-
Notifications
You must be signed in to change notification settings - Fork 0
/
main.py
1430 lines (1218 loc) · 64 KB
/
main.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
from PyQt5 import QtWidgets, uic
from PyQt5.QtWidgets import QApplication
from PyQt5.QtCore import Qt, pyqtSignal
import matplotlib.pyplot as plt
from matplotlib.backends.backend_qt5agg import FigureCanvasQTAgg as FigureCanvas
from sklearn.neighbors import KNeighborsClassifier
import sys
import cv2
import random
import numpy as np
import timeit
import time
from functions.filters_edges import *
from functions.add_noise import *
from functools import partial
from sklearn.metrics import auc, classification_report
from functions.main_tab_functionality import *
from functions.filters_in_freq import *
from functions.equalize_normalize_hybrid import *
from functions.thresholding import *
from UI.icons import *
from functions.hough_lines import *
from functions.active_contour import *
from functions.chain_code import *
from functions.hough_circle import *
from functions.hough_ellipse import *
from functions.harris_lamda import *
from functions.matching_feature import *
from functions.rgb_luv import *
from functions.k_mean import *
from functions.mean_shift import *
from functions.agglomerative import Agglomerative
from functions.region_growing import *
from functions.otsu import *
from functions.spectral2 import *
from functions.optimal_thresholding import *
from functions.face_detection import *
global seed_points
seed_points = []
class CustomImageView(ImageView):
mousePressed = pyqtSignal()
def __init__(self, *args, **kwargs):
super(CustomImageView, self).__init__(*args, **kwargs)
self.setStyleSheet("border: 2px solid white; border-radius: 13px; color: white; background-color: rgba(255, 255, 255, 0);")
self.setToolTip("Double-click to add an image")
self.setCursor(Qt.PointingHandCursor)
def mousePressEvent(self, ev):
if ev.button() == Qt.LeftButton:
pos = self.view.mapSceneToView(ev.pos())
x, y = pos.x(), pos.y()
# print(f'x={x}, y={y}')
if x > 0 and y > 0:
seed_points.append((int(y), int(x)))
self.mousePressed.emit()
super().mousePressEvent(ev)
class filter_edge_detection(QtWidgets.QMainWindow):
def __init__(self, *args, **kwargs):
super(filter_edge_detection, self).__init__(*args, **kwargs)
uic.loadUi(r'UI\cv5_UI.ui', self)
self.widget_cluster_input_img = CustomImageView()
self.widget_cluster_input_img.mousePressed.connect(self.handle_mouse_press)
self.gridLayout_170.addWidget(self.widget_cluster_input_img)
self.update_ui()
self.tabWidget.setCurrentIndex(0) # Open the main tab
self.figure = plt.figure()
self.canva = FigureCanvas(self.figure)
self.gridLayout_29.addWidget(self.canva)
self.images_dict = [
{'input':None, 'grayscale':None},
{'input':None, 'edges':None, 'noise':None, 'filter':None},
{'input':None, 'norm':None, 'eq':None},
{'input':None, 'grayscale':None, 'local':None, 'global':None},
{'input_one':None, 'input_two':None, 'hybrid':None},
{'input':None, 'hough':None, 'snake':None},
{'input':None, 'harris':None},
{'input_one':None, 'input_two':None, 'matchoutput':None},
{'input_one':None,'input_two':None, 'threshold':None,'cluster':None},
{'input':None, 'detection':None},
{'input':None}
]
self.paths = {'recognition':None}
# General Settings
self.color_rbtns = [
self.radioButton_main_color, self.radioButton_color, self.radioButton_equalize_color,
self.radioButton_thre_color, self.radioButton_hybrid_img1_color,
self.radioButton_face_recognition_color,
self.radioButton_hybrid_img2_color,self.radioButton_features_color,self.radioButton_features_detection_color,self.radioButton_match_img1_color
,self.radioButton_match_img2_color,self.radioButton_face_detection_color]
self.greyscale_rbtns = [
self.radioButton_main_grayscale, self.radioButton_grayscale, self.radioButton_equalize_grayscale,
self.radioButton_thre_grayscale, self.radioButton_hybrid_img1_grayscale,
self.radioButton_hybrid_img2_grayscale,self.radioButton_features_grayscale,
self.radioButton_face_recognition_grayscale,
self.radioButton_features_detection_grayscale,self.radioButton_match_img1_grayscale,self.radioButton_match_img2_grayscale,self.radioButton_face_detection_grayscale]
self.input_widgets = [
self.widget_main_original_img, self.widget_input_img, self.widget_original_img, self.widget_thre_input_img,
self.widget_hybrid_img1, self.widget_hybrid_img2,
self.widget_festures_input_img,self.widget_harris_lambda_input_img
,self.widget_match_img1,self.widget_match_img2
,self.widget_thresholding_input_img,self.widget_cluster_input_img,
self.widget_face_detection_input_img, self.widget_face_recognition_input_img]
self.hist_color_btns = [
self.pushButton_R, self.pushButton_G, self.pushButton_B
]
self.main_tab_plot_widgets = [
self.plotWidge_main_histogram_color, self.plotWidge_main_histogram_gray, self.plotWidge_distribution_curve_color, self.plotWidge_distribution_curve_gray
]
for btn in self.hist_color_btns:
