FrontDetection/MyTransformations.py at master · stnie/FrontDetection · GitHub

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
import random
import numpy as np


class RandomHorizontalFlip(object):
    def __init__(self, p=0.5):
        super().__init__()
        self.p = p

    def __call__(self, img):

        # THIS FLIPS 0 1 2 to  3 4 5
        #            3 4 5     0 1 2
        prob = random.random()
        if prob < self.p:
            return np.flip(img,0)
        return img

    def __repr__(self):
        return self.__class__.__name__ + '(p={})'.format(self.p)

class RandomVerticalFlip(object):
    def __init__(self, p=0.5):
        super().__init__()
        self.p = p

    def __call__(self, img):

        prob = random.random()
        if prob < self.p:
            return np.flip(img,1)
        return img

    def __repr__(self):
        return self.__class__.__name__ + '(p={})'.format(self.p)


class RandomHorizontalCoordsFlip(object):

    def __init__(self, size, p=0.5):
        super().__init__()
        self.p = p
        self.size = size

    def __call__(self, coords):

        # THIS FLIPS 0 1 2 to  3 4 5
        #            3 4 5     0 1 2
        # Assumption Sizes and coords are always given in
        # lat lon
        # above examples has len(lat) = self.size[0] = 2, len(lon) = self.size[1] = 3
        prob = random.random()
        if prob < self.p:
            # flip all horizontal coords
            for unit in range(len(coords)):
                coords[unit][:,0] = self.size[0]-1-coords[unit][:,0]
        return coords

    def __repr__(self):
        return self.__class__.__name__ + '(p={})'.format(self.p)


class RandomVerticalCoordsFlip(object):


    def __init__(self, size, p=0.5):
        super().__init__()
        self.p = p
        self.size = size

    def __call__(self, coords):

        prob = random.random()
        if prob < self.p:
            for unit in range(len(coords)):
                coords[unit][:,1] = self.size[1]-1-coords[unit][:,1]
        return coords

    def __repr__(self):
        return self.__class__.__name__ + '(p={})'.format(self.p)

class RandomTranspose(object):


    def __init__(self, p=0.5):
        super().__init__()
        self.p = p

    def __call__(self, img):

        if random.random() < self.p:
            return np.transpose(img)
        return img

    def __repr__(self):
        return self.__class__.__name__ + '(p={})'.format(self.p)