/
symmetryCalculation.py
374 lines (335 loc) · 23.2 KB
/
symmetryCalculation.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
import numpy as np
import cv2
import pdb
from IntersectionPoint import LineSegmentIntersection, Perpendicular, CurveLineIntersection, DistancePointLine, lineMagnitude
from helperFunctions import find_x
#Finds symmetry percentage considering face boundary and central symmetry line
# We need the input image, the face curve points(faceCurve) and the two intersection points(of the eye line)
def symmetryCalculationIntensity(face_boundary_points,input_img, eye_line_point1_woNP,eye_line_point2_woNP, x_symmetry, y_symmetry, vx_symmetry, vy_symmetry):
# We first need to get the points of the symmetry line
symmetry_point1_woNP = [(x_symmetry - 200*vx_symmetry),(y_symmetry - 200*vy_symmetry)]
symmetry_point2_woNP = [(x_symmetry + 200*vx_symmetry),(y_symmetry + 200*vy_symmetry)]
# symmetry_point1_woNP = x_symmetry - 200*vx_symmetry,y_symmetry - 200*vy_symmetry
# symmetry_point2_woNP = x_symmetry + 200*vx_symmetry,y_symmetry + 200*vy_symmetry
# symmetry_point1 = np.array([(x_symmetry - 200*vx_symmetry),(y_symmetry - 200*vy_symmetry)])
# symmetry_point2 = np.array([(x_symmetry + 200*vx_symmetry),(y_symmetry + 200*vy_symmetry)])
# symmetry_point1 = np.array(symmetry_point1_woNP)
# symmetry_point2 = np.array(symmetry_point2_woNP)
symmetry_point1 = symmetry_point1_woNP
symmetry_point2 = symmetry_point2_woNP
# eye_line_point1 = np.array([(eye_line_point1[0]),(eye_line_point1[1])])
# eye_line_point2 = np.array([(eye_line_point2[0]),(eye_line_point2[1])])
# eye_line_point1 = np.array(eye_line_point1_woNP)
# eye_line_point2 = np.array(eye_line_point2_woNP)
#
# # so now a hack - which also didn't work !
# eye_line_point1 = np.transpose(eye_line_point1)
# eye_line_point2 = np.transpose(eye_line_point2)
eye_line_point1 = eye_line_point1_woNP
eye_line_point2 = eye_line_point2_woNP
# we come down in y axis
row, col, depth = input_img.shape
if col<row:
input_img = np.transpose(input_img)
else:
pass
# theta = 0
# pdb.set_trace()
# We know that eye_line_point1 is going to be on the left side of the face
points_count = len(face_boundary_points[0,:])
left_boundary = face_boundary_points[:,0:((points_count/2)-1)]
right_boundary = face_boundary_points[:, (points_count/2):(points_count-1)]
# We need to scan all the points on the left boundary which have y value greater than eye_line_point1
sum = 0
symmetry_point1 = np.asarray(symmetry_point1)
symmetry_point2 = np.asarray(symmetry_point2)
symmetry_point1 = symmetry_point1.astype(int)
symmetry_point2 = symmetry_point2.astype(int)
# symmetry_perpendicular_intersection = LineSegmentIntersection([int(symmetry_point1[0]), int(symmetry_point1[1])], [int(symmetry_point2[0]), int(symmetry_point2[1])], eye_line_point1, eye_line_point2)
eye_line_point1 = np.asarray(eye_line_point1)
eye_line_point2 = np.asarray(eye_line_point2)
eye_line_point1 = eye_line_point1.reshape((2,1))
eye_line_point2 = eye_line_point2.reshape((2,1))
eye_line_point1 = eye_line_point1.astype(int)
eye_line_point2 = eye_line_point2.astype(int)
# pdb.set_trace()
symmetry_perpendicular_intersection = LineSegmentIntersection(symmetry_point1, symmetry_point2, eye_line_point1, eye_line_point2)
symmetry_perpendicular_intersection = np.asarray(symmetry_perpendicular_intersection)
symmetry_perpendicular_intersection = symmetry_perpendicular_intersection.astype(int)