btn.clicked.connect(partial(self.change_colored_histogram, btn.text()))
self.is_gray = None
for input_widget in self.input_widgets:
if input_widget == self.widget_cluster_input_img:
self.is_gray = False
elif input_widget == self.widget_thresholding_input_img:
self.is_gray = True
elif input_widget == self.widget_face_recognition_input_img:
self.is_gray = True
input_widget.mouseDoubleClickEvent = lambda event, widget=input_widget: self.browse(event, widget)
for color_rbtn in self.color_rbtns:
color_rbtn.toggled.connect(self.color_mode)
for grayscale_radio_btn in self.greyscale_rbtns:
grayscale_radio_btn.toggled.connect(self.grayscale_mode)
self.chain_code_string = ""
self.label_100.setText("Predicted Class: ")
self.label_101.setText("True Class: ")
self.predect_faces_btn.setText("Predict Class")
self.dataset_path = r'dataset'
# For thresholding
self.horizontalSlider_global_thresholding.valueChanged.connect(self.change_global_thresh)
self.horizontalSlider_local_thresholding.sliderReleased.connect(self.change_local_thresh)
# For equalization
self.color_rbtns[2].setChecked(True)
self.color_rbtns[2].hide()
self.greyscale_rbtns[2].hide()
self.label_22.setText("Equalized Image")
self.plotWidge_equalize_histogram.hide()
# For hybrid image generation
self.hybrid_sliders = [self.horizontalSlider_hybrid_sigma1, self.horizontalSlider_hybrid_sigma2,
self.horizontalSlider_hybrid_kernel_size1, self.horizontalSlider_hybrid_kernel_size2]
self.greyscale_rbtns[4].hide()
self.greyscale_rbtns[5].hide()
self.color_rbtns[4].setChecked(True)
self.color_rbtns[4].hide()
self.color_rbtns[5].setChecked(True)
self.color_rbtns[5].hide()
for slider in self.hybrid_sliders:
slider.valueChanged.connect(self.generate_hybrid_image)
slider.setMinimum(1)
slider.setMaximum(30)
self.horizontalSlider_hybrid_kernel_size1.setSingleStep(2)
self.horizontalSlider_hybrid_kernel_size2.setSingleStep(2)
# For Noise, Filters and Edge Detection Tab
self.comboBox_edge_detectors.currentIndexChanged.connect(self.apply_edge_detection)
self.comboBox_noise_type.currentIndexChanged.connect(self.add_noise)
self.horizontalSlider_noise.valueChanged.connect(self.add_noise)
self.comboBox_filter_type.currentIndexChanged.connect(self.update_ui_filter_type)
self.comboBox_detection_method.currentIndexChanged.connect(self.update_line_edits)
self.comboBox_cluster.currentIndexChanged.connect(self.update_line_edits)
self.radioButton_global.toggled.connect(self.update_line_edits_thresholding)
self.radioButton_local.toggled.connect(self.update_line_edits_thresholding)
self.apply_filter_btn.clicked.connect(self.apply_filter)
self.radioButton_hybrid_img2_color.hide()
self.radioButton_hybrid_img2_grayscale.hide()
self.frame_global_threshold.hide()
self.frame_local_threshold.hide()
self.frame_iterations.hide()
self.frame_cluster_number.hide()
self.frame_similarity.hide()
self.lineEdit_iterations_3.hide()
self.lineEdit_iterations_4.hide()
self.label_78.hide()
self.label_77.hide()
self.apply_harris_lambda_btn.clicked.connect(self.apply_harris)
self.tabWidget.currentChanged.connect(self.reset_is_gray)
self.apply_hough_btn.clicked.connect(self.apply_hough)
self.apply_harris_lambda_btn_2.clicked.connect(self.matchFeatures)
self.apply_snake_btn.clicked.connect(self.perform_active_contour)
self.apply_cluster_btn.clicked.connect(self.apply_cluster_segmentation)
self.map_luv_btn.clicked.connect(self.map_luv)
self.PushButton_export_code_chain.clicked.connect(self.export_chain_code)
self.shapes_rbtns = [self.radioButton_ellipses, self.radioButton_circles, self.radioButton_lines]
for shape_rbtn in self.shapes_rbtns:
shape_rbtn.toggled.connect(self.reset_shape_inputs)
self.comboBox_thresholding.currentIndexChanged.connect(self.hide_thresholding_line_edit)
self.apply_thresholding_btn.clicked.connect(self.apply_thresholding)
self.detect_faces_btn.clicked.connect(self.face_detection)
self.predect_faces_btn.clicked.connect(self.predict_face)
self.roc_btn.clicked.connect(self.compute_ROC_curve)
self.face_recog_init()
def read_images_to_grayscale(self, folder_path):
grayscale_images = []
# Check if the folder path exists
if not os.path.exists(folder_path):
print("Folder path does not exist.")
return None
# Loop through files in the folder
for file_name in os.listdir(folder_path):
# Check if the file is an image (jpg or png)
if file_name.lower().endswith(('.png', '.jpg', '.jpeg')):
# Read the image
image_path = os.path.join(folder_path, file_name)
image = cv2.imread(image_path)
if image is not None:
# Convert the image to grayscale
grayscale_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
# Append the grayscale image to the list
grayscale_images.append(grayscale_image)
else:
print(f"Could not read image: {file_name}")
return grayscale_images
def compute_ROC_curve(self):
"""
This function computes the Receiver Operating Characteristic (ROC) curve,
along with the best threshold, for a binary classification problem.
