Spaces:
Running
Running
Upload 4 files
Browse files- api_server.py +525 -0
- scalingtestupdated.py +184 -0
- u2netp.pth +3 -0
- u2netp.py +525 -0
api_server.py
ADDED
@@ -0,0 +1,525 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# from fastapi import FastAPI, HTTPException, UploadFile, File, Form
|
2 |
+
# from pydantic import BaseModel
|
3 |
+
# import numpy as np
|
4 |
+
# from PIL import Image
|
5 |
+
# import io, uuid, os, shutil, timeit
|
6 |
+
# from datetime import datetime
|
7 |
+
# from fastapi.staticfiles import StaticFiles
|
8 |
+
# from fastapi.middleware.cors import CORSMiddleware
|
9 |
+
|
10 |
+
# # import your three wrappers
|
11 |
+
# from app import predict_simple, predict_middle, predict_full
|
12 |
+
|
13 |
+
# app = FastAPI()
|
14 |
+
|
15 |
+
# # allow CORS if needed
|
16 |
+
# app.add_middleware(
|
17 |
+
# CORSMiddleware,
|
18 |
+
# allow_origins=["*"],
|
19 |
+
# allow_methods=["*"],
|
20 |
+
# allow_headers=["*"],
|
21 |
+
# )
|
22 |
+
|
23 |
+
# BASE_URL = "https://snapanddtraceapp-988917236820.us-central1.run.app"
|
24 |
+
# OUTPUT_DIR = os.path.abspath("./outputs")
|
25 |
+
# os.makedirs(OUTPUT_DIR, exist_ok=True)
|
26 |
+
# app.mount("/outputs", StaticFiles(directory=OUTPUT_DIR), name="outputs")
|
27 |
+
|
28 |
+
# UPDATES_DIR = os.path.abspath("./updates")
|
29 |
+
# os.makedirs(UPDATES_DIR, exist_ok=True)
|
30 |
+
# app.mount("/updates", StaticFiles(directory=UPDATES_DIR), name="updates")
|
31 |
+
|
32 |
+
|
33 |
+
# def save_and_build_urls(
|
34 |
+
# session_id: str,
|
35 |
+
# output_image: np.ndarray,
|
36 |
+
# outlines: np.ndarray,
|
37 |
+
# dxf_path: str,
|
38 |
+
# mask: np.ndarray
|
39 |
+
# ):
|
40 |
+
# """Helper to save all four artifacts and return public URLs."""
|
41 |
+
# request_dir = os.path.join(OUTPUT_DIR, session_id)
|
42 |
+
# os.makedirs(request_dir, exist_ok=True)
|
43 |
+
|
44 |
+
# # filenames
|
45 |
+
# out_fn = "overlay.jpg"
|
46 |
+
# outlines_fn = "outlines.jpg"
|
47 |
+
# mask_fn = "mask.jpg"
|
48 |
+
# current_date = datetime.now().strftime("%d-%m-%Y")
|
49 |
+
# dxf_fn = f"out_{current_date}_{session_id}.dxf"
|
50 |
+
|
51 |
+
# # full paths
|
52 |
+
# out_path = os.path.join(request_dir, out_fn)
|
53 |
+
# outlines_path = os.path.join(request_dir, outlines_fn)
|
54 |
+
# mask_path = os.path.join(request_dir, mask_fn)
|
55 |
+
# new_dxf_path = os.path.join(request_dir, dxf_fn)
|
56 |
+
|
57 |
+
# # save images
|
58 |
+
# Image.fromarray(output_image).save(out_path)
|
59 |
+
# Image.fromarray(outlines).save(outlines_path)
|
60 |
+
# Image.fromarray(mask).save(mask_path)
|
61 |
+
|
62 |
+
# # copy dx file
|
63 |
+
# if os.path.exists(dxf_path):
|
64 |
+
# shutil.copy(dxf_path, new_dxf_path)
|
65 |
+
# else:
|
66 |
+
# # fallback if your DXF generator returns bytes or string
|
67 |
+
# with open(new_dxf_path, "wb") as f:
|
68 |
+
# if isinstance(dxf_path, (bytes, bytearray)):
|
69 |
+
# f.write(dxf_path)
|
70 |
+
# else:
|
71 |
+
# f.write(str(dxf_path).encode("utf-8"))
|
72 |
+
|
73 |
+
# # build URLs
|
74 |
+
# return {
|
75 |
+
# "output_image_url": f"{BASE_URL}/outputs/{session_id}/{out_fn}",
|
76 |
+
# "outlines_url": f"{BASE_URL}/outputs/{session_id}/{outlines_fn}",
|
77 |
+
# "mask_url": f"{BASE_URL}/outputs/{session_id}/{mask_fn}",
|
78 |
+
# "dxf_url": f"{BASE_URL}/outputs/{session_id}/{dxf_fn}",
|
79 |
+
# }
|
80 |
+
|
81 |
+
|
82 |
+
# @app.post("/predict1")
|
83 |
+
# async def predict1_api(
|
84 |
+
# file: UploadFile = File(...)
|
85 |
+
# ):
|
86 |
+
# """
|
87 |
+
# Simple predict: only image → overlay, outlines, mask, DXF
|
88 |
+
# """
|
89 |
+
# session_id = str(uuid.uuid4())
|
90 |
+
# try:
|
91 |
+
# img_bytes = await file.read()
|
92 |
+
# image = np.array(Image.open(io.BytesIO(img_bytes)).convert("RGB"))
|
93 |
+
# except Exception:
|
94 |
+
# raise HTTPException(400, "Invalid image upload")
|
95 |
+
|
96 |
+
# try:
|
97 |
+
# start = timeit.default_timer()
|
98 |
+
# out_img, outlines, dxf_path, mask = predict_simple(image)
|
99 |
+
# elapsed = timeit.default_timer() - start
|
100 |
+
# print(f"[{session_id}] predict1 in {elapsed:.2f}s")
|
101 |
+
|
102 |
+
# return save_and_build_urls(session_id, out_img, outlines, dxf_path, mask)
|
103 |
+
|
104 |
+
# except Exception as e:
|
105 |
+
# raise HTTPException(500, f"predict1 failed: {e}")
|
106 |
+
# except ReferenceBoxNotDetectedError:
|
107 |
+
# raise HTTPException(status_code=400, detail="Error detecting reference battery! Please try again with a clearer image.")
|
108 |
+
# except FingerCutOverlapError:
|
109 |
+
# raise HTTPException(status_code=400, detail="There was an overlap with fingercuts!s Please try again to generate dxf.")
|
110 |
+
|
111 |
+
|
112 |
+
# @app.post("/predict2")
|
113 |
+
# async def predict2_api(
|
114 |
+
# file: UploadFile = File(...),
|
115 |
+
# enable_fillet: str = Form(..., regex="^(On|Off)$"),
|
116 |
+
# fillet_value_mm: float = Form(...)
|
117 |
+
# ):
|
118 |
+
# """
|
119 |
+
# Middle predict: image + fillet toggle + fillet value → overlay, outlines, mask, DXF
|
120 |
+
# """
|
121 |
+
# session_id = str(uuid.uuid4())
|
122 |
+
# try:
|
123 |
+
# img_bytes = await file.read()
|
124 |
+
# image = np.array(Image.open(io.BytesIO(img_bytes)).convert("RGB"))
|
125 |
+
# except Exception:
|
126 |
+
# raise HTTPException(400, "Invalid image upload")
|
127 |
+
|
128 |
+
# try:
|
129 |
+
# start = timeit.default_timer()
|
130 |
+
# out_img, outlines, dxf_path, mask = predict_middle(
|
131 |
+
# image, enable_fillet, fillet_value_mm
|
132 |
+
# )
|
133 |
+
# elapsed = timeit.default_timer() - start
|
134 |
+
# print(f"[{session_id}] predict2 in {elapsed:.2f}s")
|
135 |
+
|
136 |
+
# return save_and_build_urls(session_id, out_img, outlines, dxf_path, mask)
|
137 |
+
|
138 |
+
# except Exception as e:
|
139 |
+
# raise HTTPException(500, f"predict2 failed: {e}")
|
140 |
+
# except ReferenceBoxNotDetectedError:
|
141 |
+
# raise HTTPException(status_code=400, detail="Error detecting reference battery! Please try again with a clearer image.")