sum_left = 0
sum_right = 0
for i in range(0, (points_count/2) - 1):
# The above loop will run if eye_line_point1[1]< eye_line_point2[1]
if (eye_line_point1[1]<eye_line_point2[1]) or (eye_line_point1[1] == eye_line_point2[1]):
if face_boundary_points[1, i] < eye_line_point1[1]:
print 'pass ',i
pass
elif (face_boundary_points[1, i] < symmetry_perpendicular_intersection[1]) and (
eye_line_point1[1] < face_boundary_points[1, i]):
xf = find_x(face_boundary_points[1, i], eye_line_point1, eye_line_point2)
sum_left = sum_left + np.sum(input_img[face_boundary_points[0, i]:xf, face_boundary_points[1, i]])
print 'cond 1'
elif ((symmetry_perpendicular_intersection[1] < face_boundary_points[1, i]) and (
face_boundary_points[1, i] < eye_line_point2[1])):
xf = find_x(face_boundary_points[1, i], eye_line_point1, eye_line_point2)
xg = find_x(face_boundary_points[1, i], symmetry_point1, symmetry_point2)
sum_left = sum_left + np.sum(input_img[face_boundary_points[0, i]:xg, face_boundary_points[1, i]])
sum_right = sum_right + np.sum(input_img[xg:xf, face_boundary_points[1, i]])
print 'cond 2'
elif face_boundary_points[1, i] > eye_line_point2[1]:
xg = find_x(face_boundary_points[1, i], symmetry_point1, symmetry_point2)
sum_left = sum_left + np.sum(input_img[face_boundary_points[0, i]:xg, face_boundary_points[1, i]])
sum_right = sum_right + np.sum(input_img[xg:right_boundary[0, i], face_boundary_points[1, i]])
# sum_right = sum_right + np.sum(input_img[xg:face_boundary_points[1,i], face_boundary_points[1,i]])
print 'cond 3'
# total_sum = sum_left + sum_right
# # pdb.set_trace()set_trace
# left_percentage = (sum_left/total_sum)*100
# right_percentage = (sum_right/total_sum)*100
# total_sum = sum_left + sum_right
# # pdb.set_trace()set_trace
# left_percentage = (sum_left/total_sum)*100
# right_percentage = (sum_right/total_sum)*100
# For the case where eye_line_point1[1]>=eye_line_point2[1]:
else:
print 'Case 2 - left tilted face'
if face_boundary_points[1, i] < eye_line_point2[1]:
pass
elif (face_boundary_points[1, i] < symmetry_perpendicular_intersection[1]) and (
eye_line_point2[1] < face_boundary_points[1, i]):
xf = find_x(face_boundary_points[1, i], eye_line_point1, eye_line_point2)
sum_right = sum_right + np.sum(input_img[xf:right_boundary[0, i], face_boundary_points[1, i]])
#
# xg = find_x(face_boundary_points[1], symmetry_point1, symmetry_point2)
# sum_left = sum_left + np.sum(input_img[face_boundary_points[0,i]:xg, face_boundary_points[1,i]])
# sum_right = sum_right + np.sum(input_img[xg:xf, face_boundary_points[1,i]])
elif ((symmetry_perpendicular_intersection[1] < face_boundary_points[1, i]) and (
face_boundary_points[1, i] < eye_line_point1[1])):
xf = find_x(face_boundary_points[1, i], eye_line_point1, eye_line_point2)
xg = find_x(face_boundary_points[1, i], symmetry_point1, symmetry_point2)
sum_left = sum_left + np.sum(input_img[xf:xg, face_boundary_points[1, i]])
sum_right = sum_right + np.sum(input_img[xg:right_boundary[0, i], face_boundary_points[1, i]])
elif face_boundary_points[1, i] > eye_line_point1[1]:
xg = find_x(face_boundary_points[1, i], symmetry_point1, symmetry_point2)
# pdb.set_trace()
print "face_boundary_points[0,i] = ", face_boundary_points[0,i]
print "xg = ", xg
sum_left = sum_left + np.sum(input_img[face_boundary_points[0,i]:xg, face_boundary_points[1,i]])
sum_right = sum_right + np.sum(input_img[xg:right_boundary[0,i], face_boundary_points[1,i]])