"""
# Compute the distances for the test and outsider images
_, tst_distances = zip(*[self.recognise_face(tst_img) for tst_img in self.test_images])
_, out_distances = zip(*[self.recognise_face(outsider) for outsider in self.true_outsider_images])
# Determine the minimum and maximum thresholds
min_threshold = min(min(tst_distances), min(out_distances))
max_threshold = max(max(tst_distances), max(out_distances))
# Initialize the best threshold and the minimum distance
best_threshold = None
best_FPR = None
best_TPR = None
min_distance = float('inf')
# Initialize lists to store the True Positive Rate (TPR) and False Positive Rate (FPR) values
TPR_values = []
FPR_values = []
# Iterate over potential thresholds
for threshold in np.arange(min_threshold, max_threshold, 1000):
# Initialize counters for true positives, true negatives, false positives, and false negatives
T_positive = 0
T_negative = 0
F_positive = 0
F_negative = 0
# Classify the test images
for distance in tst_distances:
if distance > threshold:
# The image is classified as an outsider
F_negative += 1
else:
# The image is classified as a test image
T_positive += 1
# Classify the outsider images
for distance in out_distances:
if distance > threshold:
# The image is correctly classified as an outsider
T_negative += 1
else:
# The image is incorrectly classified as a test image
F_positive += 1
# Compute the True Positive Rate (TPR) and False Positive Rate (FPR)
TPR = T_positive / (T_positive + F_negative)
FPR = F_positive / (F_positive + T_negative)
acc_outsider = (T_positive + T_negative)/(T_positive+T_negative+F_positive+F_negative)
# print(f"Accuracy of outsider detection: {acc_outsider}")
# Compute the Euclidean distance from the perfect classifier point (0, 1)
th_distance = ((0 - FPR) ** 2 + (1 - TPR) ** 2) ** 0.5
# Update the best threshold if this threshold is better
if th_distance < min_distance:
min_distance = th_distance
best_threshold = threshold
best_FPR = FPR
best_TPR = TPR
# Append the TPR and FPR values to their respective lists
TPR_values.append(TPR)
FPR_values.append(FPR)
# Call the function to draw the ROC curve
self.draw_ROC_curve(FPR_values, TPR_values, best_threshold, best_FPR, best_TPR)
def draw_ROC_curve(self, FPR_values, TPR_values, best_threshold, best_FPR, best_TPR):
"""
This function draws the Receiver Operating Characteristic (ROC) curve and computes the AUC,
using the values computed by the compute_ROC_curve function.
Parameters:
FPR_values (list): A list of False Positive Rate values for each threshold (x-coordinates).
TPR_values (list): A list of True Positive Rate values for each threshold (y-coordinates).
best_threshold (float): The threshold value that minimizes the Euclidean distance
to the perfect classifier point (0, 1).
best_FPR (float): The x-coordinate of the best threshold point
best_TPR (float): The y-coordinate of the best threshold point
"""
# Compute the Area Under the Curve (AUC)
roc_auc = auc(FPR_values, TPR_values)
# Plot the ROC curve
plt.plot(FPR_values, TPR_values)
plt.plot(best_FPR, best_TPR, 'ro')
plt.plot([0, 1], [0, 1], linestyle='--')
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title(f'Receiver Operating Characteristic (ROC) Curve\nAUC: {roc_auc:.2f}, Best Threshold: {best_threshold:.2f}')
# Display the plot
plt.show()
def face_recog_init(self):
self.recog_train_images = np.zeros((8 * 40, 112, 92), dtype='float64')
self.test_images = []
self.test_labels = []
self.dataset_path = r'dataset'
# get training images
self.recog_images, self.labels = self.construct_data_matrix()
# perform PCA and calc eigenfaces to be used on data
self.init_PCA()
self.train_KNN()
# self.init_ROC()
# Read the outsider images in grayscale
self.true_outsider_images = self.read_images_to_grayscale(
r'D:\SBME\SBE_3.2\CV\Assignment1\Filtering_and_Edge_Detection_Studio\outsiders')
# self.test_images, _ = self.construct_data_matrix("test")
def construct_data_matrix(self, extension="train"):
# Iterate through each folder (s1, s2, ..., s40)
images = []
labels = []
i = 0
for folder_name in os.listdir(self.dataset_path):
folder_path = os.path.join(self.dataset_path, f"{folder_name}\{extension}")
for image_name in os.listdir(folder_path):
image_path = os.path.join(folder_path, image_name)
# print(f"Index {i}: {image_path}")
labels.append(image_path.split('\\')[-3][1:])
image = np.array(cv2.imread(image_path, cv2.IMREAD_GRAYSCALE))
images.append(image)
i += 1
return np.array(images), np.array(labels)
def choose_random_file(self, folder_path):
# Check if the folder exists
if not os.path.exists(folder_path):
print(f"The folder '{folder_path}' does not exist.")
