+
+
diff --git a/image/reports/report-server.service b/image/reports/report-server.service
new file mode 100644
index 0000000..1caac3d
--- /dev/null
+++ b/image/reports/report-server.service
@@ -0,0 +1,15 @@
+[Unit]
+Description=测试报告统一HTTP服务 (8850, 路径区分: /wave /v2 /v1)
+After=network-online.target
+Wants=network-online.target
+
+[Service]
+Type=simple
+User=ubuntu
+WorkingDirectory=/home/ubuntu/hair/image/reports
+ExecStart=/home/ubuntu/miniconda3/envs/my_hair/bin/python -m http.server 8850 --bind 0.0.0.0
+Restart=on-failure
+RestartSec=5
+
+[Install]
+WantedBy=multi-user.target
diff --git a/image/reports/v1 b/image/reports/v1
new file mode 120000
index 0000000..5d9e5ef
--- /dev/null
+++ b/image/reports/v1
@@ -0,0 +1 @@
+/home/ubuntu/hair/image/compare_test/out
\ No newline at end of file
diff --git a/image/reports/v2 b/image/reports/v2
new file mode 120000
index 0000000..6117c2e
--- /dev/null
+++ b/image/reports/v2
@@ -0,0 +1 @@
+/home/ubuntu/hair/image/res_test/out
\ No newline at end of file
diff --git a/image/reports/wave b/image/reports/wave
new file mode 120000
index 0000000..454eb76
--- /dev/null
+++ b/image/reports/wave
@@ -0,0 +1 @@
+/home/ubuntu/hair/image/wave_test/out
\ No newline at end of file
diff --git a/image/wave_test/batch_test.py b/image/wave_test/batch_test.py
new file mode 100644
index 0000000..f50a918
--- /dev/null
+++ b/image/wave_test/batch_test.py
@@ -0,0 +1,143 @@
+#!/usr/bin/env python3
+"""接口2 female wave发型 分辨率对比测试 v3: 21图 × 5档 = 105次。
+串行执行，记录耗时与成败，结果图按 图_档位 命名保存。支持断点续跑。
+"""
+import base64, csv, json, os, time, io
+import requests
+from PIL import Image
+
+API = "http://127.0.0.1:8187/api/v1/hair/grow"
+TOKEN = "dev-shared-secret-2026"
+TIMEOUT = 600
+OUT = "/home/ubuntu/hair/image/wave_test/out"
+PROGRESS = "/home/ubuntu/hair/image/wave_test/progress.json"
+
+IMG_DIR = "/home/ubuntu/hair/image"
+# 21张图：19张girl_img + asdf + qwer
+IMAGES = []
+for f in sorted(os.listdir(os.path.join(IMG_DIR, "girl_img"))):
+ if f.lower().endswith((".jpg", ".jpeg", ".png")):
+ IMAGES.append((os.path.splitext(f)[0], os.path.join(IMG_DIR, "girl_img", f)))
+IMAGES.append(("asdf", os.path.join(IMG_DIR, "asdf.jpg")))
+IMAGES.append(("qwer", os.path.join(IMG_DIR, "qwer.jpg")))
+
+# wave = female hair_style 5
+STYLE_IDX = 5
+# 5档: 原图(0) / 1024 / 896 / 768 / 640
+SIDES = [
+ ("origin", 0),
+ ("s1024", 1024),
+ ("s896", 896),
+ ("s768", 768),
+ ("s640", 640),
+]
+
+def load_progress():
+ if os.path.exists(PROGRESS):
+ try:
+ return json.load(open(PROGRESS))
+ except Exception:
+ pass
+ return {"done": [], "results": []}
+
+def save_progress(prog):
+ json.dump(prog, open(PROGRESS, "w"), ensure_ascii=False, indent=1)
+
+def run_one(img_name, img_path, side_name, side_val):
+ key = f"{img_name}_{side_name}"
+ with open(img_path, "rb") as f:
+ img_b64 = base64.b64encode(f.read()).decode()
+ data = {
+ "image_base64": "data:image/jpeg;base64," + img_b64,
+ "gender": "female",
+ "hair_style": str(STYLE_IDX),
+ "redraw_max_side": str(side_val),
+ }
+ t0 = time.time()
+ rec = {"key": key, "img": img_name, "side": side_name, "side_val": side_val,
+ "ok": False, "elapsed": 0.0, "err": "", "out_w": 0, "out_h": 0, "bytes": 0}
+ try:
+ r = requests.post(API, data=data, headers={"X-Internal-Token": TOKEN}, timeout=TIMEOUT)
+ rec["elapsed"] = round(time.time() - t0, 1)
+ d = r.json()
+ if d.get("code") != 0:
+ rec["err"] = f"code={d.get('code')} {d.get('message','')}"[:200]
+ return rec
+ results = (d.get("data") or {}).get("results") or []
+ if not results:
+ rec["err"] = "空结果"
+ return rec
+ it = results[0]
+ grown = it.get("grown_image_base64")
+ if not grown:
+ rec["err"] = "无生发图(重绘失败/OOM?)"