|
142 |
+
# except FingerCutOverlapError:
|
143 |
+
# raise HTTPException(status_code=400, detail="There was an overlap with fingercuts!s Please try again to generate dxf.")
|
144 |
+
|
145 |
+
# @app.post("/predict3")
|
146 |
+
# async def predict3_api(
|
147 |
+
# file: UploadFile = File(...),
|
148 |
+
# enable_fillet: str = Form(..., regex="^(On|Off)$"),
|
149 |
+
# fillet_value_mm: float = Form(...),
|
150 |
+
# enable_finger_cut: str = Form(..., regex="^(On|Off)$")
|
151 |
+
# ):
|
152 |
+
# """
|
153 |
+
# Full predict: image + fillet toggle/value + finger-cut toggle → overlay, outlines, mask, DXF
|
154 |
+
# """
|
155 |
+
# session_id = str(uuid.uuid4())
|
156 |
+
# try:
|
157 |
+
# img_bytes = await file.read()
|
158 |
+
# image = np.array(Image.open(io.BytesIO(img_bytes)).convert("RGB"))
|
159 |
+
# except Exception:
|
160 |
+
# raise HTTPException(400, "Invalid image upload")
|
161 |
+
|
162 |
+
# try:
|
163 |
+
# start = timeit.default_timer()
|
164 |
+
# out_img, outlines, dxf_path, mask = predict_full(
|
165 |
+
# image, enable_fillet, fillet_value_mm, enable_finger_cut
|
166 |
+
# )
|
167 |
+
# elapsed = timeit.default_timer() - start
|
168 |
+
# print(f"[{session_id}] predict3 in {elapsed:.2f}s")
|
169 |
+
|
170 |
+
# return save_and_build_urls(session_id, out_img, outlines, dxf_path, mask)
|
171 |
+
|
172 |
+
# except Exception as e:
|
173 |
+
# raise HTTPException(500, f"predict3 failed: {e}")
|
174 |
+
# except ReferenceBoxNotDetectedError:
|
175 |
+
# raise HTTPException(status_code=400, detail="Error detecting reference battery! Please try again with a clearer image.")
|
176 |
+
# except FingerCutOverlapError:
|
177 |
+
# raise HTTPException(status_code=400, detail="There was an overlap with fingercuts!s Please try again to generate dxf.")
|
178 |
+
|
179 |
+
# @app.post("/update")
|
180 |
+
# async def update_files(
|
181 |
+
# output_image: UploadFile = File(...),
|
182 |
+
# outlines_image: UploadFile = File(...),
|
183 |
+
# mask_image: UploadFile = File(...),
|
184 |
+
# dxf_file: UploadFile = File(...)
|
185 |
+
# ):
|
186 |
+
# session_id = str(uuid.uuid4())
|
187 |
+
# update_dir = os.path.join(UPDATES_DIR, session_id)
|
188 |
+
# os.makedirs(update_dir, exist_ok=True)
|
189 |
+
|
190 |
+
# try:
|
191 |
+
# upload_map = {
|
192 |
+
# "output_image": output_image,
|
193 |
+
# "outlines_image": outlines_image,
|
194 |
+
# "mask_image": mask_image,
|
195 |
+
# "dxf_file": dxf_file,
|
196 |
+
# }
|
197 |
+
# urls = {}
|
198 |
+
# for key, up in upload_map.items():
|
199 |
+
# fn = up.filename
|
200 |
+
# path = os.path.join(update_dir, fn)
|
201 |
+
# with open(path, "wb") as f:
|
202 |
+
# shutil.copyfileobj(up.file, f)
|
203 |
+
# urls[key] = f"{BASE_URL}/updates/{session_id}/{fn}"
|
204 |
+
|
205 |
+
# return {"session_id": session_id, "uploaded": urls}
|
206 |
+
|
207 |
+
# except Exception as e:
|
208 |
+
# raise HTTPException(500, f"Update failed: {e}")
|
209 |
+
|
210 |
+
|
211 |
+
# if __name__ == "__main__":
|
212 |
+
# import uvicorn
|
213 |
+
# port = int(os.environ.get("PORT", 8082))
|
214 |
+
# print(f"Starting FastAPI server on 0.0.0.0:{port}...")
|
215 |
+
# uvicorn.run(app, host="0.0.0.0", port=port)
|
216 |
+
|
217 |
+
|
218 |
+
|
219 |
+
|
220 |
+
|
221 |
+
|
222 |
+
|
223 |
+
|
224 |
+
|
225 |
+
|
226 |
+
from fastapi import FastAPI, HTTPException, UploadFile, File, Form
|
227 |
+
from pydantic import BaseModel
|
228 |
+
import numpy as np
|
229 |
+
from PIL import Image
|
230 |
+
import io, uuid, os, shutil, timeit
|
231 |
+
from datetime import datetime
|
232 |
+
from fastapi.staticfiles import StaticFiles
|
233 |
+
from fastapi.middleware.cors import CORSMiddleware
|
234 |
+
from fastapi.responses import FileResponse
|
235 |
+
|
236 |
+
# import your three wrappers
|
237 |
+
from app import predict_simple, predict_middle, predict_full
|
238 |
+
|
239 |
+
from app import (
|
240 |
+
predict_simple, predict_middle, predict_full,
|
241 |
+
ReferenceBoxNotDetectedError,
|
242 |
+
FingerCutOverlapError
|
243 |
+
)
|
244 |
+
|
245 |
+
|
246 |
+
app = FastAPI()
|
247 |
+
|
248 |
+
# allow CORS if needed
|
249 |
+
app.add_middleware(
|
250 |
+
CORSMiddleware,
|
251 |
+
allow_origins=["*"],
|
252 |
+
allow_methods=["*"],
|
253 |
+
allow_headers=["*"],
|
254 |
+
)
|
255 |
+
|
256 |
+
BASE_URL = "https://snapanddtraceapp-988917236820.us-central1.run.app"
|
257 |
+
|
258 |
+
OUTPUT_DIR = os.path.abspath("./outputs")
|
259 |
+
os.makedirs(OUTPUT_DIR, exist_ok=True)
|
260 |
+
|
261 |
+
UPDATES_DIR = os.path.abspath("./updates")
|
262 |
+
os.makedirs(UPDATES_DIR, exist_ok=True)
|
263 |
+
|
264 |
+
# Mount static directories with normal StaticFiles
|
265 |
+
app.mount("/outputs", StaticFiles(directory=OUTPUT_DIR), name="outputs")
|
266 |
+
app.mount("/updates", StaticFiles(directory=UPDATES_DIR), name="updates")
|
267 |
+
|
268 |
+
|
269 |
+
def save_and_build_urls(
|
270 |
+
session_id: str,
|
271 |
+
output_image: np.ndarray,
|
272 |
+
outlines: np.ndarray,
|
273 |
+
dxf_path: str,
|
274 |
+
mask: np.ndarray,
|
275 |
+
endpoint_type: str,
|
276 |
+
fillet_value: float = None,
|
277 |
+
finger_cut: str = None
|
278 |
+
):
|
279 |
+
"""Helper to save all four artifacts and return public URLs."""