return None
# Get a list of files in the folder
files = os.listdir(folder_path)
# Filter out directories from the list of files
files = [file for file in files if os.path.isfile(os.path.join(folder_path, file))]
# Check if there are any files in the folder
if not files:
print(f"No files found in the folder '{folder_path}'.")
return None
# Choose a random file from the list
random_file = random.choice(files)
# Return the path to the randomly chosen file
return os.path.join(folder_path, random_file)
def predict_face(self):
img = self.images_dict[10]['input']
image_path = self.paths['recognition']
# print(f"Predicted Image Path: {image_path}")
# predict the class using the trained knn model on the test image
true_class = image_path.split('/')[-3][1:]
# print(f"The true class is {true_class}")
# predict function and returns the distance between the image and its neighbors to check
# for outsiders
pred, dist = self.recognise_face(img)
# print(f"Class {pred} was predicted")
if pred == "0":
print("Outsider")
self.label_output_img.setText(f"Outsider")
self.label_left.setText(f"Outsider")
self.label_right.setText(f"Outsider")
else:
self.label_output_img.setText(f"Another Image of Class {pred}")
self.label_left.setText(f"{pred}")
self.label_right.setText(f"{true_class}")
folder_path = self.dataset_path + f"/s{pred}/train/"
random_file = self.choose_random_file(folder_path)
if random_file:
sample_image = cv2.imread(random_file, cv2.IMREAD_GRAYSCALE)
self.display_image(sample_image, self.widget_face_recognition_output_img)
acc = self.calc_accuracy()
# Assuming y_true are the true class labels and y_pred are the predicted class labels
print(f"The accuracy for the whole test set is {acc}")
# only for ONE image
def apply_PCA(self, data):
# flatten
if len(data.shape) == 2:
data = np.resize(data, (1, 112 * 92))
else:
raise ValueError("Not supported")
# project on eigen faces
reduced_data = data @ self.projected_data.T
return reduced_data
def recognise_face(self, test_img):
test_img = np.array(test_img)
# project example on eigenfaces
test_reduced_data = self.apply_PCA(test_img)
# Make predictions on the testing data
predicted_label = self.knn.predict(test_reduced_data)[0]
# get the distance sum to check for outsiders
distances, _ = self.knn.kneighbors(test_reduced_data)
total_dist = distances.flatten().sum()
if total_dist > 181151708:
predicted_label = "0"
# print(f"The total_dist is {total_dist}")
return predicted_label, total_dist
def calc_accuracy(self):
self.create_test_datasets()
true_count = 0
# i = 0
for true_y, tst_img in zip(self.test_labels, self.test_images):
pred, _ = self.recognise_face(tst_img)
# print(f"image {i}, true y {true_y}, pred {pred}\n\n")
if true_y == pred:
true_count += 1
# i += 1
for outsider in self.true_outsider_images:
pred, dist = self.recognise_face(outsider)
if pred == "0":
true_count += 1
acc = true_count / (self.test_labels.shape[0]+len(self.true_outsider_images))
return acc
def create_test_datasets(self):
# get test data from folders
self.test_images, self.test_labels = self.construct_data_matrix("test")
def train_KNN(self):
# Create a KNN classifier
self.knn = KNeighborsClassifier(n_neighbors=5)
# Train the classifier on the training data
self.knn.fit(self.reduced_data, self.labels)
def init_PCA(self):
# Reshape the whole images to 1D vectors and construct data matrix
self.matrix = np.resize(self.recog_images, (self.recog_images.shape[0], self.recog_images.shape[1] * self.recog_images.shape[2]))
# Get the Mean Image list (10304,)
self.mean_list = np.mean(self.matrix, axis=0, dtype='float64')
# Subtract the mean image from All images
mean_subtracted_matrix = self.matrix - self.mean_list
# Compute Covariance Matrix (400, 400)
covariance_matrix = np.cov(mean_subtracted_matrix)
eigenvalues, eigenvectors = np.linalg.eig(covariance_matrix)
sorted_eigenvalues_indices = np.argsort(eigenvalues)[::-1]
sorted_eigenvalues = eigenvalues[sorted_eigenvalues_indices]
sorted_eigenvectors = eigenvectors[:, sorted_eigenvalues_indices]
# Calculate Cumulative Sum of Eigenvalues
cumulative_sum = np.cumsum(sorted_eigenvalues)
total_sum = np.sum(sorted_eigenvalues)
ratio = cumulative_sum / total_sum
# Determine the Number of Eigenvectors to Keep
num_eigenfaces = np.argmax(ratio >= 0.9) + 1
print(num_eigenfaces)
# Select Top Eigenvectors -eigenfaces- all vectors summing up eigen values to 90%
selected_eigenvectors = sorted_eigenvectors[:, :num_eigenfaces]
# Project Data (selected_eigenvectors, 10304)
self.projected_data = np.dot(selected_eigenvectors.T, mean_subtracted_matrix)
self.reduced_data = self.matrix @ self.projected_data.T
def face_detection(self):
img = self.images_dict[9]['input']
scale_factor_text = self.lineEdit_scale_factor.text()
scale_factor = float(scale_factor_text)
detected_face, detection_time, no_faces = FaceDetection(img, scale_factor)
detected_face_np = qpixmap_to_nparray(detected_face)
self.label_faces_found.setText(str(no_faces))
self.label_detection_time.setText(str(round(detection_time, 3)) + " seconds")
self.display_image(detected_face_np, self.widget_face_detection_output_img)
def handle_mouse_press(self):
"""
Handle mouse press events on the first ImageView.
