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
|
#!/usr/bin/python3
import os, subprocess, cv2
# This is a much more complex script. BY ALL MEANS use base folder
pwd = os.getcwd()
f = open("ATTR.txt", "w")
bf1=[]
subprocess.call("cp ready-units/base/* ready-units/", shell=True)
FILES = list(os.listdir(pwd + "/ready-units"))
for it in FILES:
if not it.endswith("png") and not it.endswith("jpg"):
continue
if it.startswith("sq_"):
continue
name = it.split(".")[0]
path = pwd + "/ready-units/"
sqne = "sq_%s" % it
try:
unitid = int(name)
except:
print("ERROR: %s is not parseable" % it)
continue
print("%s" % it)
"""
# Force to PNG
if it.endswith("jpg"):
# WARNING: DESTRUCTIVE
subprocess.call("convert %s/%s %s/%s.png" % (path, it, path, it), shell=True)
subprocess.call("rm %s/%s" % (path, it), shell=True)
it = "%s.png" % it
"""
# Resizing
img = cv2.imread(path+it)
w, h = img.shape[:2]
if h != 640 or w != 960:
# Resize the image, relying entirely in ImageMagick
# Preserving aspect ratio is turned OFF (Crop would be better)
# WARNING: DESTRUCTIVE
subprocess.call("convert %s/%s -resize 640x960\! %s/%s" % (path, it, path, it), shell=True)
print("[OK] Image resized")
# A square doesn't exists
# Generate one with opencv
if not sqne in FILES:
cPath="lbpcascade_animeface.xml"
iPath="ready-units/%s" % it
cascade=cv2.CascadeClassifier(cPath)
org_img=cv2.imread(iPath)
img=cv2.cvtColor(org_img, cv2.COLOR_BGR2GRAY)
maxi = 3.2
step = 0.1
curr = 1.0
faces = []
# Detect faces
while len(faces) != 1:
if curr == 1.0:
faces = cascade.detectMultiScale(img, scaleFactor=1.01, minNeighbors=5)
else:
faces = cascade.detectMultiScale(img, scaleFactor=curr, minNeighbors=5)
curr+=step
if curr > maxi:
break
if len(faces) != 1:
print("INFO: %s has %d anime faces, no square possible!" % (it, len(faces)))
print("Retrying with human face...")
cPath="lbpcascade_face.xml"
iPath="ready-units/%s" % it
cascade=cv2.CascadeClassifier(cPath)
org_img=cv2.imread(iPath)
img=cv2.cvtColor(org_img, cv2.COLOR_BGR2GRAY)
maxi = 3.2
step = 0.1
curr = 1.0
faces = []
# Detect faces
while len(faces) != 1:
if curr == 1.0:
faces = cascade.detectMultiScale(img, scaleFactor=1.01, minNeighbors=5)
else:
faces = cascade.detectMultiScale(img, scaleFactor=curr, minNeighbors=5)
curr+=step
if curr > maxi:
break
if len(faces) != 1:
print("\033[1mERROR: %s has %d faces, no square possible!\033[0m" % (it, len(faces)))
bf1.append("* ERROR: %s needs a square\n" % it)
continue
face=faces[0]
# Calculate offset
xoff=max(0, 340-face[2])/2
yoff=max(0, 340-face[3])/2
x=max(0, face[0]-xoff)
y=max(0, face[1]-yoff)
subprocess.call("convert %s -crop %dx%d+%d+%d %s/%s" % (iPath, 340, 340, x, y, path, sqne), shell=True)
print("[OK] Success finding face at (%d,%d)" % (x, y))
# Resize the squared image (DESTRUCTIVE)
# FIXME: Add alpha channel
subprocess.call("convert %s/%s -resize 340x340\! -alpha on %s/%s.png" % (path, sqne, path, sqne), shell=True)
# Format the squared image (DESTRUCTIVE)
subprocess.call("convert %s/%s.png mask.png -alpha off -compose CopyOpacity -composite %s/%s.webp" % (path, sqne, path, sqne), shell=True)
sqne+=".webp"
# Make copies
i=0
while i < 3:
subprocess.call("convert %s/%s %s/unit/10%04d%02d.webp" % (path, it, path, unitid, i), shell=True)
subprocess.call("convert %s/%s %s/square/10%04d%02d.webp" % (path, sqne, path, unitid, i), shell=True)
i+=1
# Save to attribution
bf1.append("\t10%04d NAME (AUTHOR)\t(CC BY SA)\t(SOURCE)\n" % (unitid))
# Save attribution file
for it in sorted(bf1):
f.write(it)
f.close()
# Remove the working files
# IF YOU DID NOT USE BASE FOLDER, CONSIDER THE FILES FORSAKEN
subprocess.call("rm ready-units/*.*", shell=True)
|