|
280 |
+
request_dir = os.path.join(OUTPUT_DIR, session_id)
|
281 |
+
os.makedirs(request_dir, exist_ok=True)
|
282 |
+
|
283 |
+
# filenames
|
284 |
+
out_fn = "overlay.jpg"
|
285 |
+
outlines_fn = "outlines.jpg"
|
286 |
+
mask_fn = "mask.jpg"
|
287 |
+
|
288 |
+
# Get current date
|
289 |
+
current_date = datetime.utcnow().strftime("%d-%m-%Y")
|
290 |
+
|
291 |
+
|
292 |
+
# Format fillet value with underscore instead of dot
|
293 |
+
fillet_str = f"{fillet_value:.2f}".replace(".", "_") if fillet_value is not None else None
|
294 |
+
|
295 |
+
# Determine DXF filename based on endpoint type
|
296 |
+
if endpoint_type == "predict1":
|
297 |
+
dxf_fn = f"DXF_{current_date}.dxf"
|
298 |
+
elif endpoint_type == "predict2":
|
299 |
+
dxf_fn = f"DXF_{current_date}.dxf"
|
300 |
+
elif endpoint_type == "predict3":
|
301 |
+
dxf_fn = f"DXF_{current_date}.dxf"
|
302 |
+
|
303 |
+
# full paths
|
304 |
+
out_path = os.path.join(request_dir, out_fn)
|
305 |
+
outlines_path = os.path.join(request_dir, outlines_fn)
|
306 |
+
mask_path = os.path.join(request_dir, mask_fn)
|
307 |
+
new_dxf_path = os.path.join(request_dir, dxf_fn)
|
308 |
+
|
309 |
+
# save images
|
310 |
+
Image.fromarray(output_image).save(out_path)
|
311 |
+
Image.fromarray(outlines).save(outlines_path)
|
312 |
+
Image.fromarray(mask).save(mask_path)
|
313 |
+
|
314 |
+
# copy dxf file
|
315 |
+
if os.path.exists(dxf_path):
|
316 |
+
shutil.copy(dxf_path, new_dxf_path)
|
317 |
+
else:
|
318 |
+
# fallback if your DXF generator returns bytes or string
|
319 |
+
with open(new_dxf_path, "wb") as f:
|
320 |
+
if isinstance(dxf_path, (bytes, bytearray)):
|
321 |
+
f.write(dxf_path)
|
322 |
+
else:
|
323 |
+
f.write(str(dxf_path).encode("utf-8"))
|
324 |
+
|
325 |
+
# build URLs with /download prefix for DXF
|
326 |
+
return {
|
327 |
+
"output_image_url": f"{BASE_URL}/outputs/{session_id}/{out_fn}",
|
328 |
+
"outlines_url": f"{BASE_URL}/outputs/{session_id}/{outlines_fn}",
|
329 |
+
"mask_url": f"{BASE_URL}/outputs/{session_id}/{mask_fn}",
|
330 |
+
"dxf_url": f"{BASE_URL}/download/{session_id}/{dxf_fn}", # Changed to use download endpoint
|
331 |
+
}
|
332 |
+
|
333 |
+
# Add new endpoint for downloading DXF files
|
334 |
+
@app.get("/download/{session_id}/{filename}")
|
335 |
+
async def download_file(session_id: str, filename: str):
|
336 |
+
file_path = os.path.join(OUTPUT_DIR, session_id, filename)
|
337 |
+
if not os.path.exists(file_path):
|
338 |
+
raise HTTPException(status_code=404, detail="File not found")
|
339 |
+
|
340 |
+
return FileResponse(
|
341 |
+
path=file_path,
|
342 |
+
filename=filename,
|
343 |
+
media_type="application/x-dxf",
|
344 |
+
headers={"Content-Disposition": f"attachment; filename={filename}"}
|
345 |
+
)
|
346 |
+
|
347 |
+
|
348 |
+
@app.post("/predict1")
|
349 |
+
async def predict1_api(
|
350 |
+
file: UploadFile = File(...)
|
351 |
+
):
|
352 |
+
"""
|
353 |
+
Simple predict: only image → overlay, outlines, mask, DXF
|
354 |
+
DXF naming format: DXF_DD-MM-YYYY.dxf
|
355 |
+
"""
|
356 |
+
session_id = str(uuid.uuid4())
|
357 |
+
try:
|
358 |
+
img_bytes = await file.read()
|
359 |
+
image = np.array(Image.open(io.BytesIO(img_bytes)).convert("RGB"))
|
360 |
+
except Exception:
|
361 |
+
raise HTTPException(400, "Invalid image upload")
|
362 |
+
|
363 |
+
try:
|
364 |
+
start = timeit.default_timer()
|
365 |
+
out_img, outlines, dxf_path, mask = predict_simple(image)
|
366 |
+
elapsed = timeit.default_timer() - start
|
367 |
+
print(f"[{session_id}] predict1 in {elapsed:.2f}s")
|
368 |
+
|
369 |
+
return save_and_build_urls(
|
370 |
+
session_id=session_id,
|
371 |
+
output_image=out_img,
|
372 |
+
outlines=outlines,
|
373 |
+
dxf_path=dxf_path,
|
374 |
+
mask=mask,
|
375 |
+
endpoint_type="predict1"
|
376 |
+
)
|
377 |
+
|
378 |
+
except ReferenceBoxNotDetectedError:
|
379 |
+
raise HTTPException(status_code=400, detail="Error detecting reference battery! Please try again with a clearer image.")
|
380 |
+
except FingerCutOverlapError:
|
381 |
+
raise HTTPException(status_code=400, detail="There was an overlap with fingercuts! Please try again to generate dxf.")
|
382 |
+
except HTTPException as e:
|
383 |
+
raise e
|
384 |
+
except Exception as e:
|
385 |
+
raise HTTPException(status_code=500, detail="Error detecting reference battery! Please try again with a clearer image.")
|
386 |
+
|
387 |
+
@app.post("/predict2")
|
388 |
+
async def predict2_api(
|
389 |
+
file: UploadFile = File(...),
|
390 |
+
enable_fillet: str = Form(..., regex="^(On|Off)$"),
|
391 |
+
fillet_value_mm: float = Form(...)
|
392 |
+
):
|
393 |
+
"""
|
394 |
+
Middle predict: image + fillet toggle + fillet value → overlay, outlines, mask, DXF
|
395 |
+
DXF naming format: DXF_DD-MM-YYYY_fillet-value_mm.dxf
|
396 |
+
"""
|
397 |
+
session_id = str(uuid.uuid4())
|
398 |
+
try:
|
399 |
+
img_bytes = await file.read()
|
400 |
+
image = np.array(Image.open(io.BytesIO(img_bytes)).convert("RGB"))
|
401 |
+
except Exception:
|
402 |
+
raise HTTPException(400, "Invalid image upload")
|
403 |
+
|
404 |
+
try:
|
405 |
+
start = timeit.default_timer()
|
406 |
+
out_img, outlines, dxf_path, mask = predict_middle(
|
407 |
+
image, enable_fillet, fillet_value_mm
|
408 |
+
)
|
409 |
+
elapsed = timeit.default_timer() - start
|
410 |
+
print(f"[{session_id}] predict2 in {elapsed:.2f}s")
|
411 |
+
|
412 |
+
return save_and_build_urls(
|
413 |
+
session_id=session_id,
|
414 |
+
output_image=out_img,
|
415 |
+
outlines=outlines,
|
416 |
+
dxf_path=dxf_path,
|
417 |
+
mask=mask,
|
418 |
+
endpoint_type="predict2",
|
419 |
+
fillet_value=fillet_value_mm
|
420 |
+
)
|
421 |
+
|
422 |
+
except ReferenceBoxNotDetectedError:
|
423 |
+
raise HTTPException(status_code=400, detail="Error detecting reference battery! Please try again with a clearer image.")