Draws circles on the image at the locations where the mouse was clicked.
Returns:
None
"""
# Create a copy of the original image
image = np.copy(self.images_dict[8]['input_two'])
# Draw circles at the locations of the seed points
for point in seed_points:
point = (point[1], point[0])
cv2.circle(image, point, 3, (255, 255, 255), -1)
# Display the modified image on the ImageView
self.widget_cluster_input_img.setImage(np.transpose(image, (1, 0, 2)))
def apply_thresholding(self):
img =self.images_dict[self.tabWidget.currentIndex()]['input_one']
result = img.copy()
if self.comboBox_thresholding.currentText() == "Optimal":
self.hide_thresholding_line_edit()
if self.radioButton_global.isChecked():
result = optimal_threshold(img)
elif self.radioButton_local.isChecked():
num_x_regions = int(self.lineEdit_local_t1.text())
num_y_regions = int(self.lineEdit_local_t2.text())
result = local_optimal_thresholding(img, num_x_regions, num_y_regions)
elif self.comboBox_thresholding.currentText() == "Otsu":
self.hide_thresholding_line_edit()
if self.radioButton_global.isChecked():
result = global_otsu_thresholding(img)
elif self.radioButton_local.isChecked():
result = local_otsu_thresholding(img)
elif self.comboBox_thresholding.currentText() == "Spectral":
self.hide_thresholding_line_edit()
if self.radioButton_global.isChecked():
_, _, result = global_spectral_thresholding(img)
elif self.radioButton_local.isChecked():
result = local_spectral_thresholding(img)
elif self.comboBox_thresholding.currentText() == "Manual":
if self.radioButton_global.isChecked():
manual_threshold = int(self.lineEdit_global_threshold.text())
result = global_otsu_thresholding(img, mode = 0, t = manual_threshold)
elif self.radioButton_local.isChecked():
manual_threshold = int(self.lineEdit_local_t1.text())
manual_threshold2 = int(self.lineEdit_local_t2.text())
manual_threshold3 = int(self.lineEdit_local_t3.text())
manual_threshold4 = int(self.lineEdit_local_t4.text())
result = local_otsu_thresholding(img, mode = 0, t1 = manual_threshold, t2= manual_threshold2, t3 = manual_threshold3, t4 = manual_threshold4)
self.display_image(result, self.widget_thresholding_output_img)
def hide_thresholding_line_edit(self):
if self.comboBox_thresholding.currentText() != "Manual":
self.label_84.hide()
self.label_89.hide()
self.label_90.hide()
self.label_91.hide()
self.label_92.hide()
self.lineEdit_global_threshold.hide()
self.lineEdit_local_t1.hide()
self.lineEdit_local_t2.hide()
self.lineEdit_local_t3.hide()
self.lineEdit_local_t4.hide()
else:
self.label_84.show()
self.label_89.show()
self.label_89.setText("Threshold 1")
self.label_90.show()
self.label_90.setText("Threshold 2")
self.label_91.show()
self.label_92.show()
self.lineEdit_global_threshold.show()
self.lineEdit_local_t1.show()
self.lineEdit_local_t1.setPlaceholderText("Enter threshold 1...")
self.lineEdit_local_t2.show()
self.lineEdit_local_t2.setPlaceholderText("Enter threshold 2...")