total_sum = sum_left + sum_right
# pdb.set_trace()
left_percentage = (sum_left/total_sum)*100
right_percentage = (sum_right/total_sum)*100
return left_percentage, right_percentage
def symmetryCalculationBoundaryDifference(face_boundary_points, input_img, eye_line_point1, eye_line_point2, x_symmetry,
y_symmetry, vx_symmetry, vy_symmetry):
# We first need to get the points of the symmetry line
symmetry_point1 = [(x_symmetry - 200 * vx_symmetry), (y_symmetry - 200 * vy_symmetry)]
symmetry_point2 = [(x_symmetry + 200 * vx_symmetry), (y_symmetry + 200 * vy_symmetry)]
# we come down in y axis
# row, col, depth = input_img.shape
# if col < row:
# input_img = np.transpose(input_img)
# else:
# pass
points_count = len(face_boundary_points[0, :])
left_boundary = face_boundary_points[:, 0:((points_count / 2) - 1)]
right_boundary = face_boundary_points[:, (points_count / 2):(points_count - 1)]
# We need to scan all the points on the left boundary which have y value greater than eye_line_point1
sum = 0
symmetry_point1 = np.asarray(symmetry_point1)
symmetry_point2 = np.asarray(symmetry_point2)
symmetry_point1 = symmetry_point1.astype(int)
symmetry_point2 = symmetry_point2.astype(int)
eye_line_point1 = np.asarray(eye_line_point1)
eye_line_point2 = np.asarray(eye_line_point2)
eye_line_point1 = eye_line_point1.reshape((2, 1))
eye_line_point2 = eye_line_point2.reshape((2, 1))
eye_line_point1 = eye_line_point1.astype(int)
eye_line_point2 = eye_line_point2.astype(int)
# pdb.set_trace()
symmetry_perpendicular_intersection = LineSegmentIntersection(symmetry_point1, symmetry_point2, eye_line_point1,
eye_line_point2)
symmetry_perpendicular_intersection = np.asarray(symmetry_perpendicular_intersection)
symmetry_perpendicular_intersection = symmetry_perpendicular_intersection.astype(int)
sum_left = 0
sum_right = 0
y_init = symmetry_perpendicular_intersection[1]
y_jump = 1
# from face_boundary_points find out the maximum value of y_max and replace it below
y_max = max(face_boundary_points[1,:])
y_value = y_init
perpendicular_vectors = Perpendicular([vx_symmetry, vy_symmetry])
DistanceDifference = []
# print "Reached symmetryCalculationBoundaryDifference"
for i in range(y_init, y_max, y_jump):
x_value = find_x(y_value, symmetry_point1, symmetry_point2)
perpendicular_point_1 = [(x_value - 200 * perpendicular_vectors[0]), (y_value - 200 * perpendicular_vectors[1])]
perpendicular_point_2 = [(x_value + 200 * perpendicular_vectors[0]), (y_value + 200 * perpendicular_vectors[1])]
left_intersection_point = CurveLineIntersection(left_boundary, perpendicular_point_1, perpendicular_point_2)
right_intersection_point = CurveLineIntersection(right_boundary, perpendicular_point_1, perpendicular_point_2)
dist_left = np.linalg.norm(left_intersection_point-np.asarray([x_value, y_value]).reshape(1,2))
dist_right = np.linalg.norm(right_intersection_point-np.asarray([x_value, y_value]).reshape(1,2))
leftRightDifference = dist_left - dist_right
DistanceDifference.append(leftRightDifference)
# print "Reached symmetryCalculationBoundaryDifference after for"
pdb.set_trace()
sumNegative = 0
for number in DistanceDifference:
if number < 0:
sumNegative += number
# print "Reached symmetryCalculationBoundaryDifference after second for"
sumPositive = 0
for number in DistanceDifference:
if number >= 0:
sumPositive += number
# sumNegative = sum(number2 for number2 in DistanceDifference if number2 < 0)
positivePercentage = 100*(sumPositive/(sumPositive+sumNegative))
