Sephira pantie converter Ver.1.0, fixed Eyo

TenteEEEE · Jul 24, 2021 · eac4def · eac4def
1 parent a51ffa1
commit eac4def
Show file tree

Hide file tree

Showing 15 changed files with 143 additions and 2 deletions.
diff --git a/body/body_sephira.png b/body/body_sephira.png
diff --git a/mask/bra_sephira.png b/mask/bra_sephira.png
diff --git a/mask/bra_sephira_center.png b/mask/bra_sephira_center.png
diff --git a/mask/mask_sephira.png b/mask/mask_sephira.png
diff --git a/material/bra_sephira_shade.png b/material/bra_sephira_shade.png
diff --git a/material/sephira_bra_ext1.png b/material/sephira_bra_ext1.png
diff --git a/material/sephira_bra_ext2.png b/material/sephira_bra_ext2.png
diff --git a/material/sephira_pantie_ext.png b/material/sephira_pantie_ext.png
diff --git a/material/sephira_ribbon.png b/material/sephira_ribbon.png
diff --git a/material/sephira_ribbon_shade.png b/material/sephira_ribbon_shade.png
diff --git a/readme.md b/readme.md
@@ -102,6 +102,8 @@
 |![test](./sample/wolferia_pantie.png)|![test](./sample/inakokoa_pantie.png)|
 |[先呼巣～ホーニーサキュバス](https://booth.pm/ja/items/2294683)|[イヨ](https://booth.pm/ja/items/2630437)|
 |![test](./sample/succubus_pantie.png)|![test](./sample/eyo_pantie.png)|
+|[セフィラ](https://booth.pm/ja/items/2953001)||
+|![test](./sample/sephira_pantie.png)||
 
 ## スペシャルサンクス
 [Booth:キッシュちゃん](https://mutachannel.booth.pm/items/954376)  
@@ -220,6 +222,8 @@ When you want to know the help, please run `python patch.py -h`
 |![test](./sample/wolferia_pantie.png)|![test](./sample/inakokoa_pantie.png)|
 |[Succubus](https://booth.pm/ja/items/2294683)|[Eyo](https://booth.pm/ja/items/2630437)|
 |![test](./sample/succubus_pantie.png)|![test](./sample/eyo_pantie.png)|
+|[Sephira](https://booth.pm/ja/items/2953001)||
+|![test](./sample/sephira_pantie.png)||
 
 ## Special thanks
 [Quiche model](https://mutachannel.booth.pm/items/954376)  

diff --git a/sample/sephira_pantie.png b/sample/sephira_pantie.png
diff --git a/src/models/eyo.py b/src/models/eyo.py
@@ -10,7 +10,6 @@ class patcher(patcher):
     def __init__(self, body='./body/body_eyo.png', **options):
         super().__init__('イヨ', body=body, pantie_position=[1423, 877], **options)
         self.mask = io.imread('./mask/mask_eyo.png')
-        self.bra_position = [298, 1301]
         try:
             self.use_ribbon_mesh = self.options['use_ribbon_mesh']
         except:
@@ -77,7 +76,7 @@ def gen_bra(self, image):
         front_color, front_shade_color = self.extract_bra_color(pantie)
         center = pantie[20:170, -200:-15, :3][:, ::-1]
         center = resize(center, [3.7, 3.7])
-        bra_center = np.float32(io.imread('./mask/bra_eyo_center.png') / 255)
+        bra_center = np.copy(self.bra_center)
         bra_center[:center.shape[0], 55:55 + center.shape[1], :3] = center * np.float32(bra_center[:center.shape[0], 55:55 + center.shape[1], :3] > 0)
         bra = self.bra[:, :, :3] * front_color
         bra_shade = (self.bra_shade[:, :, -1])[:, :, None] * front_shade_color