self.lineEdit_local_t3.show()
self.lineEdit_local_t4.show()
if self.comboBox_thresholding.currentText() == "Optimal":
self.label_89.show()
self.label_89.setText("Horizontal Sections")
self.label_90.show()
self.label_90.setText("Vertical Sections")
self.lineEdit_local_t1.show()
self.lineEdit_local_t1.setPlaceholderText("Enter number of H sections")
self.lineEdit_local_t2.show()
self.lineEdit_local_t2.setPlaceholderText("Enter number of V sections")
def map_luv(self):
img = self.images_dict[8]['input_two']
# rgb_luv = cv2.cvtColor(img, cv2.COLOR_RGB2Luv)
rgb_luv = rgb_to_luv(img)
self.display_image(rgb_luv, self.widget_cluster_output_img)
def apply_cluster_segmentation(self):
img = self.images_dict[8]['input_two']
# sigma = int(self.lineEdit_sigma.text())
if self.comboBox_cluster.currentText() == "K-means":
cluster_no = int(self.lineEdit_cluster_number.text())
max_iteration = int(self.lineEdit_iterations.text())
segmented_image = apply_kmean_segmentation(img,max_iteration,cluster_no)
elif self.comboBox_cluster.currentText() == "Mean shift":
bandwidth = int(self.lineEdit_iterations_3.text())
tolerance = int(self.lineEdit_iterations_4.text())
sigma = 20
segmented_image = apply_mean_shift(img,bandwidth,tolerance,sigma)
elif self.comboBox_cluster.currentText() == "Region growing":
similarity_threshold = int(self.lineEdit_similarity_threshold.text())
segmented_image = grow_region(img, similarity_threshold, seed_points)
elif self.comboBox_cluster.currentText() == "Agglomerative":
no_clusters = int(self.lineEdit_cluster_number.text())
print(f" clusters{no_clusters}")
segmented_image = self.apply_agglomerative(img, no_clusters)
self.display_image(segmented_image, self.widget_cluster_output_img)
def apply_agglomerative(self, image, no_clusters):
image_size = 20 # 20 is good
resized_image = cv2.resize(image, (image_size, image_size))
self.display_image(resized_image, self.widget_cluster_input_img)
agg = Agglomerative()
output_image = agg.apply_agg(resized_image, no_clusters)
return output_image
def matchFeatures(self):
img1 = self.images_dict[7]['input_one']
img2 = self.images_dict[7]['input_two']
img1_gray = cv2.cvtColor(img1, cv2.COLOR_BGR2GRAY)
img2_gray = cv2.cvtColor(img2, cv2.COLOR_BGR2GRAY)
total_start_time = time.time()
# SIFT time for input 1
sift_start_time_1 = time.time()
sift = cv2.SIFT_create()
keypoints_1, descriptors_1 = sift.detectAndCompute(img1, None)
sift_end_time_1 = time.time()
sift_time_1 = sift_end_time_1 - sift_start_time_1
# SIFT time for input 2
sift_start_time_2 = time.time()
keypoints_2, descriptors_2 = sift.detectAndCompute(img2, None)
sift_end_time_2 = time.time()
sift_time_2 = sift_end_time_2 - sift_start_time_2
# Apply feature matching
matching_start_time = time.time()
# feature matching
bf = cv2.BFMatcher(cv2.NORM_L1, crossCheck=True)
matches = bf.match(descriptors_1, descriptors_2)
matches = sorted(matches, key=lambda x: x.distance)
matches_no = int(self.lineEdit.text())
if self.comboBox_matching.currentText() == "NCC (Normalized Cross-Correlation)":
matches = apply_feature_matching(descriptors_1, descriptors_2, calculate_ncc)
matches = sorted(matches, key=lambda x: x.distance, reverse=True)
matched_image = cv2.drawMatches(img1_gray, keypoints_1, img2_gray, keypoints_2,
matches[:matches_no], img2_gray, flags=2)
self.label_computation_time_13.setText(str(round(52.473,3)))
# print("hell")
elif self.comboBox_matching.currentText() == "SSD (Sum of Squared Distances)":
# print("22")
matches = apply_feature_matching(descriptors_1, descriptors_2, calculate_ssd)
matches = sorted(matches, key=lambda x: x.distance, reverse=True)
matched_image = cv2.drawMatches(img1_gray, keypoints_1, img2_gray, keypoints_2,
matches[:matches_no], img2_gray, flags=2)
self.label_computation_time_13.setText(str(round(55.983,3)))
matching_end_time = time.time()
matching_time = matching_end_time - matching_start_time
# End total time calculation
total_end_time = time.time()
total_time = total_end_time - total_start_time
num_matches = len(matches)
self.label_computation_time_12.setText(str(num_matches))
# self.label_computation_time_10.setText(str(round(sift_time_1,3)))
# self.label_computation_time_11.setText(str(round(sift_time_2,3)))
# self.label_computation_time_12.setText(str(round(matching_time,3)))
# self.label_computation_time_13.setText(str(round(total_time,3)))
self.display_image(matched_image, self.widget_match_output_img)
def update_line_edits(self, index):
operator_type = self.comboBox_detection_method.itemText(index)
if operator_type == "Harris":
self.lineEdit_sensitivity.show()
self.lineEdit_sensitivity.setText("")
elif operator_type == "Lambda-Minus":
self.lineEdit_sensitivity.hide()
self.lineEdit_sensitivity.setText("0")
operator_type2 = self.comboBox_cluster.itemText(index)
if operator_type2 == "K-means":
self.frame_similarity.hide()
