-
Notifications
You must be signed in to change notification settings - Fork 1
/
chain_coding.py
285 lines (242 loc) · 8.96 KB
/
chain_coding.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
import numpy as np
import sys
import os
import cv2
#import matplotlib.pyplot as plt
import math
import sys
def createX(path):
X = []
for file in os.listdir(path):
im = cv2.imread(path + os.sep + file)
gray_image = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)
X.append(gray_image)
return np.array(X)
def findChain(firstPixel, image):
prevPixel = None
currentPixel = None
currentDirection = None
pixels = []
prevPixel = firstPixel
# directions = [(-1, -1), (-1, 0), (-1, 1),
# (0, -1), (0, 1),
# (1, -1), (1, 0), (1, 1)]
directions = [(-1, -1), (-1, 0), (-1, 1), (0, 1),
(1, 1), (1, 0), (1, -1), (0, -1)]
count = 0
for directionIndex in range(len(directions)):
direction = directions[directionIndex]
newPixelRow = prevPixel[0] + direction[0]
newPixelCol = prevPixel[1] + direction[1]
if(image[newPixelRow][newPixelCol] > 0):
currentDirection = directionIndex
break
currentPixel = (newPixelRow, newPixelCol)
pixels.append(currentPixel)
while(currentPixel != firstPixel):
count += 1
startingDirection = (currentDirection + 5) % 8
for directionOffset in range(len(directions)):
directionIndex = (startingDirection + directionOffset) % 8
direction = directions[directionIndex]
newPixelRow = currentPixel[0] + direction[0]
newPixelCol = currentPixel[1] + direction[1]
if(image[newPixelRow][newPixelCol] > 0):
currentPixel = (newPixelRow, newPixelCol)
pixels.append(currentPixel)
currentDirection = directionIndex
break
return pixels
def findCentroid(pixels):
avgX = 0
avgY = 0
for pixel in pixels:
pixelX, pixelY = pixel
avgX += pixelX
avgY += pixelY
return (avgX//len(pixels), avgY//len(pixels))
def findPerimeter(image):
pixelFound = False
for rowIndex in range(len(image)//2, len(image)):
for colIndex in range(len(image[0])):
pixel = image[rowIndex][colIndex]
if(pixel > 0):
firstPixel = (rowIndex, colIndex)
chain = findChain(firstPixel, image)
pixelFound = True
if(pixelFound):
break
if(pixelFound):
break
return chain
def makeNewImage(pixels):
image = [[255] * 1000 for i in range(1000)]
for row in range(1000):
for col in range(1000):
if((row, col) in pixels):
image[row][col] = 0
cv2.imwrite("test.jpg", np.array(image))
img = cv2.imread("test.jpg")
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
cv2.imshow('perimeter', gray)
cv2.waitKey(0)
def distance(x1, y1, x2, y2):
return math.sqrt((x1 - x2)**2 + (y1 - y2)**2)
def findDistances(centroid, perimeter):
centroidX, centroidY = centroid
sample = len(perimeter)//360 + 1
distances = []
for i in range(0, 360):
pixelIndex = (sample*i)%len(perimeter)
pixelX, pixelY = perimeter[pixelIndex]
pixDist = distance(centroidX, centroidY, pixelX, pixelY)
distances.append((pixelIndex, pixDist))
return distances
def getAmplitude(distances):
return max(distances) - min(distances)
def crossConvolution(signal, template):
maximum = None
maxshift = None
correlations = []
correlations = np.correlate(signal, template, "full")
#plt.plot(signal, 'r', template, 'g')
#plt.plot(correlations)
#plt.show()
# for shift in range(0, 360):
# shiftedTemplate = np.roll(template, shift)
# correlation = np.correlate(signal, shiftedTemplate)
# #print(correlation, np.dot(signal, shiftedTemplate))
# correlations.append(correlation[0])
# # if(shift % 60 == 0):
# # plt.plot(signal, 'r', shiftedTemplate, 'g')
# # plt.show()
# if(maximum == None or correlation[0] > maximum):
# maximum = correlation
# maxshift = shift
# print(np.shape(correlations))
# print(maxshift)
# plt.plot(correlations)
# plt.show()
amplitude = getAmplitude(correlations)
return max(abs(correlations))#amplitude#correlation
def makeTemplate(X, index, filename):
showImage = False
blur = cv2.blur(X[index], (5,5))
pixels = findPerimeter(blur)
if (showImage):
cv2.imshow('blurred_image', blur)
cv2.waitKey(0)
centroid = findCentroid(pixels)
distances = findDistances(centroid, pixels)
data = np.array(distances)[:,1]
plt.plot(data)
plt.show()
block_template = np.save(filename + ".npy", data)
def loadTemplates():
spike_template = np.load("spike_template.npy")
block_template = np.load("block3_template.npy")
ball_template = np.load("ball_template.npy")
spike_template = spike_template/(np.linalg.norm(spike_template))
block_template = block_template/(np.linalg.norm(block_template))
ball_template = ball_template/(np.linalg.norm(ball_template))
return spike_template, block_template, ball_template
def classify(data):
spike_template, block_template, ball_template = loadTemplates()
amplitude = getAmplitude(data)
variance = np.var(data)
oldData = copy.deepcopy(data)
ballCorrelation = crossConvolution(data, ball_template)
blockCorrelation = crossConvolution(data, block_template)
spikeCorrelation = crossConvolution(data, spike_template)
maxCorrelation = max(ballCorrelation, blockCorrelation, spikeCorrelation)
if(ballCorrelation == maxCorrelation):
print("This is a ball!!!!")