|
424 |
+
except FingerCutOverlapError:
|
425 |
+
raise HTTPException(status_code=400, detail="There was an overlap with fingercuts! Please try again to generate dxf.")
|
426 |
+
except HTTPException as e:
|
427 |
+
raise e
|
428 |
+
except Exception as e:
|
429 |
+
raise HTTPException(status_code=500, detail="Error detecting reference battery! Please try again with a clearer image.")
|
430 |
+
|
431 |
+
|
432 |
+
@app.post("/predict3")
|
433 |
+
async def predict3_api(
|
434 |
+
file: UploadFile = File(...),
|
435 |
+
enable_fillet: str = Form(..., regex="^(On|Off)$"),
|
436 |
+
fillet_value_mm: float = Form(...),
|
437 |
+
enable_finger_cut: str = Form(..., regex="^(On|Off)$")
|
438 |
+
):
|
439 |
+
"""
|
440 |
+
Full predict: image + fillet toggle/value + finger-cut toggle → overlay, outlines, mask, DXF
|
441 |
+
DXF naming format: DXF_DD-MM-YYYY_fillet-value_mm_fingercut-On|Off.dxf
|
442 |
+
"""
|
443 |
+
session_id = str(uuid.uuid4())
|
444 |
+
try:
|
445 |
+
img_bytes = await file.read()
|
446 |
+
image = np.array(Image.open(io.BytesIO(img_bytes)).convert("RGB"))
|
447 |
+
except Exception:
|
448 |
+
raise HTTPException(400, "Invalid image upload")
|
449 |
+
|
450 |
+
try:
|
451 |
+
start = timeit.default_timer()
|
452 |
+
out_img, outlines, dxf_path, mask = predict_full(
|
453 |
+
image, enable_fillet, fillet_value_mm, enable_finger_cut
|
454 |
+
)
|
455 |
+
elapsed = timeit.default_timer() - start
|
456 |
+
print(f"[{session_id}] predict3 in {elapsed:.2f}s")
|
457 |
+
|
458 |
+
return save_and_build_urls(
|
459 |
+
session_id=session_id,
|
460 |
+
output_image=out_img,
|
461 |
+
outlines=outlines,
|
462 |
+
dxf_path=dxf_path,
|
463 |
+
mask=mask,
|
464 |
+
endpoint_type="predict3",
|
465 |
+
fillet_value=fillet_value_mm,
|
466 |
+
finger_cut=enable_finger_cut
|
467 |
+
)
|
468 |
+
|
469 |
+
except ReferenceBoxNotDetectedError:
|
470 |
+
raise HTTPException(status_code=400, detail="Error detecting reference battery! Please try again with a clearer image.")
|
471 |
+
except FingerCutOverlapError:
|
472 |
+
raise HTTPException(status_code=400, detail="There was an overlap with fingercuts! Please try again to generate dxf.")
|
473 |
+
except HTTPException as e:
|
474 |
+
raise e
|
475 |
+
except Exception as e:
|
476 |
+
raise HTTPException(status_code=500, detail="Error detecting reference battery! Please try again with a clearer image.")
|
477 |
+
|
478 |
+
|
479 |
+
@app.post("/update")
|
480 |
+
async def update_files(
|
481 |
+
output_image: UploadFile = File(...),
|
482 |
+
outlines_image: UploadFile = File(...),
|
483 |
+
mask_image: UploadFile = File(...),
|
484 |
+
dxf_file: UploadFile = File(...)
|
485 |
+
):
|
486 |
+
session_id = str(uuid.uuid4())
|
487 |
+
update_dir = os.path.join(UPDATES_DIR, session_id)
|
488 |
+
os.makedirs(update_dir, exist_ok=True)
|
489 |
+
|
490 |
+
try:
|
491 |
+
upload_map = {
|
492 |
+
"output_image": output_image,
|
493 |
+
"outlines_image": outlines_image,
|
494 |
+
"mask_image": mask_image,
|
495 |
+
"dxf_file": dxf_file,
|
496 |
+
}
|
497 |
+
urls = {}
|
498 |
+
for key, up in upload_map.items():
|
499 |
+
fn = up.filename
|
500 |
+
path = os.path.join(update_dir, fn)
|
501 |
+
with open(path, "wb") as f:
|
502 |
+
shutil.copyfileobj(up.file, f)
|
503 |
+
urls[key] = f"{BASE_URL}/updates/{session_id}/{fn}"
|
504 |
+
|
505 |
+
return {"session_id": session_id, "uploaded": urls}
|
506 |
+
|
507 |
+
except Exception as e:
|
508 |
+
raise HTTPException(500, f"Update failed: {e}")
|
509 |
+
|
510 |
+
|
511 |
+
from fastapi import Response
|
512 |
+
|
513 |
+
@app.get("/health")
|
514 |
+
def health():
|
515 |
+
return Response(content="OK", status_code=200)
|
516 |
+
|
517 |
+
|
518 |
+
if __name__ == "__main__":
|
519 |
+
import uvicorn
|
520 |
+
port = int(os.environ.get("PORT", 8080))
|
521 |
+
print(f"Starting FastAPI server on 0.0.0.0:{port}...")
|
522 |
+
uvicorn.run(app, host="0.0.0.0", port=port)
|
523 |
+
|
524 |
+
|
525 |
+
|
scalingtestupdated.py
ADDED
@@ -0,0 +1,184 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import cv2
|
2 |
+
import numpy as np
|
3 |
+
import os
|
4 |
+
import argparse
|
5 |
+
from typing import Union
|
6 |
+
from matplotlib import pyplot as plt
|
7 |
+
|
8 |
+
class ScalingSquareDetector:
|
9 |
+
def __init__(self, feature_detector="ORB", debug=False):
|
10 |
+
"""
|
11 |
+
Initialize the detector with the desired feature matching algorithm.
|
12 |
+
:param feature_detector: "ORB" or "SIFT" (default is "ORB").
|
13 |
+
:param debug: If True, saves intermediate images for debugging.
|
14 |
+
"""
|
15 |
+
self.feature_detector = feature_detector
|
16 |
+
self.debug = debug
|
17 |
+
self.detector = self._initialize_detector()
|
18 |
+
|
19 |
+
def _initialize_detector(self):
|
20 |
+
"""
|
21 |
+
Initialize the chosen feature detector.
|
22 |
+
:return: OpenCV detector object.
|
23 |
+
"""
|
24 |
+
if self.feature_detector.upper() == "SIFT":
|
25 |
+
return cv2.SIFT_create()
|
26 |
+
elif self.feature_detector.upper() == "ORB":
|
27 |
+
return cv2.ORB_create()
|
28 |
+
else:
|
29 |
+
raise ValueError("Invalid feature detector. Choose 'ORB' or 'SIFT'.")
|
30 |
+
|
31 |
+
def find_scaling_square(
|
32 |
+
self, target_image, known_size_mm, roi_margin=30
|
33 |
+
):
|
34 |
+
"""
|
35 |
+
Detect the scaling square in the target image based on the reference image.
|
36 |
+
:param reference_image_path: Path to the reference image of the square.
|
37 |
+
:param target_image_path: Path to the target image containing the square.
|
38 |
+
:param known_size_mm: Physical size of the square in millimeters.
|
39 |
+
:param roi_margin: Margin to expand the ROI around the detected square (in pixels).
|
40 |
+
:return: Scaling factor (mm per pixel).
|
41 |
+
"""
|
42 |
+
contours, _ = cv2.findContours(
|
43 |
+
target_image, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE
|
44 |
+
)
|
45 |
+
|
46 |
+
if not contours:
|
47 |
+
raise ValueError("No contours found in the cropped ROI.")
