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
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
| '''
Date: 2022-06-27 14:57:55
LastEditTime: 2022-06-27 16:06:43
'''
from PIL import Image
from pylab import *
import numpy as np
import cv2
from scipy.spatial import Delaunay
import matplotlib.pyplot as plt
def morphed_im(im1, im2, im1_pts, im2_pts, warp_func, dissolve_func):
"""生成 morphing 图像
Args:
im1 (ndarray): 原图像1的像素数组
im2 (ndarray): 原图像2的像素数组
im1_pts (ndarray): 原图像1上标注的点
im2_pts (ndarray): 原图像2上标注的点
warp_func (float): 向图像2变形(warping)的权重
dissolve_func (float): 向图像2取色的权重
Return:
morphed_im(ndarray): 生成的图像
"""
if (np.size(im1, 2) != 3 or np.size(im2, 2) != 3 or np.size(im1, 1) != np.size(im2, 1) or np.size(im1, 2) != np.size(im2, 2)):
return
xsz = np.size(im1, 0)
ysz = np.size(im1, 1)
pts_mean = (1 - warp_func) * im1_pts + warp_func * im2_pts
tri = Delaunay(pts_mean)
# 显示三角分割
figure()
# 图形1的三角分割
sub1 = subplot(2, 2, 1)
plt.imshow(im1)
print(im1_pts)
sub1.triplot(im1_pts[:, 0], im1_pts[:, 1], tri.simplices.copy())
sub1.plot(im1_pts[:, 0], im1_pts[:, 1], 'o')
# 图形2的三角分割
sub2 = subplot(2, 2, 2)
plt.imshow(im2)
sub2.triplot(im2_pts[:, 0], im2_pts[:, 1], tri.simplices.copy())
sub2.plot(im2_pts[:, 0], im2_pts[:, 1], 'o')
# 平均三角分割的形状图形
sub_mean = subplot(2, 2, 3)
sub_mean.triplot(pts_mean[:, 0], ysz - pts_mean[:, 1], tri.simplices.copy())
sub_mean.plot(pts_mean[:, 0], ysz - pts_mean[:, 1], 'o')
# 关闭三角分割图片
plt.pause(10)
plt.close()
# 获取重心坐标参数
barycentric = zeros((xsz * ysz, 3), dtype=float32)
tris = tri.simplices
xy_mesh = np.array(list((x,y) for x in range(xsz) for y in range(ysz))) # 生成网格
# 网格点对应像素点所在三角形在平均图形坐标点集中的索引
triang = tri.find_simplex([xy_mesh])[0]
for i in arange(0, xsz * ysz):
xp,yp = xy_mesh[i]
trindex = tri.find_simplex([xp, yp])
if trindex == -1:
# 当前像素点不在三角分割内时不计算入生成图形内
barycentric[i, :] = [float('nan'), float('nan'), float('nan')]
continue
ax = pts_mean[tris[trindex, 0], 0]
bx = pts_mean[tris[trindex, 1], 0]
cx = pts_mean[tris[trindex, 2], 0]
ay = pts_mean[tris[trindex, 0], 1]
by = pts_mean[tris[trindex, 1], 1]
cy = pts_mean[tris[trindex, 2], 1]
A = np.mat([[ax, bx, cx], [ay, by, cy], [1, 1, 1]])
b = np.mat([[xp], [yp], [1]])
mul = inv(A) * b
barycentric[i] = [mul[0,0], mul[1,0], mul[2,0]]
# 用重心坐标计算生成图像每个像素对应的原图像像素的坐标
im1_crsp = zeros((xsz * ysz, 3), dtype=float32)
for i in arange(0, xsz * ysz):
trindex = triang[i]
if trindex == -1:
im1_crsp[i] = [255,255,255]
continue
ax = im1_pts[tris[trindex, 0], 0]
bx = im1_pts[tris[trindex, 1], 0]
cx = im1_pts[tris[trindex, 2], 0]
ay = im1_pts[tris[trindex, 0], 1]
by = im1_pts[tris[trindex, 1], 1]
cy = im1_pts[tris[trindex, 2], 1]
A = np.mat([[ax, bx, cx], [ay, by, cy], [1, 1, 1]])
X = A * np.mat(barycentric[i]).reshape(3, 1)
X = uint8(around(X))
im1_crsp[i] = im1[X[1, 0], X[0, 0]] # 这个地方xy是反着的，因为不知名原因im1的像素点的索引和颜色值是y=x的直线对称的
# 用重心坐标计算生成图像每个像素对应的原图像像素的坐标
im2_crsp = zeros((xsz * ysz, 3), dtype=float32)
for i in arange(0, xsz * ysz):
trindex = triang[i]
if trindex == -1:
im2_crsp[i] = [255,255,255]
continue
ax = im2_pts[tris[trindex, 0], 0]
bx = im2_pts[tris[trindex, 1], 0]
cx = im2_pts[tris[trindex, 2], 0]
ay = im2_pts[tris[trindex, 0], 1]
by = im2_pts[tris[trindex, 1], 1]
cy = im2_pts[tris[trindex, 2], 1]
A = [[ax, bx, cx], [ay, by, cy], [1, 1, 1]]
X = A * np.mat(barycentric[i]).reshape(3, 1)
X = uint8(around(X))
im2_crsp[i] = im2[X[1,0], X[0,0]]
print("finish warping")
morphed_im = zeros((xsz * ysz, 3), dtype=float32).reshape(xsz, ysz, 3)
morphed_im1 = zeros((xsz * ysz, 3), dtype=uint8).reshape(xsz, ysz, 3)
morphed_im2 = zeros((xsz * ysz, 3), dtype=uint8).reshape(xsz, ysz, 3)
# 将前面计算好的平均图像的像素值填充到数组中
for i in arange(0, xsz * ysz):
xp,yp = xy_mesh[i]
morphed_im1[yp, xp] = im1_crsp[i]
morphed_im2[yp, xp] = im2_crsp[i]
morphed_im = (1 - dissolve_func) * morphed_im1 + dissolve_func * morphed_im2
morphed_im = uint8(around(morphed_im))
print("finish morphing")
figure()
sub1 = subplot(2, 2, 1)
plt.imshow(morphed_im1)
sub2 = subplot(2, 2, 2)
plt.imshow(morphed_im2)
sub2 = subplot(2, 2, 3)
plt.imshow(morphed_im)
sub_mean = subplot(2, 2, 4)
plt.imshow(morphed_im)
plt.pause(10)
plt.close()
return morphed_im
if __name__ == '__main__':
# 展示两张图片
im1 = array(Image.open('p1.jpg'))
im2 = array(Image.open('p2.jpg'))
p1 = np.array(np.loadtxt(fname="p1.csv", dtype=np.float32, delimiter=","))
p2 = np.array(np.loadtxt(fname="p2.csv", dtype=np.float32, delimiter=","))
im = morphed_im(im1, im2, p1, p2, 0.5, 0.5)
cv2.cvtColor(im, cv2.COLOR_BGR2RGB, im); # 转换色域
cv2.imwrite('./res.jpg', im)
|
