-
Notifications
You must be signed in to change notification settings - Fork 0
/
kmCues.py
153 lines (115 loc) · 7.8 KB
/
kmCues.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
##############################################################################
# Copyright (c) 2013, Konstantinos Makantasis
# All rights reserved.
#
# Distributed under the terms of the BSD Simplified License
#
##############################################################################
import cv2
import numpy as np
import kmLowLevelFeatures
import kmBlockDivision
def kmPyramidFeatures(frame):
pyr1 = frame
edgePyr1 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindEdges(pyr1, 25, 3, 3), 4)
laplacianPyr1 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindFrequency(pyr1), 4)
linesPyr1 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindLines(pyr1), 4)
colorPyr1 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindColor(pyr1), 4)
entropyPyr1 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindEntropy(pyr1), 4)
pyr2 = cv2.pyrDown(pyr1)
edgePyr2 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindEdges(pyr2, 25, 3, 3), 4)
laplacianPyr2 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindFrequency(pyr2), 4)
linesPyr2 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindLines(pyr2), 4)
colorPyr2 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindColor(pyr2), 4)
entropyPyr2 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindEntropy(pyr2), 4)
pyr3 = cv2.pyrDown(pyr2)
edgePyr3 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindEdges(pyr3, 25, 3, 3), 4)
laplacianPyr3 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindFrequency(pyr3), 4)
linesPyr3 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindLines(pyr3), 4)
colorPyr3 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindColor(pyr3), 4)
entropyPyr3 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindEntropy(pyr3), 4)
pyr4 = cv2.pyrDown(pyr3)
edgePyr4 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindEdges(pyr4, 25, 3, 3), 4)
laplacianPyr4 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindFrequency(pyr4), 4)
linesPyr4 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindLines(pyr4), 4)
colorPyr4 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindColor(pyr4), 4)
entropyPyr4 = kmBlockDivision.kmBlockDivision(kmLowLevelFeatures.kmFindEntropy(pyr4), 4)
edgePyrs = [edgePyr1] + [edgePyr2] + [edgePyr3] + [edgePyr4]
laplacianPyrs = [laplacianPyr1] + [laplacianPyr2] + [laplacianPyr3] + [laplacianPyr4]
linesPyrs = [linesPyr1] + [linesPyr2] + [linesPyr3] + [linesPyr4]
colorPyrs = [colorPyr1] + [colorPyr2] + [colorPyr3] + [colorPyr4]
entropyPyrs = [entropyPyr1] + [entropyPyr2] + [entropyPyr3] + [entropyPyr4]
return edgePyrs, laplacianPyrs, linesPyrs, colorPyrs, entropyPyrs
def kmPyramidSummation(pyr1, pyr2, pyr3, pyr4):
pyr2R = cv2.resize(pyr2, (pyr1.shape[1], pyr1.shape[0]), interpolation=cv2.INTER_LINEAR)
pyr3R = cv2.resize(pyr3, (pyr1.shape[1], pyr1.shape[0]), interpolation=cv2.INTER_LINEAR)
pyr4R = cv2.resize(pyr4, (pyr1.shape[1], pyr1.shape[0]), interpolation=cv2.INTER_LINEAR)
tempPyr = cv2.addWeighted(pyr4R, 0.25, pyr3R, 0.25, 0.0)
tempPyr = cv2.addWeighted(pyr2R, 0.25, tempPyr, 1.0, 0.0)
tempPyr = cv2.addWeighted(pyr1, 0.25, tempPyr, 1.0, 0.0)
return tempPyr.astype(np.uint8)
def kmLocalCues(edgePyrs, laplacianPyrs, linesPyrs, colorPyrs, entropyPyrs):
edgeLocal = kmPyramidSummation(edgePyrs[0], edgePyrs[1], edgePyrs[2], edgePyrs[3])
laplacianLocal = kmPyramidSummation(laplacianPyrs[0], laplacianPyrs[1], laplacianPyrs[2], laplacianPyrs[3])
linesLocal = kmPyramidSummation(linesPyrs[0], linesPyrs[1], linesPyrs[2], linesPyrs[3])
