixels or times')
  175. ax2.set_xticks(range(len(images_indices)))
  176. ax2.set_xticklabels(list(map(int, images_indices)))
  177. ax2.plot(error_data)
  178. plt.show()
  179. def main():
  180. # by default p_step value is 10 to enable all photos
  181. p_step = 10
  182. p_ylim = (0, 1)
  183. if len(sys.argv) <= 1:
  184. print('Run with default parameters...')
  185. print('python display_svd_data_scene.py --scene A --interval "0,800" --indices "0, 900" --metric lab --mode svdne --step 50 --norm 0 --error mae --ylim "0, 0.1"')
  186. sys.exit(2)
  187. try:
  188. opts, args = getopt.getopt(sys.argv[1:], "hs:i:i:z:l:m:s:n:e:y", ["help=", "scene=", "interval=", "indices=", "metric=", "mode=", "step=", "norm=", "error=", "ylim="])
  189. except getopt.GetoptError:
  190. # print help information and exit:
  191. print('python display_svd_data_scene.py --scene A --interval "0,800" --indices "0, 900" --metric lab --mode svdne --step 50 --norm 0 --error mae --ylim "0, 0.1"')
  192. sys.exit(2)
  193. for o, a in opts:
  194. if o == "-h":
  195. print('python display_svd_data_scene.py --scene A --interval "0,800" --indices "0, 900" --metric lab --mode svdne --step 50 --norm 0 --error mae --ylim "0, 0.1"')
  196. sys.exit()
  197. elif o in ("-s", "--scene"):
  198. p_scene = a
  199. if p_scene not in scenes_indices:
  200. assert False, "Invalid scene choice"
  201. else:
  202. p_scene = scenes_list[scenes_indices.index(p_scene)]
  203. elif o in ("-i", "--interval"):
  204. p_interval = list(map(int, a.split(',')))
  205. elif o in ("-i", "--indices"):
  206. p_indices = list(map(int, a.split(',')))
  207. elif o in ("-m", "--metric"):
  208. p_metric = a
  209. if p_metric not in metric_choices:
  210. assert False, "Invalid metric choice"
  211. elif o in ("-m", "--mode"):
  212. p_mode = a
  213. if p_mode not in choices:
  214. assert False, "Invalid normalization choice, expected ['svd', 'svdn', 'svdne']"
  215. elif o in ("-s", "--step"):
  216. p_step = int(a)
  217. elif o in ("-n", "--norm"):
  218. p_norm = int(a)
  219. elif o in ("-e", "--error"):
  220. p_error = a
  221. elif o in ("-y", "--ylim"):
  222. p_ylim = list(map(float, a.split(',')))
  223. else:
  224. assert False, "unhandled option"
  225. display_svd_values(p_scene, p_interval, p_indices, p_metric, p_mode, p_step, p_norm, p_error, p_ylim)
  226. if __name__== "__main__":
  227. main()