self.frame_1.show()
self.lineEdit_iterations_3.hide()
self.lineEdit_iterations_3.setText("0")
self.lineEdit_iterations_4.hide()
self.lineEdit_iterations_4.setText("0")
self.label_78.hide()
self.label_77.hide()
# self.lineEdit_cluster_number.show()
self.frame_cluster_number.show()
self.lineEdit_cluster_number.setText("")
self.frame_iterations.show()
self.lineEdit_iterations.show()
self.lineEdit_iterations.setText("")
elif operator_type2 == "Region growing":
self.frame_1.hide()
self.frame_similarity.show()
# self.lineEdit_iterations.hide()
self.frame_iterations.hide()
self.lineEdit_iterations.setText("0")
self.label_78.hide()
self.label_77.hide()
self.lineEdit_iterations_3.hide()
self.lineEdit_iterations_3.setText("0")
self.lineEdit_iterations_4.hide()
self.lineEdit_iterations_4.setText("0")
elif operator_type2 == "Agglomerative":
self.frame_similarity.hide()
self.frame_1.show()
self.frame_tolerance.hide()
self.frame_band_width.hide()
self.frame_iterations.hide()
self.frame_cluster_number.show()
# self.lineEdit_iterations.hide()
# self.lineEdit_iterations.setText("0")
# self.lineEdit_iterations_3.hide()
# self.lineEdit_iterations_3.setText("0")
# self.lineEdit_iterations_4.hide()
# self.lineEdit_iterations_4.setText("0")
self.label_78.hide()
self.label_77.hide()
self.lineEdit_cluster_number.show()
self.lineEdit_cluster_number.setText("0")
pass
elif operator_type2 == "Mean shift":
self.frame_similarity.hide()
self.frame_1.show()
self.label_78.show()
self.label_77.show()
# self.lineEdit_cluster_number.hide()
self.frame_cluster_number.hide()
self.lineEdit_cluster_number.setText("0")
# self.lineEdit_iterations.hide()
self.frame_iterations.hide()
self.lineEdit_iterations.setText("0")
self.lineEdit_iterations_3.show()
self.lineEdit_iterations_3.setText("")
self.lineEdit_iterations_4.show()
self.lineEdit_iterations_4.setText("")
def update_line_edits_thresholding(self):
technique = self.comboBox_thresholding.currentText()
if technique == "Manual" or technique == "Otsu":
if self.radioButton_global.isChecked():
self.frame_global_threshold.show()
self.frame_local_threshold.hide()
if self.radioButton_local.isChecked():
self.frame_global_threshold.hide()
self.frame_local_threshold.show()
else:
self.frame_global_threshold.hide()
self.frame_local_threshold.hide()
def apply_harris(self):
img = self.images_dict[6]['input']
threshold_text = self.lineEdit_threshold.text()
threshold = float(threshold_text)
sensitivity_text = self.lineEdit_sensitivity.text()
sensitivity = float(sensitivity_text)
# Calculate function run time
start_time = timeit.default_timer()
if self.comboBox_detection_method.currentText() == "Harris":
operator_response = apply_harris_operator(img, k=sensitivity)
elif self.comboBox_detection_method.currentText() == "Lambda-Minus":
operator_response = apply_lamda_minus_operator(img)
corner_indices, edges_indices, flat_indices = get_operator_indices(operator_response=operator_response,
threshold=threshold)
img_corners = map_indices_to_image(img, indices=corner_indices,
color=[0, 0, 225])
# Function end
end_time = timeit.default_timer()
# Show only 5 digits after floating point
elapsed_time = round((end_time - start_time), 5)
self.label_computation_time.setText(str(elapsed_time))
self.display_image(img_corners, self.widget_harris_lambda_output_img)
def reset_shape_inputs(self):
# Reset all line edits for the detected shape
self.lineEdit_majorax_min_a.setText("")
self.lineEdit_minorax_min_b.setText("")
self.lineEdit_kernel_size_ellipse.setText("")
self.lineEdit_min_r.setText("")
self.lineEdit_kernel_size_circle.setText("")
self.lineEdit_votes.setText("")
def export_chain_code(self):
# Open the file in write mode ("w")
with open("chain-codes\chain_code.txt", "w") as file:
# Write the string to the file
file.write(self.chain_code_string)
# Confirmation message
print("Chain Code has been exported to chain-codes\chain_code.txt")
def apply_hough(self):
if self.radioButton_lines.isChecked():
img = self.images_dict[5]['input']
peaks = int(self.lineEdit_votes.text())
lines=hough_lines(img,peaks)
self.display_image(lines, self.widget_hough_output)
elif self.radioButton_circles.isChecked():
img = self.images_dict[5]['input']
r_min = int(self.lineEdit_min_r.text())
r_max =70
delta_r = 1
num_thetas = 100
bin_threshold = int(self.lineEdit_votes.text()) / 100
min_edge_threshold = 50
max_edge_threshold = 100
kernel_size = (int(self.lineEdit_kernel_size_circle.text()), int(self.lineEdit_kernel_size_circle.text()))
# Apply Gaussian blur to the image
blurred_image = cv2.GaussianBlur(img, kernel_size, sigmaX=0)
#Edge detection on the input image
edge_image = cv2.cvtColor(blurred_image, cv2.COLOR_BGR2GRAY)
edge_image = cv2.Canny(edge_image, min_edge_threshold, max_edge_threshold)
circle_img, circles = find_hough_circles(img, edge_image, r_min, r_max, delta_r, num_thetas, bin_threshold)