|
48 |
+
|
49 |
+
# # Select the largest square-like contour
|
50 |
+
print(f"No of contours: {len(contours)}")
|
51 |
+
largest_square = None
|
52 |
+
# largest_square_area = 0
|
53 |
+
# for contour in contours:
|
54 |
+
# x_c, y_c, w_c, h_c = cv2.boundingRect(contour)
|
55 |
+
# aspect_ratio = w_c / float(h_c)
|
56 |
+
# if 0.9 <= aspect_ratio <= 1.1:
|
57 |
+
# peri = cv2.arcLength(contour, True)
|
58 |
+
# approx = cv2.approxPolyDP(contour, 0.02 * peri, True)
|
59 |
+
# if len(approx) == 4:
|
60 |
+
# area = cv2.contourArea(contour)
|
61 |
+
# if area > largest_square_area:
|
62 |
+
# largest_square = contour
|
63 |
+
# largest_square_area = area
|
64 |
+
|
65 |
+
for contour in contours:
|
66 |
+
largest_square = contour
|
67 |
+
|
68 |
+
# if largest_square is None:
|
69 |
+
# raise ValueError("No square-like contour found in the ROI.")
|
70 |
+
|
71 |
+
# Draw the largest contour on the original image
|
72 |
+
target_image_color = cv2.cvtColor(target_image, cv2.COLOR_GRAY2BGR)
|
73 |
+
cv2.drawContours(
|
74 |
+
target_image_color, largest_square, -1, (255, 0, 0), 3
|
75 |
+
)
|
76 |
+
|
77 |
+
# if self.debug:
|
78 |
+
cv2.imwrite("largest_contour.jpg", target_image_color)
|
79 |
+
|
80 |
+
# Calculate the bounding rectangle of the largest contour
|
81 |
+
x, y, w, h = cv2.boundingRect(largest_square)
|
82 |
+
square_width_px = w
|
83 |
+
square_height_px = h
|
84 |
+
print(f"Reference object size: {known_size_mm} mm")
|
85 |
+
print(f"width: {square_width_px} px")
|
86 |
+
print(f"height: {square_height_px} px")
|
87 |
+
|
88 |
+
# Calculate the scaling factor
|
89 |
+
avg_square_size_px = (square_width_px + square_height_px) / 2
|
90 |
+
print(f"avg square size: {avg_square_size_px} px")
|
91 |
+
scaling_factor = known_size_mm / avg_square_size_px # mm per pixel
|
92 |
+
print(f"scaling factor: {scaling_factor} mm per pixel")
|
93 |
+
|
94 |
+
return scaling_factor #, square_height_px, square_width_px, roi_binary
|
95 |
+
|
96 |
+
def draw_debug_images(self, output_folder):
|
97 |
+
"""
|
98 |
+
Save debug images if enabled.
|
99 |
+
:param output_folder: Directory to save debug images.
|
100 |
+
"""
|
101 |
+
if self.debug:
|
102 |
+
if not os.path.exists(output_folder):
|
103 |
+
os.makedirs(output_folder)
|
104 |
+
debug_images = ["largest_contour.jpg"]
|
105 |
+
for img_name in debug_images:
|
106 |
+
if os.path.exists(img_name):
|
107 |
+
os.rename(img_name, os.path.join(output_folder, img_name))
|
108 |
+
|
109 |
+
|
110 |
+
def calculate_scaling_factor(
|
111 |
+
target_image,
|
112 |
+
reference_obj_size_mm,
|
113 |
+
feature_detector="ORB",
|
114 |
+
debug=False,
|
115 |
+
roi_margin=30,
|
116 |
+
):
|
117 |
+
# Initialize detector
|
118 |
+
detector = ScalingSquareDetector(feature_detector=feature_detector, debug=debug)
|
119 |
+
|
120 |
+
# Find scaling square and calculate scaling factor
|
121 |
+
scaling_factor = detector.find_scaling_square(
|
122 |
+
target_image=target_image,
|
123 |
+
known_size_mm=reference_obj_size_mm,
|
124 |
+
roi_margin=roi_margin,
|
125 |
+
)
|
126 |
+
|
127 |
+
# Save debug images
|
128 |
+
if debug:
|
129 |
+
detector.draw_debug_images("debug_outputs")
|
130 |
+
|
131 |
+
return scaling_factor
|
132 |
+
|
133 |
+
|
134 |
+
# Example usage:
|
135 |
+
if __name__ == "__main__":
|
136 |
+
import os
|
137 |
+
from PIL import Image
|
138 |
+
from ultralytics import YOLO
|
139 |
+
from app import yolo_detect, shrink_bbox
|
140 |
+
from ultralytics.utils.plotting import save_one_box
|
141 |
+
|
142 |
+
for idx, file in enumerate(os.listdir("./sample_images")):
|
143 |
+
img = np.array(Image.open(os.path.join("./sample_images", file)))
|
144 |
+
img = yolo_detect(img, ['box'])
|
145 |
+
model = YOLO("./best.pt")
|
146 |
+
res = model.predict(img, conf=0.6)
|
147 |
+
|
148 |
+
box_img = save_one_box(res[0].cpu().boxes.xyxy, im=res[0].orig_img, save=False)
|
149 |
+
# img = shrink_bbox(box_img, 1.20)
|
150 |
+
cv2.imwrite(f"./outputs/{idx}_{file}", box_img)
|
151 |
+
|
152 |
+
print("File: ",f"./outputs/{idx}_{file}")
|
153 |
+
try:
|
154 |
+
|
155 |
+
scaling_factor = calculate_scaling_factor(
|
156 |
+
target_image=box_img,
|
157 |
+
known_square_size_mm=20,
|
158 |
+
feature_detector="ORB",
|
159 |
+
debug=False,
|
160 |
+
roi_margin=90,
|
161 |
+
)
|
162 |
+
# cv2.imwrite(f"./outputs/{idx}_binary_{file}", roi_binary)
|
163 |
+
|
164 |
+
# Square size in mm
|
165 |
+
# square_size_mm = 12.7
|
166 |
+
|
167 |
+
# # Compute the calculated scaling factors and compare
|
168 |
+
# calculated_scaling_factor = square_size_mm / height_px
|
169 |
+
# discrepancy = abs(calculated_scaling_factor - scaling_factor)
|
170 |
+
# import pprint
|
171 |
+
# pprint.pprint({
|
172 |
+
# "height_px": height_px,
|
173 |
+
# "width_px": width_px,
|
174 |
+
# "given_scaling_factor": scaling_factor,
|
175 |
+
# "calculated_scaling_factor": calculated_scaling_factor,
|
176 |
+
# "discrepancy": discrepancy,
|
177 |
+
# })
|
178 |
+
|
179 |
+
|
180 |
+
print(f"Scaling Factor (mm per pixel): {scaling_factor:.6f}")
|
181 |
+
except Exception as e:
|
182 |
+
from traceback import print_exc
|
183 |
+
print(print_exc())
|
184 |
+
print(f"Error: {e}")
|
u2netp.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e7567cde013fb64813973ce6e1ecc25a80c05c3ca7adbc5a54f3c3d90991b854
|
3 |
+
size 4683258
|
u2netp.py
ADDED
@@ -0,0 +1,525 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import torch
|
2 |
+
import torch.nn as nn
|
3 |
+
import torch.nn.functional as F
|
4 |
+
|
5 |
+
class REBNCONV(nn.Module):
|
6 |
+
def __init__(self,in_ch=3,out_ch=3,dirate=1):
|
7 |
+
super(REBNCONV,self).__init__()
|
8 |
+
|
9 |
+
self.conv_s1 = nn.Conv2d(in_ch,out_ch,3,padding=1*dirate,dilation=1*dirate)
|
10 |
+
self.bn_s1 = nn.BatchNorm2d(out_ch)
|
11 |
+