negativePercentage = 100*(sumNegative/(sumPositive+sumNegative))
pdb.set_trace()
return negativePercentage, positivePercentage
def symmetryCalculationLandmarkPoints(points, x_symmetry,y_symmetry,vx_symmetry,vy_symmetry):
# cv2.line(img, (points[18][0],points[18][1]), (points[18][0],points[18][1]), (255,224,0),5)
# cv2.line(img, (points[21][0],points[21][1]), (points[21][0],points[21][1]), (255,224,0),5)
# cv2.line(img, (points[22][0],points[22][1]), (points[22][0],points[22][1]), (255,224,0),5)
# cv2.line(img, (points[25][0],points[25][1]), (points[25][0],points[25][1]), (255,224,0),5)
#
# cv2.line(img, (points[30][0],points[30][1]), (points[30][0],points[30][1]), (255,0,0),5)
# cv2.line(img, (points[40][0],points[40][1]), (points[40][0],points[40][1]), (255,0,0),5)
#
# cv2.line(img, (points[54][0],points[54][1]), (points[54][0],points[54][1]), (0,255,0),5)
# cv2.line(img, (points[56][0],points[56][1]), (points[56][0],points[56][1]), (0,255,0),5)
# cv2.line(img, (points[58][0],points[58][1]), (points[58][0],points[58][1]), (0,255,0),5)
#
# cv2.line(img, (points[59][0],points[59][1]), (points[59][0],points[59][1]), (0,0,255),5)
# cv2.line(img, (points[65][0],points[65][1]), (points[65][0],points[65][1]), (0,0,255),5)
# Difference in landmark points
difference_array = []
difference_array2 = []
symmetry_point1 = np.asarray([(x_symmetry - 200 * vx_symmetry), (y_symmetry - 200 * vy_symmetry)])
symmetry_point2 = np.asarray([(x_symmetry + 200 * vx_symmetry), (y_symmetry + 200 * vy_symmetry)])
# pdb.set_trace()
Dist_18 = DistancePointLine(points[18][0],points[18][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
Dist_25 = DistancePointLine(points[25][0],points[25][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
difference_array.append((Dist_18-Dist_25)/Dist_18)
Dist_21 = DistancePointLine(points[21][0],points[21][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
Dist_22 = DistancePointLine(points[22][0],points[22][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
difference_array.append((Dist_21-Dist_22)/Dist_21)
Dist_30 = DistancePointLine(points[30][0],points[30][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
Dist_40 = DistancePointLine(points[40][0],points[40][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
difference_array.append((Dist_30-Dist_40)/Dist_30)
Dist_58 = DistancePointLine(points[58][0],points[58][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
Dist_54 = DistancePointLine(points[54][0],points[54][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
difference_array.append((Dist_58-Dist_54)/Dist_58)
Dist_59 = DistancePointLine(points[59][0],points[59][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
Dist_65 = DistancePointLine(points[65][0],points[65][1], symmetry_point1[0], symmetry_point1[1], symmetry_point2[0], symmetry_point2[1])
difference_array.append((Dist_59-Dist_65)/Dist_59)
# Difference in perpendicular projection on the symmetry line
temp1 = []
temp2 = []
print "Line 265 symmetryCal"
perpendicular_vectors = Perpendicular([vx_symmetry, vy_symmetry])
temp1 = np.asanyarray([(points[18][0] - 200 * perpendicular_vectors[0]), (points[18][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[18][0] + 200 * perpendicular_vectors[0]), (points[18][1] + 200 * perpendicular_vectors[1])])
point1 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