diff --git a/src/models/sephira.py b/src/models/sephira.py
@@ -0,0 +1,135 @@
+from PIL import Image, ImageOps
+import numpy as np
+import skimage.io as io
+from src.models.class_patcher import patcher
+from src.utils.imgproc import *
+from skimage.color import rgb2hsv, hsv2rgb
+
+
+class patcher(patcher):
+    def __init__(self, body='./body/body_sephira.png', **options):
+        super().__init__('セフィラ', body=body, pantie_position=[737, 2421], **options)
+        self.mask = io.imread('./mask/mask_sephira.png')
+        try:
+            self.use_ribbon_mesh = self.options['use_ribbon_mesh']
+        except:
+            self.use_ribbon_mesh = self.ask(question='Use ribbon mesh?', default=True)
+
+        self.bra_position = [136, 1583]
+        self.bra = np.float32(io.imread('./mask/bra_sephira.png') / 255)
+        self.bra_center = np.float32(io.imread('./mask/bra_sephira_center.png') / 255)
+        self.bra_shade = np.float32(io.imread('./material/bra_sephira_shade.png') / 255)
+        self.bra_shade_alpha = self.bra_shade[:, :, -1]
+
+        self.pantie_extra_position = [1624, 3255]
+        self.pantie_extra_shade = np.float32(io.imread('./material/sephira_pantie_ext.png') / 255)
+
+        self.bra_extra_position1 = [10, 1786]
+        self.bra_extra_position2 = [10, 2958]
+        self.bra_extra_shade1 = np.float32(io.imread('./material/sephira_bra_ext1.png') / 255)
+        self.bra_extra_shade2 = np.float32(io.imread('./material/sephira_bra_ext2.png') / 255)
+
+        if self.use_ribbon_mesh:
+            self.ribbon_position = [1585, 2602]
+            self.ribbon_shade = np.float32(io.imread('./material/sephira_ribbon_shade.png') / 255)[:, :, -1]
+            self.ribbon = np.float32(io.imread('./material/sephira_ribbon.png') / 255)
+
+    def convert(self, image):
+        pantie = np.array(image)
+        if self.use_ribbon_mesh:
+            pantie = ribbon_inpaint(pantie)
+
+        patch = np.copy(pantie[-140:-5, 546:, :])
+        [pr, pc, d] = patch.shape
+        pantie[120:120 + pr, :pc, :] = patch[::-1, ::-1]
+        front = pantie[:300, :350]
+        back = pantie[:-8, 300:][:, ::-1]
+        back = np.pad(back, [(0, 100), (0, 0), (0, 0)], 'constant')
+
+        arry = np.zeros(49)
+        arry[2:25] = np.sin(np.linspace(0, np.pi, 23)) * -10
+        arrx = (np.sin(np.linspace(-np.pi / 2, np.pi, 49)) + 1) * 11.5 - 35
+        arrx[:25] -= np.sin(np.linspace(0, np.pi / 2, 25)) * 4
+        arrx[-14:] += np.linspace(0, 1, 14)**2 * 20
+        front = affine_transform_by_arr(front, arrx, arry)
+        front = np.uint8(resize(front, [1.87, 1.87]) * 255)[:, 16:]
+        front = np.concatenate((front[:, ::-1], front), axis=1)
+
+        arry = np.linspace(0, 1, 49) * -10
+        arry[7:21] -= np.sin(np.linspace(0, np.pi / 2, 14)) * 20
+        arrx = np.linspace(0, 1, 49)**1.5 * 155 - 155
+        arrx[7:-7] -= np.sin(np.linspace(-np.pi, np.pi, 35)) * 6.5
+        arrx[:21] -= np.sin(np.linspace(0, np.pi / 2, 21)) * 5
+        back = affine_transform_by_arr(back, arrx, arry)
+        back = np.uint8(resize(back, [1.87, 1.87]) * 255)[:-50, 18:]
+        back = np.concatenate((back[:, ::-1], back), axis=1)
+        [fr, fc, d] = front.shape
+        [br, bc, d] = back.shape
+        shiftx = 45
+        shifty = 293
+        pantie = np.zeros((fr + br - shifty + 25, fc, d), dtype=np.uint8)
+        pantie[:fr, :fc] = front
+        pantie[shifty:shifty + br, shiftx:shiftx + bc] = alpha_brend(back, pantie[shifty:shifty + br, shiftx:shiftx + bc], back[:, :, -1] > 0.5)