self.relu_s1 = nn.ReLU(inplace=True)
|
12 |
+
|
13 |
+
def forward(self,x):
|
14 |
+
|
15 |
+
hx = x
|
16 |
+
xout = self.relu_s1(self.bn_s1(self.conv_s1(hx)))
|
17 |
+
|
18 |
+
return xout
|
19 |
+
|
20 |
+
## upsample tensor 'src' to have the same spatial size with tensor 'tar'
|
21 |
+
def _upsample_like(src,tar):
|
22 |
+
|
23 |
+
src = F.upsample(src,size=tar.shape[2:],mode='bilinear')
|
24 |
+
|
25 |
+
return src
|
26 |
+
|
27 |
+
|
28 |
+
### RSU-7 ###
|
29 |
+
class RSU7(nn.Module):#UNet07DRES(nn.Module):
|
30 |
+
|
31 |
+
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
|
32 |
+
super(RSU7,self).__init__()
|
33 |
+
|
34 |
+
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
|
35 |
+
|
36 |
+
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
|
37 |
+
self.pool1 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
38 |
+
|
39 |
+
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
40 |
+
self.pool2 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
41 |
+
|
42 |
+
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
43 |
+
self.pool3 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
44 |
+
|
45 |
+
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
46 |
+
self.pool4 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
47 |
+
|
48 |
+
self.rebnconv5 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
49 |
+
self.pool5 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
50 |
+
|
51 |
+
self.rebnconv6 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
52 |
+
|
53 |
+
self.rebnconv7 = REBNCONV(mid_ch,mid_ch,dirate=2)
|
54 |
+
|
55 |
+
self.rebnconv6d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
56 |
+
self.rebnconv5d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
57 |
+
self.rebnconv4d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
58 |
+
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
59 |
+
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
60 |
+
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
|
61 |
+
|
62 |
+
def forward(self,x):
|
63 |
+
|
64 |
+
hx = x
|
65 |
+
hxin = self.rebnconvin(hx)
|
66 |
+
|
67 |
+
hx1 = self.rebnconv1(hxin)
|
68 |
+
hx = self.pool1(hx1)
|
69 |
+
|
70 |
+
hx2 = self.rebnconv2(hx)
|
71 |
+
hx = self.pool2(hx2)
|
72 |
+
|
73 |
+
hx3 = self.rebnconv3(hx)
|
74 |
+
hx = self.pool3(hx3)
|
75 |
+
|
76 |
+
hx4 = self.rebnconv4(hx)
|
77 |
+
hx = self.pool4(hx4)
|
78 |
+
|
79 |
+
hx5 = self.rebnconv5(hx)
|
80 |
+
hx = self.pool5(hx5)
|
81 |
+
|
82 |
+
hx6 = self.rebnconv6(hx)
|
83 |
+
|
84 |
+
hx7 = self.rebnconv7(hx6)
|
85 |
+
|
86 |
+
hx6d = self.rebnconv6d(torch.cat((hx7,hx6),1))
|
87 |
+
hx6dup = _upsample_like(hx6d,hx5)
|
88 |
+
|
89 |
+
hx5d = self.rebnconv5d(torch.cat((hx6dup,hx5),1))
|
90 |
+
hx5dup = _upsample_like(hx5d,hx4)
|
91 |
+
|
92 |
+
hx4d = self.rebnconv4d(torch.cat((hx5dup,hx4),1))
|
93 |
+
hx4dup = _upsample_like(hx4d,hx3)
|
94 |
+
|
95 |
+
hx3d = self.rebnconv3d(torch.cat((hx4dup,hx3),1))
|
96 |
+
hx3dup = _upsample_like(hx3d,hx2)
|
97 |
+
|
98 |
+
hx2d = self.rebnconv2d(torch.cat((hx3dup,hx2),1))
|
99 |
+
hx2dup = _upsample_like(hx2d,hx1)
|
100 |
+
|
101 |
+
hx1d = self.rebnconv1d(torch.cat((hx2dup,hx1),1))
|
102 |
+
|
103 |
+
return hx1d + hxin
|
104 |
+
|
105 |
+
### RSU-6 ###
|
106 |
+
class RSU6(nn.Module):#UNet06DRES(nn.Module):
|
107 |
+
|
108 |
+
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
|
109 |
+
super(RSU6,self).__init__()
|
110 |
+
|
111 |
+
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
|
112 |
+
|
113 |
+
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
|
114 |
+
self.pool1 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
115 |
+
|
116 |
+
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
117 |
+
self.pool2 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
118 |
+
|
119 |
+
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
120 |
+
self.pool3 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
121 |
+
|
122 |
+
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
123 |
+
self.pool4 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
124 |
+
|
125 |
+
self.rebnconv5 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
126 |
+
|
127 |
+
self.rebnconv6 = REBNCONV(mid_ch,mid_ch,dirate=2)
|
128 |
+
|
129 |
+
self.rebnconv5d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
130 |
+
self.rebnconv4d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
131 |
+
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
132 |
+
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
133 |
+
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
|
134 |
+
|
135 |
+
def forward(self,x):
|
136 |
+
|
137 |
+
hx = x
|
138 |
+
|
139 |
+
hxin = self.rebnconvin(hx)
|
140 |
+
|
141 |
+
hx1 = self.rebnconv1(hxin)
|
142 |
+
hx = self.pool1(hx1)
|
143 |
+
|
144 |
+
hx2 = self.rebnconv2(hx)
|
145 |
+
hx = self.pool2(hx2)
|
146 |
+
|
147 |
+
hx3 = self.rebnconv3(hx)
|
148 |
+
hx = self.pool3(hx3)
|
149 |
+
|
150 |
+
hx4 = self.rebnconv4(hx)
|
151 |
+
hx = self.pool4(hx4)
|
152 |
+
|
153 |
+
hx5 = self.rebnconv5(hx)
|
154 |
+
|
155 |
+
hx6 = self.rebnconv6(hx5)
|
156 |
+
|
157 |
+
|
158 |
+
hx5d = self.rebnconv5d(torch.cat((hx6,hx5),1))
|
159 |
+
hx5dup = _upsample_like(hx5d,hx4)
|
160 |
+
|
161 |
+
hx4d = self.rebnconv4d(torch.cat((hx5dup,hx4),1))
|
162 |
+
hx4dup = _upsample_like(hx4d,hx3)
|
163 |
+
|
164 |
+
hx3d = self.rebnconv3d(torch.cat((hx4dup,hx3),1))
|
165 |
+
hx3dup = _upsample_like(hx3d,hx2)
|
166 |
+
|
167 |
+
hx2d = self.rebnconv2d(torch.cat((hx3dup,hx2),1))
|
168 |
+
hx2dup = _upsample_like(hx2d,hx1)
|
169 |
+
|
170 |
+
hx1d = self.rebnconv1d(torch.cat((hx2dup,hx1),1))
|
171 |
+
|
172 |
+
return hx1d + hxin
|
173 |
+
|
174 |
+
### RSU-5 ###
|
175 |
+
class RSU5(nn.Module):#UNet05DRES(nn.Module):
|
176 |
+