temp1 = np.asarray([(points[25][0] - 200 * perpendicular_vectors[0]), (points[25][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[25][0] + 200 * perpendicular_vectors[0]), (points[25][1] + 200 * perpendicular_vectors[1])])
point2 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
difference_array2.append(lineMagnitude(point1[0], point1[1], point2[0], point2[1]))
temp1 = np.asarray([(points[21][0] - 200 * perpendicular_vectors[0]), (points[21][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[21][0] + 200 * perpendicular_vectors[0]), (points[21][1] + 200 * perpendicular_vectors[1])])
point1 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
temp1 = np.asarray([(points[22][0] - 200 * perpendicular_vectors[0]), (points[22][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[22][0] + 200 * perpendicular_vectors[0]), (points[22][1] + 200 * perpendicular_vectors[1])])
point2 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
difference_array2.append(lineMagnitude(point1[0], point1[1], point2[0], point2[1]))
temp1 = np.asarray([(points[30][0] - 200 * perpendicular_vectors[0]), (points[30][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[30][0] + 200 * perpendicular_vectors[0]), (points[30][1] + 200 * perpendicular_vectors[1])])
point1 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
temp1 = np.asarray([(points[40][0] - 200 * perpendicular_vectors[0]), (points[40][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[40][0] + 200 * perpendicular_vectors[0]), (points[40][1] + 200 * perpendicular_vectors[1])])
point2 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
difference_array2.append(lineMagnitude(point1[0], point1[1], point2[0], point2[1]))
temp1 = np.asarray([(points[58][0] - 200 * perpendicular_vectors[0]), (points[58][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[58][0] + 200 * perpendicular_vectors[0]), (points[58][1] + 200 * perpendicular_vectors[1])])
point1 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
temp1 = np.asarray([(points[54][0] - 200 * perpendicular_vectors[0]), (points[54][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[54][0] + 200 * perpendicular_vectors[0]), (points[54][1] + 200 * perpendicular_vectors[1])])
point2 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
difference_array2.append(lineMagnitude(point1[0], point1[1], point2[0], point2[1]))
temp1 = np.asarray([(points[59][0] - 200 * perpendicular_vectors[0]), (points[59][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[59][0] + 200 * perpendicular_vectors[0]), (points[59][1] + 200 * perpendicular_vectors[1])])
point1 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
temp1 = np.asarray([(points[65][0] - 200 * perpendicular_vectors[0]), (points[65][1] - 200 * perpendicular_vectors[1])])
temp2 = np.asarray([(points[65][0] + 200 * perpendicular_vectors[0]), (points[65][1] + 200 * perpendicular_vectors[1])])
point2 = LineSegmentIntersection(temp1, temp2, symmetry_point1, symmetry_point2)
difference_array2.append(lineMagnitude(point1[0], point1[1], point2[0], point2[1]))
return difference_array, difference_array2