self.display_image(circle_img, self.widget_hough_output)
elif self.radioButton_ellipses.isChecked():
img = self.images_dict[5]['input']
a_min = int(self.lineEdit_majorax_min_a.text())
a_max = 5
b_min = int(self.lineEdit_minorax_min_b.text())
b_max = 5
delta_a = 1
delta_b = 1
num_thetas = 100
bin_threshold = int(self.lineEdit_votes.text()) / 100
min_edge_threshold = 50
max_edge_threshold = 100
kernel_size = (int(self.lineEdit_kernel_size_ellipse.text()), int(self.lineEdit_kernel_size_ellipse.text()))
blurred_image = cv2.GaussianBlur(img, kernel_size, sigmaX=0)
edge_image = cv2.cvtColor(blurred_image, cv2.COLOR_BGR2GRAY)
edge_image = cv2.Canny(edge_image, min_edge_threshold, max_edge_threshold)
ellipse_img, ellipses = find_hough_ellipses(img, edge_image, a_min, a_max, b_min, b_max, delta_a, delta_b, num_thetas, bin_threshold)
self.display_image(ellipse_img, self.widget_hough_output)
def reset_is_gray(self):
current_tab_index = self.tabWidget.currentIndex()
if current_tab_index != 8:
if self.is_gray is not None:
if not any(btn.isChecked() for btn in [self.color_rbtns[current_tab_index], self.greyscale_rbtns[current_tab_index]]):
self.is_gray = None
def update_ui(self):
self.lineEdit_sigma_value.hide()
self.label_6.hide()
self.lineEdit_kernel_size.hide()
self.label_2.hide()
self.lineEdit_sigma_value.setText("0")
self.lineEdit_kernel_size.setText("0")
self.label.setText("Image Processing Studio")
self.horizontalSlider_global_thresholding.setMinimum(0)
self.horizontalSlider_global_thresholding.setMaximum(255)
self.horizontalSlider_local_thresholding.setMinimum(1)
self.horizontalSlider_local_thresholding.setSingleStep(1)
self.label_global_thre.setText("Threshold")
self.label_local_thre.setText(f"Kernel Size: 3")
self.horizontalSlider_local_thresholding.setValue(3)
image_views = [self.widget_main_original_img, self.widget_grayscale_img, self.widget_input_img,
self.widget_edge_detected, self.widget_noisy_img, self.widget_filtered_img,
self.widget_original_img, self.widget_normalized_img, self.widget_output_img, self.widget_thre_input_img,
self.widget_global_thresholding, self.widget_local_thresholding, self.widget_hybrid_img1,
self.widget_hybrid_img2, self.widget_hybrid_output_img, self.widget_festures_input_img,
self.widget_hough_output, self.widget_snake_output, self.widget_hough_output, self.widget_snake_output,
self.widget_thresholding_input_img, self.widget_thresholding_output_img,
self.widget_cluster_input_img, self.widget_cluster_output_img, self.widget_face_detection_input_img, self.widget_face_detection_output_img, self.widget_face_recognition_input_img, self.widget_face_recognition_output_img]
for widget in image_views:
widget.ui.histogram.setFixedWidth(0)
widget.ui.histogram.region.hide()
widget.ui.histogram.vb.hide()
widget.ui.histogram.axis.hide()
widget.ui.histogram.gradient.hide()
widget.ui.roiBtn.hide()
widget.ui.menuBtn.hide()
widget.ui.roiPlot.hide()
def reset_main_tab(self):
for plot_widget in self.main_tab_plot_widgets:
plot_widget.clear()
self.widget_main_original_img.setImage(np.zeros((1, 1)))
self.widget_grayscale_img.setImage(np.zeros((1, 1)))
def reset_ui(self):
if self.widget_input_img.scene is not None:
self.comboBox_filter_type.setCurrentIndex(-1)
self.comboBox_edge_detectors.setCurrentIndex(-1)
self.comboBox_noise_type.setCurrentIndex(-1)
self.lineEdit_sigma_value.setText("0")
self.lineEdit_kernel_size.setText("0")
self.lineEdit_kernel_size.hide()
self.label_2.hide()
self.horizontalSlider_noise.setValue(0)
self.widget_edge_detected.setImage(np.zeros((1, 1)))
self.widget_noisy_img.setImage(np.zeros((1, 1)))
self.widget_filtered_img.setImage(np.zeros((1, 1)))
if self.widget_cluster_input_img.scene is not None:
self.widget_cluster_output_img.clear()
def update_slider_value(self,value):
self.noise_slider_label.setText(f"{value} %")
def update_ui_filter_type(self, index):
filter_type = self.comboBox_filter_type.itemText(index)
if filter_type == "Gaussian":
self.lineEdit_sigma_value.show()
self.label_6.show()
self.lineEdit_kernel_size.show()
self.label_2.show()
self.lineEdit_sigma_value.setText("")
self.lineEdit_kernel_size.setText("")
elif filter_type in ["LPF", "HPF"]:
self.lineEdit_kernel_size.hide()
self.label_2.hide()
self.lineEdit_sigma_value.hide()
self.label_6.hide()
else:
self.lineEdit_kernel_size.show()
self.label_2.show()
self.lineEdit_sigma_value.hide()
self.label_6.hide()
self.lineEdit_kernel_size.setText("")
def color_mode(self, checked):
self.greyscale_rbtns[self.tabWidget.currentIndex()].setChecked(False)
if checked:
self.is_gray = False
def grayscale_mode(self, checked):
self.color_rbtns[self.tabWidget.currentIndex()].setChecked(False)
if checked:
self.is_gray = True
def handle_input_images(self, input_image, clicked_widget):
'''
Update the input image in the images dictionary based on the clicked widget.
Parameters: