handle checks.

Author: PritamTheCoder

.gitignore

@@ -10,4 +10,5 @@ logs
 logs_extreme
 htmlcov
 coverage.xml
-Makefile
+Makefile
+.coverage

src/yolo_augmentor/aug/augment_dataset.py

@@ -106,7 +106,7 @@ def _load_config(self, config_path: str) -> Dict:
     def _setup_logging(self):
         """Setup comprehensive logging for debugging."""
         log_dir = Path(self.config['dataset'].get('logs_dir', 'logs'))
-        log_dir.mkdir(exist_ok=True)
+        log_dir.mkdir(parents=True, exist_ok=True)
         log_file = log_dir / f"augmentation_{datetime.now():%Y%m%d_%H%M%S}.log"
 
         logging.basicConfig(

src/yolo_augmentor/data/repair_labels.py

@@ -36,8 +36,12 @@ def _clean_line(line: str):
     x, y = max(0, min(1, x)), max(0, min(1, y))
     w, h = max(0, min(1, w)), max(0, min(1, h))
 
-    # invalid bounding boxes
-    if w <= 0 or h <= 0:
+    # Coordinates must be strictly inside (0,1)
+    if not (0 < x < 1 and 0 < y < 1):
+        return None
+
+    # width/height must be positive AND ensure box doesn't exceed bounds
+    if not (0 < w < 1 and 0 < h < 1):
         return None
 
     return cls, x, y, w, h

src/yolo_augmentor/data/scan_dataset.py

@@ -37,15 +37,36 @@ def scan_dataset(dataset_path: str) -> Dict:
     labels = find_files(root, [".txt"])
     label_map = {f.stem: f for f in labels}
 
+    # Build a function to check if label file has valid YOLO lines
+    def has_valid_label(file: Path) -> bool:
+        try:
+            lines = file.read_text().strip().split("\n")
+        except:
+            return False
+
+        for line in lines:
+            parts = line.strip().split()
+            if len(parts) == 5:
+                try:
+                    float(parts[1]); float(parts[2])
+                    float(parts[3]); float(parts[4])
+                    return True
+                except ValueError:
+                    continue
+        return False
+
+    # NEW: count only files with valid labels
+    valid_label_files = [f for f in labels if has_valid_label(f)]
+
     missing_labels = [img for img in images if img.stem not in label_map]
     image_dirs = {img.parent for img in images}
-    label_dirs = {lbl.parent for lbl in labels}
+    label_dirs = {lbl.parent for lbl in valid_label_files}
 
     result = {
         "total_images": len(images),
-        "total_labels": len(labels),
+        "total_labels": len(valid_label_files),     # <-- FIXED
         "missing_pairs": len(missing_labels),
-        "missing_label_files": [str(i) for i in missing_labels[:20]],  # preview
+        "missing_label_files": [str(i) for i in missing_labels[:20]],
         "structure_type": classify_structure(image_dirs, label_dirs),
         "image_dirs": list(map(str, image_dirs)),
         "label_dirs": list(map(str, label_dirs)),
@@ -55,6 +76,7 @@ def scan_dataset(dataset_path: str) -> Dict:
     return result
 
 
+
 if __name__ == "__main__":
     import pprint
     path = input("Enter dataset directory: ")