FaceMorphing/align_image_code.py at master · kevin-lii/FaceMorphing · 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
import math
import numpy as np
import matplotlib.pyplot as plt
import skimage.transform as sktr


def get_points(im1, im2):
    print('Please select 2 points in each image for alignment.')
    plt.imshow(im1)
    p1, p2 = plt.ginput(2)
    plt.close()
    plt.imshow(im2)
    p3, p4 = plt.ginput(2)
    plt.close()
    return (p1, p2, p3, p4)

def recenter(im, r, c):
    R, C, _ = im.shape
    rpad = (int) (np.abs(2*r+1 - R))
    cpad = (int) (np.abs(2*c+1 - C))
    return np.pad(
        im, [(0 if r > (R-1)/2 else rpad, 0 if r < (R-1)/2 else rpad),
             (0 if c > (C-1)/2 else cpad, 0 if c < (C-1)/2 else cpad),
             (0, 0)], 'constant')

def find_centers(p1, p2):
    cx = np.round(np.mean([p1[0], p2[0]]))
    cy = np.round(np.mean([p1[1], p2[1]]))
    return cx, cy

def align_image_centers(im1, im2, pts):
    p1, p2, p3, p4 = pts
    h1, w1, b1 = im1.shape
    h2, w2, b2 = im2.shape

    cx1, cy1 = find_centers(p1, p2)
    cx2, cy2 = find_centers(p3, p4)

    im1 = recenter(im1, cy1, cx1)
    im2 = recenter(im2, cy2, cx2)
    return im1, im2

def rescale_images(im1, im2, pts):
    p1, p2, p3, p4 = pts
    len1 = np.sqrt((p2[1] - p1[1])**2 + (p2[0] - p1[0])**2)
    len2 = np.sqrt((p4[1] - p3[1])**2 + (p4[0] - p3[0])**2)
    dscale = len2/len1
    if dscale < 1:
        im1 = sktr.rescale(im1, dscale, multichannel = True)
    else:
        im2 = sktr.rescale(im2, 1./dscale, multichannel = True)
    return im1, im2

def rotate_im1(im1, im2, pts):
    p1, p2, p3, p4 = pts
    theta1 = math.atan2(-(p2[1] - p1[1]), (p2[0] - p1[0]))
    theta2 = math.atan2(-(p4[1] - p3[1]), (p4[0] - p3[0]))
    dtheta = theta2 - theta1
    im1 = sktr.rotate(im1, dtheta*180/np.pi)
    return im1, dtheta

def match_img_size(im1, im2):
    # Make images the same size
    h1, w1, c1 = im1.shape
    h2, w2, c2 = im2.shape
    if h1 < h2:
        im2 = im2[int(np.floor((h2-h1)/2.)) : -int(np.ceil((h2-h1)/2.)), :, :]
    elif h1 > h2:
        im1 = im1[int(np.floor((h1-h2)/2.)) : -int(np.ceil((h1-h2)/2.)), :, :]
    if w1 < w2:
        im2 = im2[:, int(np.floor((w2-w1)/2.)) : -int(np.ceil((w2-w1)/2.)), :]
    elif w1 > w2:
        im1 = im1[:, int(np.floor((w1-w2)/2.)) : -int(np.ceil((w1-w2)/2.)), :]
    assert im1.shape == im2.shape
    return im1, im2

def align_images(im1, im2):
    pts = get_points(im1, im2)
    im1, im2 = align_image_centers(im1, im2, pts)
    im1, im2 = rescale_images(im1, im2, pts)
    im1, angle = rotate_im1(im1, im2, pts)
    im1, im2 = match_img_size(im1, im2)
    return im1, im2


if __name__ == "__main__":
    # 1. load the image
    # 2. align the two images by calling align_images
    # Now you are ready to write your own code for creating hybrid images!
    pass