|
177 |
+
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
|
178 |
+
super(RSU5,self).__init__()
|
179 |
+
|
180 |
+
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
|
181 |
+
|
182 |
+
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
|
183 |
+
self.pool1 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
184 |
+
|
185 |
+
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
186 |
+
self.pool2 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
187 |
+
|
188 |
+
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
189 |
+
self.pool3 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
190 |
+
|
191 |
+
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
192 |
+
|
193 |
+
self.rebnconv5 = REBNCONV(mid_ch,mid_ch,dirate=2)
|
194 |
+
|
195 |
+
self.rebnconv4d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
196 |
+
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
197 |
+
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
198 |
+
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
|
199 |
+
|
200 |
+
def forward(self,x):
|
201 |
+
|
202 |
+
hx = x
|
203 |
+
|
204 |
+
hxin = self.rebnconvin(hx)
|
205 |
+
|
206 |
+
hx1 = self.rebnconv1(hxin)
|
207 |
+
hx = self.pool1(hx1)
|
208 |
+
|
209 |
+
hx2 = self.rebnconv2(hx)
|
210 |
+
hx = self.pool2(hx2)
|
211 |
+
|
212 |
+
hx3 = self.rebnconv3(hx)
|
213 |
+
hx = self.pool3(hx3)
|
214 |
+
|
215 |
+
hx4 = self.rebnconv4(hx)
|
216 |
+
|
217 |
+
hx5 = self.rebnconv5(hx4)
|
218 |
+
|
219 |
+
hx4d = self.rebnconv4d(torch.cat((hx5,hx4),1))
|
220 |
+
hx4dup = _upsample_like(hx4d,hx3)
|
221 |
+
|
222 |
+
hx3d = self.rebnconv3d(torch.cat((hx4dup,hx3),1))
|
223 |
+
hx3dup = _upsample_like(hx3d,hx2)
|
224 |
+
|
225 |
+
hx2d = self.rebnconv2d(torch.cat((hx3dup,hx2),1))
|
226 |
+
hx2dup = _upsample_like(hx2d,hx1)
|
227 |
+
|
228 |
+
hx1d = self.rebnconv1d(torch.cat((hx2dup,hx1),1))
|
229 |
+
|
230 |
+
return hx1d + hxin
|
231 |
+
|
232 |
+
### RSU-4 ###
|
233 |
+
class RSU4(nn.Module):#UNet04DRES(nn.Module):
|
234 |
+
|
235 |
+
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
|
236 |
+
super(RSU4,self).__init__()
|
237 |
+
|
238 |
+
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
|
239 |
+
|
240 |
+
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
|
241 |
+
self.pool1 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
242 |
+
|
243 |
+
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
244 |
+
self.pool2 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
245 |
+
|
246 |
+
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=1)
|
247 |
+
|
248 |
+
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=2)
|
249 |
+
|
250 |
+
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
251 |
+
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
|
252 |
+
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
|
253 |
+
|
254 |
+
def forward(self,x):
|
255 |
+
|
256 |
+
hx = x
|
257 |
+
|
258 |
+
hxin = self.rebnconvin(hx)
|
259 |
+
|
260 |
+
hx1 = self.rebnconv1(hxin)
|
261 |
+
hx = self.pool1(hx1)
|
262 |
+
|
263 |
+
hx2 = self.rebnconv2(hx)
|
264 |
+
hx = self.pool2(hx2)
|
265 |
+
|
266 |
+
hx3 = self.rebnconv3(hx)
|
267 |
+
|
268 |
+
hx4 = self.rebnconv4(hx3)
|
269 |
+
|
270 |
+
hx3d = self.rebnconv3d(torch.cat((hx4,hx3),1))
|
271 |
+
hx3dup = _upsample_like(hx3d,hx2)
|
272 |
+
|
273 |
+
hx2d = self.rebnconv2d(torch.cat((hx3dup,hx2),1))
|
274 |
+
hx2dup = _upsample_like(hx2d,hx1)
|
275 |
+
|
276 |
+
hx1d = self.rebnconv1d(torch.cat((hx2dup,hx1),1))
|
277 |
+
|
278 |
+
return hx1d + hxin
|
279 |
+
|
280 |
+
### RSU-4F ###
|
281 |
+
class RSU4F(nn.Module):#UNet04FRES(nn.Module):
|
282 |
+
|
283 |
+
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
|
284 |
+
super(RSU4F,self).__init__()
|
285 |
+
|
286 |
+
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
|
287 |
+
|
288 |
+
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
|
289 |
+
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=2)
|
290 |
+
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=4)
|
291 |
+
|
292 |
+
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=8)
|
293 |
+
|
294 |
+
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=4)
|
295 |
+
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=2)
|
296 |
+
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
|
297 |
+
|
298 |
+
def forward(self,x):
|
299 |
+
|
300 |
+
hx = x
|
301 |
+
|
302 |
+
hxin = self.rebnconvin(hx)
|
303 |
+
|
304 |
+
hx1 = self.rebnconv1(hxin)
|
305 |
+
hx2 = self.rebnconv2(hx1)
|
306 |
+
hx3 = self.rebnconv3(hx2)
|
307 |
+
|
308 |
+
hx4 = self.rebnconv4(hx3)
|
309 |
+
|
310 |
+
hx3d = self.rebnconv3d(torch.cat((hx4,hx3),1))
|
311 |
+
hx2d = self.rebnconv2d(torch.cat((hx3d,hx2),1))
|
312 |
+
hx1d = self.rebnconv1d(torch.cat((hx2d,hx1),1))
|
313 |
+
|
314 |
+
return hx1d + hxin
|
315 |
+
|
316 |
+
|
317 |
+
##### U^2-Net ####
|
318 |
+
class U2NET(nn.Module):
|
319 |
+
|
320 |
+
def __init__(self,in_ch=3,out_ch=1):
|
321 |
+
super(U2NET,self).__init__()
|
322 |
+
|
323 |
+
self.stage1 = RSU7(in_ch,32,64)
|
324 |
+
self.pool12 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
325 |
+
|
326 |
+
self.stage2 = RSU6(64,32,128)
|
327 |
+
self.pool23 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
328 |
+
|
329 |
+
self.stage3 = RSU5(128,64,256)
|
330 |
+
self.pool34 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
331 |
+
|
332 |
+
self.stage4 = RSU4(256,128,512)
|
333 |
+
self.pool45 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
334 |
+
|
335 |
+
self.stage5 = RSU4F(512,256,512)
|
336 |
+
self.pool56 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
337 |
+
|
338 |
+
self.stage6 = RSU4F(512,256,512)
|
339 |
+
|
340 |
+
# decoder
|
341 |
+
self.stage5d = RSU4F(1024,256,512)
|
342 |
+