# 18(viewer left), 25(right) outer points of eyebrows
# 21(left), 22(right) inner points of eyebrows
# 30(left), 40(right) eye edge detection(towards nose)
# 58(left), 54(right) edge points of nose
# 59(left), 65(right) lip edge points
# for i in range(0, (points_count / 2) - 1):
# # The above loop will run if eye_line_point1[1]< eye_line_point2[1]
# if (eye_line_point1[1] < eye_line_point2[1]) or (eye_line_point1[1] == eye_line_point2[1]):
# if face_boundary_points[1, i] < eye_line_point1[1]:
# print 'pass ', i
# pass
# elif (face_boundary_points[1, i] < symmetry_perpendicular_intersection[1]) and (
# eye_line_point1[1] < face_boundary_points[1, i]):
# xf = find_x(face_boundary_points[1, i], eye_line_point1, eye_line_point2)
# sum_left = sum_left + np.sum(input_img[face_boundary_points[0, i]:xf, face_boundary_points[1, i]])
# print 'cond 1'
# elif ((symmetry_perpendicular_intersection[1] < face_boundary_points[1, i]) and (
# face_boundary_points[1, i] < eye_line_point2[1])):
# xf = find_x(face_boundary_points[1, i], eye_line_point1, eye_line_point2)
# xg = find_x(face_boundary_points[1, i], symmetry_point1, symmetry_point2)
# sum_left = sum_left + np.sum(input_img[face_boundary_points[0, i]:xg, face_boundary_points[1, i]])
# sum_right = sum_right + np.sum(input_img[xg:xf, face_boundary_points[1, i]])
# print 'cond 2'
#
# elif face_boundary_points[1, i] > eye_line_point2[1]:
# xg = find_x(face_boundary_points[1, i], symmetry_point1, symmetry_point2)
# sum_left = sum_left + np.sum(input_img[face_boundary_points[0, i]:xg, face_boundary_points[1, i]])
# sum_right = sum_right + np.sum(input_img[xg:right_boundary[0, i], face_boundary_points[1, i]])
# # sum_right = sum_right + np.sum(input_img[xg:face_boundary_points[1,i], face_boundary_points[1,i]])
# print 'cond 3'
# # total_sum = sum_left + sum_right
# # # pdb.set_trace()set_trace
# # left_percentage = (sum_left/total_sum)*100
# # right_percentage = (sum_right/total_sum)*100
#
#
#
# # For the case where eye_line_point1[1]>=eye_line_point2[1]:
# else:
# print 'Case 2 - left tilted face'
# if face_boundary_points[1, i] < eye_line_point2[1]:
# pass
# elif (face_boundary_points[1, i] < symmetry_perpendicular_intersection[1]) and (
# eye_line_point2[1] < face_boundary_points[1, i]):
# xf = find_x(face_boundary_points[1, i], eye_line_point1, eye_line_point2)
# sum_right = sum_right + np.sum(input_img[xf:right_boundary[0, i], face_boundary_points[1, i]])
# #
# # xg = find_x(face_boundary_points[1], symmetry_point1, symmetry_point2)
# # sum_left = sum_left + np.sum(input_img[face_boundary_points[0,i]:xg, face_boundary_points[1,i]])
# # sum_right = sum_right + np.sum(input_img[xg:xf, face_boundary_points[1,i]])
# elif ((symmetry_perpendicular_intersection[1] < face_boundary_points[1, i]) and (
# face_boundary_points[1, i] < eye_line_point1[1])):
# xf = find_x(face_boundary_points[1, i], eye_line_point1, eye_line_point2)
# xg = find_x(face_boundary_points[1, i], symmetry_point1, symmetry_point2)
# sum_left = sum_left + np.sum(input_img[xf:xg, face_boundary_points[1, i]])
# sum_right = sum_right + np.sum(input_img[xg:right_boundary[0, i], face_boundary_points[1, i]])
# elif face_boundary_points[1, i] > eye_line_point1[1]:
# xg = find_x(face_boundary_points[1, i], symmetry_point1, symmetry_point2)
# # pdb.set_trace()
# print "face_boundary_points[0,i] = ", face_boundary_points[0, i]
# print "xg = ", xg
# sum_left = sum_left + np.sum(input_img[face_boundary_points[0, i]:xg, face_boundary_points[1, i]])
# sum_right = sum_right + np.sum(input_img[xg:right_boundary[0, i], face_boundary_points[1, i]])
# total_sum = sum_left + sum_right
# # pdb.set_trace()
# left_percentage = (sum_left / total_sum) * 100
# right_percentage = (sum_right/total_sum) * 100
# return left_percentage, right_percentage