+        pantie = np.bitwise_and(pantie, self.mask)
+        return Image.fromarray(pantie)
+
+    def pick_color(self, arr):
+        return np.mean(np.mean(arr, axis=0), axis=0)
+
+    def extract_bra_color(self, pantie):
+        front = pantie[20:100, 30:80, :3] / 255.0
+        front_shade = pantie[130:150, 0:40, :3] / 255.0
+        front_color = self.pick_color(front)
+        front_shade_color = self.pick_color(front_shade)
+        return front_color, front_shade_color
+
+    def gen_bra(self, image):
+        pantie = np.array(image)
+        front_color, front_shade_color = self.extract_bra_color(pantie)
+        center = pantie[20:170, -200:-15, :3][:, ::-1]
+        center = resize(center, [4.3, 4.3])
+        bra_center = np.copy(self.bra_center)
+        bra_center[1715:1715 + center.shape[0], 100:100 + center.shape[1], :3] = center * np.float32(bra_center[1715:1715 + center.shape[0], 100:100 + center.shape[1], :3] > 0)
+        bra = self.bra[:, :, :3] * front_color
+        bra_shade = np.float32(self.bra_shade[:, :, -1] > 0)[:, :, None] * front_shade_color
+        bra = alpha_brend(bra_center[:, :, :3], bra[:, :, :3], bra_center[:, :, 0] > 0.1)
+        bra = alpha_brend(bra_shade, bra, self.bra_shade_alpha)
+        bra = np.dstack((bra, self.bra[:, :, 0] > 0.8))
+        return Image.fromarray(np.uint8(np.clip(bra, 0, 1) * 255))
+
+    def gen_extra_using_shademap(self, image, shade):
+        pantie = np.array(image)
+        front_color, front_shade_color = self.extract_bra_color(pantie)
+        dst = np.zeros_like(shade) + 1.
+        dst = dst[:, :, :3] * front_color
+        dst_shade = np.float32(shade[:, :, -1] > 0)[:, :, None] * front_shade_color
+        dst = alpha_brend(dst_shade, dst, shade[:, :, -1])
+        dst = np.dstack((dst, dst[:, :, 0] > 0))
+        return Image.fromarray(np.uint8(np.clip(dst, 0, 1) * 255))
+
+    def gen_ribbon(self, image):
+        pantie = np.array(image)
+        ribbon = pantie[24:32, 15:27, :3] / 255.0
+        ribbon_color = self.pick_color(ribbon)
+        ribbon_shade = pantie[26:30, 12:15, :3] / 255.0
+        ribbon_shade_color = self.pick_color(ribbon_shade)
+        ribbon = self.ribbon[:, :, :3] * ribbon_color
+        ribbon_shade = np.float32(self.ribbon_shade > 0)[:, :, None] * ribbon_shade_color
+        ribbon = alpha_brend(ribbon_shade, ribbon, self.ribbon_shade)
+        ribbon = np.dstack((ribbon, self.ribbon[:, :, -1] > 0.5))
+        return Image.fromarray(np.uint8(ribbon * 255))
+
+    def patch(self, image, transparent=False):
+        pantie = self.convert(image)
+        if transparent:
+            patched = Image.new("RGBA", (4096, 4096))
+        else:
+            patched = self.body.copy()
+        patched = self.paste(patched, pantie, self.pantie_position)
+        patched = self.paste(patched, self.gen_bra(image), self.bra_position)
+        patched = self.paste(patched, self.gen_extra_using_shademap(image, self.pantie_extra_shade), self.pantie_extra_position)
+        patched = self.paste(patched, self.gen_extra_using_shademap(image, self.bra_extra_shade1), self.bra_extra_position1)
+        patched = self.paste(patched, self.gen_extra_using_shademap(image, self.bra_extra_shade2), self.bra_extra_position2)
+        if self.use_ribbon_mesh:
+            patched = self.paste(patched, self.gen_ribbon(image), self.ribbon_position)
+        return patched
diff --git a/webapp.json b/webapp.json
@@ -111,6 +111,9 @@
   "sakurana":[
     "with_bra"
   ],
+  "sephira":[
+    "use_ribbon_mesh"
+  ],
   "shaclo_winter":[
     "with_bra"
   ],