colorLocal = kmPyramidSummation(colorPyrs[0], colorPyrs[1], colorPyrs[2], colorPyrs[3])
entropyLocal = kmPyramidSummation(entropyPyrs[0], entropyPyrs[1], entropyPyrs[2], entropyPyrs[3])
return edgeLocal, laplacianLocal, linesLocal, colorLocal, entropyLocal
def kmGlobalCues(edgePyrs, laplacianPyrs, linesPyrs, colorPyrs, entropyPyrs):
edgeGlobal = kmPyramidSummation(kmGMC(edgePyrs[0]), kmGMC(edgePyrs[1]), kmGMC(edgePyrs[2]), kmGMC(edgePyrs[3]))
laplacianGlobal = kmPyramidSummation(kmGMC(laplacianPyrs[0]), kmGMC(laplacianPyrs[1]), kmGMC(laplacianPyrs[2]), kmGMC(laplacianPyrs[3]))
linesGlobal = kmPyramidSummation(kmGMC(linesPyrs[0]), kmGMC(linesPyrs[1]), kmGMC(linesPyrs[2]), kmGMC(linesPyrs[3]))
colorGlobal = kmPyramidSummation(kmGMC(colorPyrs[0]), kmGMC(colorPyrs[1]), kmGMC(colorPyrs[2]), kmGMC(colorPyrs[3]))
entropyGlobal = kmPyramidSummation(kmGMC(entropyPyrs[0]), kmGMC(entropyPyrs[1]), kmGMC(entropyPyrs[2]), kmGMC(entropyPyrs[3]))
return edgeGlobal, laplacianGlobal, linesGlobal, colorGlobal, entropyGlobal
def kmGMC(featureImg):
globalImg = np.zeros((featureImg.shape[0], featureImg.shape[1]))
for i in range(featureImg.shape[0]):
for j in range(featureImg.shape[1]):
pixelVal = featureImg[i,j]
temp = np.absolute(featureImg.astype(np.int) - pixelVal)
globalImg[i,j] = temp.mean()
globalImg = (globalImg / globalImg.max()) * 255
return globalImg
def kmCSCues(edgePyrs, laplacianPyrs, linesPyrs, colorPyrs, entropyPyrs):
edgeCS = kmPyramidSummation(kmCSMC(edgePyrs[0]), kmCSMC(edgePyrs[1]), kmCSMC(edgePyrs[2]), kmCSMC(edgePyrs[3]))
laplacianCS = kmPyramidSummation(kmCSMC(laplacianPyrs[0]), kmCSMC(laplacianPyrs[1]), kmCSMC(laplacianPyrs[2]), kmCSMC(laplacianPyrs[3]))
linesCS = kmPyramidSummation(kmCSMC(linesPyrs[0]), kmCSMC(linesPyrs[1]), kmCSMC(linesPyrs[2]), kmCSMC(linesPyrs[3]))
colorCS = kmPyramidSummation(kmCSMC(colorPyrs[0]), kmCSMC(colorPyrs[1]), kmCSMC(colorPyrs[2]), kmCSMC(colorPyrs[3]))
entropyCS = kmPyramidSummation(kmCSMC(entropyPyrs[0]), kmCSMC(entropyPyrs[1]), kmCSMC(entropyPyrs[2]), kmCSMC(entropyPyrs[3]))
return edgeCS, laplacianCS, linesCS, colorCS, entropyCS
def kmCSMC(featureImg):
csImg = np.zeros((featureImg.shape[0], featureImg.shape[1]))
x = featureImg.shape[0]
y = featureImg.shape[1]
windowSizeX = 0
windowSizeY = 0
for i in range(featureImg.shape[0]):
if i < (x - i):
windowSizeX = i
else:
windowSizeX = x - i
for j in range(featureImg.shape[1]):
if j < (y - j):
windowSizeY = j
else:
windowSizeY = y - j
pixelVal = featureImg[i,j]
if i < 2 or j < 2 or i > x-2 or j > y-2:
temp = np.absolute(featureImg.astype(np.int) - pixelVal)
csImg[i,j] = temp.mean()
else:
tempImg = featureImg[i-windowSizeX:i+windowSizeX, j-windowSizeY:j+windowSizeY].astype(np.int)
temp = np.absolute(tempImg - pixelVal)
csImg[i,j] = temp.mean()
csImg = (csImg / csImg.max()) * 255
return csImg
def kmAllCues(frame):
edgePyrs, laplacianPyrs, linesPyrs, colorPyrs, entropyPyrs = kmPyramidFeatures(frame)
edgeLocal, laplacianLocal, linesLocal, colorLocal, entropyLocal = kmLocalCues(edgePyrs, laplacianPyrs, linesPyrs, colorPyrs, entropyPyrs)
localCues = [edgeLocal] + [laplacianLocal] + [linesLocal] + [colorLocal] +[entropyLocal]
edgeGlobal, laplacianGlobal, linesGlobal, colorGlobal, entropyGlobal = kmGlobalCues(edgePyrs, laplacianPyrs, linesPyrs, colorPyrs, entropyPyrs)
globalCues = [edgeGlobal] + [laplacianGlobal] + [linesGlobal] + [colorGlobal] +[entropyGlobal]
edgeCS, laplacianCS, linesCS, colorCS, entropyCS = kmCSCues(edgePyrs, laplacianPyrs, linesPyrs, colorPyrs, entropyPyrs)
csCues = [edgeCS] + [laplacianCS] + [linesCS] + [colorCS] +[entropyCS]
return localCues, globalCues, csCues