-
Notifications
You must be signed in to change notification settings - Fork 4
/
Kernel_Operations.py
107 lines (90 loc) · 3.41 KB
/
Kernel_Operations.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
import numpy as np
import random
class kernelOperations:
def __init__(self):
self.randomInit = 0
def rgb2gray(self, rgb):
r, g, b = rgb[:,:,0], rgb[:,:,1], rgb[:,:,2]
gray = 0.2989 * r + 0.5870 * g + 0.1140 * b
return gray
def inseperableKernel(self, im, kernel):
kernel = np.flipud(np.fliplr(kernel))
im_arr = np.array(im)
# im_arr = self.rgb2gray(im_arr)
n = kernel.shape[0]
midPt = int(np.floor(n/2))
divisionFactor = np.sum(kernel)
if divisionFactor==0:
divisionFactor = n*n
arr = np.pad(im_arr, midPt, mode='constant')
# print(arr)
h, w = arr.shape
# kernel = np.random.uniform(0,1,n*n).reshape(n,n)
outImg = np.random.randn(im_arr.shape[0], im_arr.shape[1])
for i in range(0,h-n+1):
for j in range(0,w-n+1):
ex = arr[i:i+n,j:j+n]*kernel
sum1 = np.sum(ex)
# outImg[i,j] = ex[midPt,midPt]
outImg[i,j] = int(sum1/divisionFactor)
outImg = outImg.astype(int)
outImg = np.where(outImg>255,255,outImg)
outImg = np.where(outImg<0,0,outImg)
return outImg
def seperableKernel(self, im, hx, hy):
im_arr = np.array(im)
n = hy.shape[1]
midPt = int(np.floor(n/2))
arr = np.pad(im_arr, midPt, mode='constant')
h, w = arr.shape
tempImg = np.random.randn(h-(2*midPt), w)
# print(tempImg.shape)
for i in range(h-n+1):
for j in range(w):
temp = arr[i:i+n,j].reshape(n,1) * hx.T
# tempImg[i,j] = int((np.sum(np.abs(temp)))/(divisionFactor))
tempImg[i,j] = np.abs(np.sum(temp))
divisionFactor = np.sum(hx.T*hy)
if divisionFactor==0:
divisionFactor = n
outImg = np.random.randn(tempImg.shape[0], tempImg.shape[1]-(2*midPt))
# print(outImg.shape)
for i in range(tempImg.shape[0]):
for j in range(tempImg.shape[1]-n):
temp = tempImg[i,j:j+n] * hy
outImg[i,j] = int((np.abs(np.sum(temp)))/(divisionFactor))
outImg = outImg.astype(int)
outImg = np.where(outImg>255,255,outImg)
outImg = np.where(outImg<0,0,outImg)
return outImg
def inseperableSobelX(self):
arr = np.array([[1,2,1],
[0,0,0],
[-1,-2,-1]])
return arr
def inseperableSobelY(self):
arr = np.array([[-1,0,1],
[-2,0,2],
[-1,0,1]])
return arr
def inseperableRandomIntegers(self, kernel_size=3):
arr = np.random.randint(0,3,size=(kernel_size,kernel_size))
arr = np.array(arr)
return arr
def inseperableRandomNumbers(self, kernel_size=3):
arr = np.abs(np.random.randn(kernel_size,kernel_size))
arr = np.array(arr)
return arr
def seperableSobelX(self):
hx = np.array([[1,0,-1]])
hy = np.array([[1,2,1]])
return hx,hy
def seperableSobelY(self):
hx = np.array([[1,2,1]])
hy = np.array([[-1,0,1]])
return hx,hy
def seperableRandomNumbers(self):
scaling1, scaling2 = np.random.randint(1,4),np.random.randint(1,4)
hx = np.abs(np.random.randn(1,3)*scaling1)
hy = np.abs(np.random.randn(1,3)*scaling2)
return hx, hy