self.stage4d = RSU4(1024,128,256)
|
343 |
+
self.stage3d = RSU5(512,64,128)
|
344 |
+
self.stage2d = RSU6(256,32,64)
|
345 |
+
self.stage1d = RSU7(128,16,64)
|
346 |
+
|
347 |
+
self.side1 = nn.Conv2d(64,out_ch,3,padding=1)
|
348 |
+
self.side2 = nn.Conv2d(64,out_ch,3,padding=1)
|
349 |
+
self.side3 = nn.Conv2d(128,out_ch,3,padding=1)
|
350 |
+
self.side4 = nn.Conv2d(256,out_ch,3,padding=1)
|
351 |
+
self.side5 = nn.Conv2d(512,out_ch,3,padding=1)
|
352 |
+
self.side6 = nn.Conv2d(512,out_ch,3,padding=1)
|
353 |
+
|
354 |
+
self.outconv = nn.Conv2d(6*out_ch,out_ch,1)
|
355 |
+
|
356 |
+
def forward(self,x):
|
357 |
+
|
358 |
+
hx = x
|
359 |
+
|
360 |
+
#stage 1
|
361 |
+
hx1 = self.stage1(hx)
|
362 |
+
hx = self.pool12(hx1)
|
363 |
+
|
364 |
+
#stage 2
|
365 |
+
hx2 = self.stage2(hx)
|
366 |
+
hx = self.pool23(hx2)
|
367 |
+
|
368 |
+
#stage 3
|
369 |
+
hx3 = self.stage3(hx)
|
370 |
+
hx = self.pool34(hx3)
|
371 |
+
|
372 |
+
#stage 4
|
373 |
+
hx4 = self.stage4(hx)
|
374 |
+
hx = self.pool45(hx4)
|
375 |
+
|
376 |
+
#stage 5
|
377 |
+
hx5 = self.stage5(hx)
|
378 |
+
hx = self.pool56(hx5)
|
379 |
+
|
380 |
+
#stage 6
|
381 |
+
hx6 = self.stage6(hx)
|
382 |
+
hx6up = _upsample_like(hx6,hx5)
|
383 |
+
|
384 |
+
#-------------------- decoder --------------------
|
385 |
+
hx5d = self.stage5d(torch.cat((hx6up,hx5),1))
|
386 |
+
hx5dup = _upsample_like(hx5d,hx4)
|
387 |
+
|
388 |
+
hx4d = self.stage4d(torch.cat((hx5dup,hx4),1))
|
389 |
+
hx4dup = _upsample_like(hx4d,hx3)
|
390 |
+
|
391 |
+
hx3d = self.stage3d(torch.cat((hx4dup,hx3),1))
|
392 |
+
hx3dup = _upsample_like(hx3d,hx2)
|
393 |
+
|
394 |
+
hx2d = self.stage2d(torch.cat((hx3dup,hx2),1))
|
395 |
+
hx2dup = _upsample_like(hx2d,hx1)
|
396 |
+
|
397 |
+
hx1d = self.stage1d(torch.cat((hx2dup,hx1),1))
|
398 |
+
|
399 |
+
|
400 |
+
#side output
|
401 |
+
d1 = self.side1(hx1d)
|
402 |
+
|
403 |
+
d2 = self.side2(hx2d)
|
404 |
+
d2 = _upsample_like(d2,d1)
|
405 |
+
|
406 |
+
d3 = self.side3(hx3d)
|
407 |
+
d3 = _upsample_like(d3,d1)
|
408 |
+
|
409 |
+
d4 = self.side4(hx4d)
|
410 |
+
d4 = _upsample_like(d4,d1)
|
411 |
+
|
412 |
+
d5 = self.side5(hx5d)
|
413 |
+
d5 = _upsample_like(d5,d1)
|
414 |
+
|
415 |
+
d6 = self.side6(hx6)
|
416 |
+
d6 = _upsample_like(d6,d1)
|
417 |
+
|
418 |
+
d0 = self.outconv(torch.cat((d1,d2,d3,d4,d5,d6),1))
|
419 |
+
|
420 |
+
return F.sigmoid(d0), F.sigmoid(d1), F.sigmoid(d2), F.sigmoid(d3), F.sigmoid(d4), F.sigmoid(d5), F.sigmoid(d6)
|
421 |
+
|
422 |
+
### U^2-Net small ###
|
423 |
+
class U2NETP(nn.Module):
|
424 |
+
|
425 |
+
def __init__(self,in_ch=3,out_ch=1):
|
426 |
+
super(U2NETP,self).__init__()
|
427 |
+
|
428 |
+
self.stage1 = RSU7(in_ch,16,64)
|
429 |
+
self.pool12 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
430 |
+
|
431 |
+
self.stage2 = RSU6(64,16,64)
|
432 |
+
self.pool23 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
433 |
+
|
434 |
+
self.stage3 = RSU5(64,16,64)
|
435 |
+
self.pool34 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
436 |
+
|
437 |
+
self.stage4 = RSU4(64,16,64)
|
438 |
+
self.pool45 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
439 |
+
|
440 |
+
self.stage5 = RSU4F(64,16,64)
|
441 |
+
self.pool56 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
|
442 |
+
|
443 |
+
self.stage6 = RSU4F(64,16,64)
|
444 |
+
|
445 |
+
# decoder
|
446 |
+
self.stage5d = RSU4F(128,16,64)
|
447 |
+
self.stage4d = RSU4(128,16,64)
|
448 |
+
self.stage3d = RSU5(128,16,64)
|
449 |
+
self.stage2d = RSU6(128,16,64)
|
450 |
+
self.stage1d = RSU7(128,16,64)
|
451 |
+
|
452 |
+
self.side1 = nn.Conv2d(64,out_ch,3,padding=1)
|
453 |
+
self.side2 = nn.Conv2d(64,out_ch,3,padding=1)
|
454 |
+
self.side3 = nn.Conv2d(64,out_ch,3,padding=1)
|
455 |
+
self.side4 = nn.Conv2d(64,out_ch,3,padding=1)
|
456 |
+
self.side5 = nn.Conv2d(64,out_ch,3,padding=1)
|
457 |
+
self.side6 = nn.Conv2d(64,out_ch,3,padding=1)
|
458 |
+
|
459 |
+
self.outconv = nn.Conv2d(6*out_ch,out_ch,1)
|
460 |
+
|
461 |
+
def forward(self,x):
|
462 |
+
|
463 |
+
hx = x
|
464 |
+
|
465 |
+
#stage 1
|
466 |
+
hx1 = self.stage1(hx)
|
467 |
+
hx = self.pool12(hx1)
|
468 |
+
|
469 |
+
#stage 2
|
470 |
+
hx2 = self.stage2(hx)
|
471 |
+
hx = self.pool23(hx2)
|
472 |
+
|
473 |
+
#stage 3
|
474 |
+
hx3 = self.stage3(hx)
|
475 |
+
hx = self.pool34(hx3)
|
476 |
+
|
477 |
+
#stage 4
|
478 |
+
hx4 = self.stage4(hx)
|
479 |
+
hx = self.pool45(hx4)
|
480 |
+
|
481 |
+
#stage 5
|
482 |
+
hx5 = self.stage5(hx)
|
483 |
+
hx = self.pool56(hx5)
|
484 |
+
|
485 |
+
#stage 6
|
486 |
+
hx6 = self.stage6(hx)
|
487 |
+
hx6up = _upsample_like(hx6,hx5)
|
488 |
+
|
489 |
+
#decoder
|
490 |
+
hx5d = self.stage5d(torch.cat((hx6up,hx5),1))
|
491 |
+
hx5dup = _upsample_like(hx5d,hx4)
|
492 |
+
|
493 |
+
hx4d = self.stage4d(torch.cat((hx5dup,hx4),1))
|
494 |
+
hx4dup = _upsample_like(hx4d,hx3)
|
495 |
+
|
496 |
+
hx3d = self.stage3d(torch.cat((hx4dup,hx3),1))
|
497 |
+
hx3dup = _upsample_like(hx3d,hx2)
|
498 |
+
|
499 |
+
hx2d = self.stage2d(torch.cat((hx3dup,hx2),1))
|
500 |
+
hx2dup = _upsample_like(hx2d,hx1)
|
501 |
+
|
502 |
+
hx1d = self.stage1d(torch.cat((hx2dup,hx1),1))
|
503 |
+
|
504 |
+
|
505 |
+
#side output
|
506 |
+
d1 = self.side1(hx1d)
|
507 |
+
|
508 |
+
d2 = self.side2(hx2d)
|
509 |
+
d2 = _upsample_like(d2,d1)
|
510 |
+
|
511 |
+
d3 = self.side3(hx3d)
|
512 |
+
d3 = _upsample_like(d3,d1)
|
513 |
+
|
514 |
+
d4 = self.side4(hx4d)
|
515 |
+
d4 = _upsample_like(d4,d1)
|
516 |
+
|
517 |
+
d5 = self.side5(hx5d)
|
518 |
+
d5 = _upsample_like(d5,d1)
|
519 |
+
|
520 |
+
d6 = self.side6(hx6)
|
521 |
+
d6 = _upsample_like(d6,d1)
|
522 |
+
|
523 |
+
d0 = self.outconv(torch.cat((d1,d2,d3,d4,d5,d6),1))
|
524 |
+
|
525 |
+
return F.sigmoid(d0), F.sigmoid(d1), F.sigmoid(d2), F.sigmoid(d3), F.sigmoid(d4), F.sigmoid(d5), F.sigmoid(d6)
|