-
Notifications
You must be signed in to change notification settings - Fork 0
/
siftOnePixel.py
146 lines (126 loc) · 4.07 KB
/
siftOnePixel.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
import cv2 as cv2
import numpy as np
from collections import Counter
import math
import matplotlib.pyplot as plt
#test images
# s='add.png'
s='lenna.jpg'
s='ttt.jpg'
#pixel used for SIFT
pixelX=200
pixelY=200
#functions
def drawContours(img,contours,color):
contours=np.array(contours)
for i in range(contours.shape[0]):
for j in range(contours[i].shape[0]):
for k in range(contours[i][j].shape[0]):
img[contours[i][j][k][1]][contours[i][j][k][0]]=color
def getContours(img,seuil=30):
contours=[]
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
h,w=np.shape(img)
imgContours=np.zeros((h,w),np.double)
imgContoursX=np.zeros((h,w),np.double)
imgContoursY=np.zeros((h,w),np.double)
for i in range(0,h):
for j in range (0,w):
if(j==0 or j==w-1 or i==0 or i==h-1):
imgContoursX[i][j]=0
imgContoursY[i][j]=0
else:
imgContoursX[i][j] = (np.multiply( convX,img[i-1:i+2,j-1:j+2]).sum(axis=1).sum(axis=0))
imgContoursY[i][j] = (np.multiply( convY,img[i-1:i+2,j-1:j+2]).sum(axis=1).sum(axis=0))
a=math.sqrt(math.pow(imgContoursX[i][j],2)+math.pow(imgContoursY[i][j],2))
a=min(a,255)
a=max(a,0)
if(a>seuil):
imgContours[i][j]=a
contours.append([i,j])
return imgContours,contours,imgContoursX,imgContoursY
def getOrientation(img,x,y):
global imgContoursX,imgContoursY
if(x<0 or x>w-1 or y<0 or y>h-1):
d=0
else:
d=math.atan2(imgContoursY[y][x],imgContoursX[y][x])
d+=math.pi
return d
def getBlock(img,x,y):
a=[]
block=np.zeros((16,16),np.double)
for i in range(x-8,x+8):
for j in range(y-8,y+8):
angle=roundAngle(getOrientation(img,i,j))
block[j-y+8][i-x+8]=angle
# block=np.zeros((4,4),np.double)
# for i in range(x-8,x+8):
# for j in range(y-8,y+8):
# xb=(i-x+8)//4
# yb=(j-y+8)//4
# angle=roundAngle(getOrientation(img,i,j))
# block[yb][xb]=max(block[yb][xb],angle)
# a.append(angle)
return block
def anglesArray():
angles = []
i=0
while True:
angles.append(math.pi*i)
i+=1/4
if(i==2):
break
return angles
def anglesStringsArray():
return ["0","π/4","π/2","3π/4","π","5π/4","3π/2","7π/4"]
def roundAngle(angle):
angles = anglesArray()
index=np.argmin(np.abs(np.subtract(angles,angle)))
angles = anglesArray()
a=angles[index]
return a
def roundAngleTitle(angle):
anglesStrings = anglesStringsArray()
angles = anglesArray()
a=np.argmin(np.abs(np.subtract(angles,angle)))
return anglesStrings[a]
def roundAngleIndex(angle):
angles = anglesArray()
a=np.argmin(np.abs(np.subtract(angles,angle)))
return a
def showHist(tempdicArray):
fig, ax = plt.subplots(4,4,figsize=(14,8))
for i in range (0,len(tempdicArray)):
x=(int)(i/4)
ax[x][i%4].bar(list(tempdicArray[i].keys()), tempdicArray[i].values(), color='b')
fig.tight_layout()
plt.show()
#main
img=cv2.imread(s)
h,w,d = np.shape(img)
#convolution matrix
c=1
convX=np.zeros((3,3),np.double)
convX[0,0]=0;convX[0,1]=0;convX[0,2]=0;convX[1,0]=-c;convX[1,1]=0
convX[1,2]= c;convX[2,0]= -0;convX[2,1]=0;convX[2,2]=0
convY=np.zeros((3,3),np.double)
convY[0,0]=-0;convY[0,1]=-c;convY[0,2]=-0;convY[1,0]=0;convY[1,1]=0
convY[1,2]= 0;convY[2,0]= 0;convY[2,1]=c;convY[2,2]=0
#threshold for contours
seuil=30
img,contours,imgContoursX,imgContoursY=getContours(img,seuil)
blocks=getBlock(img,pixelX,pixelY)
dic={}
histogrammes=[]
for block in blocks:
#count orientations for histogramme
array=np.matrix.flatten(block)
count=Counter(array)
for c in count:
dic[roundAngleTitle(c)]=count[c]
histogrammes.append(dic.copy())
dic={}
showHist(histogrammes)
cv2.imshow('image : '+s,img)
cv2.waitKey(0)