Added bubble split

regenerate_debug.py

@@ -0,0 +1,56 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Regenerate debug_clusters.png with the new split bubbles.json
+"""
+
+import json
+import cv2
+import numpy as np
+
+def quad_bbox(quad):
+    """Convert quad to bounding box"""
+    xs = [p[0] for p in quad]
+    ys = [p[1] for p in quad]
+    return (min(xs), min(ys), max(xs), max(ys))
+
+def save_debug_clusters_from_json(
+    image_path="004.png",
+    bubbles_path="bubbles.json",
+    out_path="debug_clusters.png"
+):
+    img = cv2.imread(image_path)
+    if img is None:
+        print(f"❌ Cannot load image: {image_path}")
+        return
+    
+    # Load bubbles.json
+    with open(bubbles_path, "r", encoding="utf-8") as f:
+        bubbles_data = json.load(f)
+    
+    # Draw all quad polygons in white (erasing original text)
+    for bid_str, bubble_info in bubbles_data.items():
+        for quad in bubble_info.get("quads", []):
+            pts = np.array(quad, dtype=np.int32)
+            cv2.fillPoly(img, [pts], (255, 255, 255))
+            cv2.polylines(img, [pts], True, (180, 180, 180), 1)
+    
+    # Draw bounding boxes with labels
+    for bid_str, bubble_info in bubbles_data.items():
+        bid = int(bid_str)
+        x = bubble_info["x"]
+        y = bubble_info["y"]
+        w = bubble_info["w"]
+        h = bubble_info["h"]
+        x2 = x + w
+        y2 = y + h
+        
+        cv2.rectangle(img, (x, y), (x2, y2), (0, 220, 0), 2)
+        cv2.putText(img, f"BOX#{bid}", (x + 2, max(15, y + 16)),
+                    cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 220, 0), 2)
+    
+    cv2.imwrite(out_path, img)
+    print(f"✅ Saved: {out_path}")
+
+if __name__ == "__main__":
+    save_debug_clusters_from_json()