os.system("python ./braille/n_char.py " + '"ball"')
return "ball"
elif(blockCorrelation == maxCorrelation):
print("This is a block!!!!")
os.system("python ./braille/n_char.py " + '"block"')
return "block"
else:
print("This is a spike!!!!")
os.system("python ./braille/n_char.py " + '"spike"')
return "spike"
def computeData(image):
blur = cv2.blur(image, (5,5))
pixels = findPerimeter(blur)
# cv2.imshow('blurred_image', blur)
# cv2.waitKey(0)
centroid = findCentroid(pixels)
distances = findDistances(centroid, pixels)
data = np.array(distances)[:,1]
data = data/(np.linalg.norm(data))
return data
def main():
print(sys.argv)
# if(sys.argc < 2):
# print("Not enough input arguments!!")
im = cv2.imread(sys.argv[1])
gray_image = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)
data = computeData(gray_image)
return classify(data)
import random, copy
def getAccuracy():
X = createX('segmented_data_train')
dictionary = dict()
spike_template = np.load("spike_template.npy")
block_template = np.load("block3_template.npy")
ball_template = np.load("ball_template.npy")
spike_template = spike_template/(np.linalg.norm(spike_template))
block_template = block_template/(np.linalg.norm(block_template))
ball_template = ball_template/(np.linalg.norm(ball_template))
for index in range(0, 75):
print(index)
try:
blur = cv2.blur(X[index], (5,5))
pixels = findPerimeter(blur)
# cv2.imshow('blurred_image', blur)
# cv2.waitKey(0)
centroid = findCentroid(pixels)
distances = findDistances(centroid, pixels)
data = np.array(distances)[:,1]
data = data/(np.linalg.norm(data))
amplitude = getAmplitude(np.array(distances)[:,1])
variance = np.var(data)
oldData = copy.deepcopy(data)
ballCorrelation = crossConvolution(data, ball_template)
blockCorrelation = crossConvolution(data, block_template)
spikeCorrelation = crossConvolution(data, spike_template)
maxCorrelation = max(ballCorrelation, blockCorrelation, spikeCorrelation)
print("ball", maxCorrelation, ballCorrelation)
print("block", maxCorrelation, blockCorrelation)
print("spike", maxCorrelation, spikeCorrelation)
if(ballCorrelation == maxCorrelation):
print("This is a ball!!!!")
dictionary[index] = "ball"
elif(blockCorrelation == maxCorrelation):
print("This is a block!!!!")
dictionary[index] = "block"
else:
print("This is a spike!!!!")
dictionary[index] = "spike"
except:
continue
print(dictionary)
total = 0
for key in dictionary:
if(key < 25 and dictionary[key] == "ball"):
total += 1
elif(key < 50 and dictionary[key] == "block"):
total += 1
elif(key < 75 and dictionary[key] == "spike"):
total += 1
print('accuracy', total/len(dictionary))
#makeNewImage(pixels)
#cv2.imshow("blah", blur)
import time
start = time.time()
main()
end = time.time()
print(end-start)
# X = createX('segmented_data_train')
# makeTemplate(X, 27, "block3_template")