+ return rec
+ raw = base64.b64decode(grown)
+ im = Image.open(io.BytesIO(raw))
+ rec["out_w"], rec["out_h"] = im.size
+ out_path = f"{OUT}/{key}.jpg"
+ with open(out_path, "wb") as fo:
+ fo.write(raw)
+ rec["ok"] = True
+ rec["bytes"] = len(raw)
+ except requests.exceptions.Timeout:
+ rec["elapsed"] = round(time.time() - t0, 1)
+ rec["err"] = f"超时(>{TIMEOUT}s)"
+ except Exception as e:
+ rec["elapsed"] = round(time.time() - t0, 1)
+ rec["err"] = f"{type(e).__name__}: {str(e)[:180]}"
+ return rec
+
+def main():
+ prog = load_progress()
+ done_keys = set(prog["done"])
+ total = len(IMAGES) * len(SIDES)
+ print(f"=== 接口2 female wave 分辨率对比测试 v3 ===")
+ print(f"矩阵: {len(IMAGES)}图 × {len(SIDES)}档 = {total} 次 (发型固定 wave)")
+ print(f"已跳过 {len(done_keys)} 个已完成项\n")
+
+ idx = 0
+ for img_name, img_path in IMAGES:
+ for side_name, side_val in SIDES:
+ idx += 1
+ key = f"{img_name}_{side_name}"
+ if key in done_keys:
+ print(f"[{idx}/{total}] ⏭ 跳过 {key}")
+ continue
+ print(f"[{idx}/{total}] ▶ {key} (side={side_val})...", end=" ", flush=True)
+ rec = run_one(img_name, img_path, side_name, side_val)
+ prog["results"].append(rec)
+ prog["done"].append(key)
+ save_progress(prog)
+ if rec["ok"]:
+ print(f"✅ {rec['elapsed']}s {rec['out_w']}x{rec['out_h']} ({rec['bytes']}B)")
+ else:
+ print(f"❌ {rec['elapsed']}s {rec['err']}")
+ time.sleep(2)
+
+ print("\n" + "=" * 70)
+ print("汇总")
+ print("=" * 70)
+ write_report(prog["results"])
+ print(f"\n结果图: {OUT}/")
+ print(f"CSV: {OUT}/report.csv JSON: {OUT}/report.json")
+
+def write_report(results):
+ with open(f"{OUT}/report.csv", "w", newline="") as f:
+ w = csv.writer(f)
+ w.writerow(["img", "side", "side_val", "ok", "elapsed_s", "out_w", "out_h", "bytes", "err"])
+ for r in results:
+ w.writerow([r["img"], r["side"], r["side_val"], r["ok"],
+ r["elapsed"], r["out_w"], r["out_h"], r["bytes"], r["err"]])
+ json.dump(results, open(f"{OUT}/report.json", "w"), ensure_ascii=False, indent=1)
+
+ # 各档平均耗时
+ print("\n--- 各档平均耗时（秒）---")
+ for side_name, side_val in SIDES:
+ ts = [r["elapsed"] for r in results if r["side"] == side_name and r["ok"]]
+ if ts:
+ avg = sum(ts) / len(ts)
+ print(f" {side_name:8s} (={side_val:>4}): 平均 {avg:5.1f}s [{min(ts):.1f}~{max(ts):.1f}] 成功 {len(ts)}")
+
+if __name__ == "__main__":
+ main()
diff --git a/image/wave_test/gen_report.py b/image/wave_test/gen_report.py
new file mode 100644
index 0000000..4a727ae
--- /dev/null
+++ b/image/wave_test/gen_report.py
@@ -0,0 +1,176 @@
+#!/usr/bin/env python3
+"""生成 v3 wave发型 分辨率对比报告：每行=原图+5档，21行；含各档耗时统计。"""
+import json, os, html
+
+OUT = "/home/ubuntu/hair/image/wave_test/out"
+REPORT = os.path.join(OUT, "report.html")
+results = json.load(open(os.path.join(OUT, "report.json")))
+
+SIDES = [("origin", "原图直送", 0), ("s1024", "1024", 1024),
+ ("s896", "896", 896), ("s768", "768", 768), ("s640", "640", 640)]
+
+# 图片顺序（与测试脚本一致）：girl_img 排序 + asdf + qwer
+IMG_DIR = "/home/ubuntu/hair/image"
+IMAGES = []
+for f in sorted(os.listdir(os.path.join(IMG_DIR, "girl_img"))):
+ if f.lower().endswith((".jpg", ".jpeg", ".png")):
+ IMAGES.append(os.path.splitext(f)[0])
+IMAGES.append("asdf")
+IMAGES.append("qwer")
+
+def get(img, side):
+ for r in results:
+ if r["img"] == img and r["side"] == side:
+ return r
+ return None
+
+# 统计
+total = len(results)
+ok = sum(1 for r in results if r["ok"])
+
+# 各档统计
+def avg(side):
+ ts = [r["elapsed"] for r in results if r["side"] == side and r["ok"]]
+ return sum(ts)/len(ts) if ts else 0
+avgs = {s[0]: avg(s[0]) for s in SIDES}
+
+# 速度色阶
+all_t = sorted(r["elapsed"] for r in results if r["ok"])
+tmin, tmax = all_t[0], all_t[-1]
+def speed_color(t):
+ if tmax == tmin: return "#16a34a"
+ ratio = (t - tmin) / (tmax - tmin)
+ if ratio < 0.33: return "#16a34a"
+ elif ratio < 0.66: return "#f59e0b"
+ else: return "#dc2626"
+
+# 表格行
+rows_html = []
+for img in IMAGES:
+ orig_cell = (f'
'
+ f'
'
+ f'
📷 原图
')
+ side_cells = []
+ for side_name, side_lbl, side_val in SIDES:
+ r = get(img, side_name)
+ if r and r["ok"]:
+ col = speed_color(r["elapsed"])
+ cell = (f'
