Avoid unnecessary image loading in Datumaro format on save (cvat-ai#176)

* Avoid unnecessary image loading in Datumaro format

* update changelog

* add test

CHANGELOG.md

@@ -25,6 +25,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Fixed
 - The ability to work with file names containing Cyrillic and spaces (<https://github.com/openvinotoolkit/datumaro/pull/148>)
 - Image reading and saving in ICDAR formats (<https://github.com/openvinotoolkit/datumaro/pull/174>)
+- Unnecessary image loading on dataset saving (<https://github.com/openvinotoolkit/datumaro/pull/176>)
 
 ### Security
 -

datumaro/plugins/datumaro_format/converter.py

@@ -62,9 +62,10 @@ def write_item(self, item):
                 self._context._save_image(item, path)
 
             item_desc['image'] = {
-                'size': item.image.size,
                 'path': path,
             }
+            if item.image.has_size: # avoid occasional loading
+                item_desc['image']['size'] = item.image.size
         self.items.append(item_desc)
 
         for ann in item.annotations:

datumaro/util/image.py

@@ -250,6 +250,9 @@ def __init__(self, data=None, path=None, loader=None, cache=None,
                 data = lazy_image(path, loader=loader, cache=cache)
         self._data = data
 
+        if not self._size and isinstance(data, np.ndarray):
+            self._size = data.shape[:2]
+
     @property
     def path(self):
         return self._path
@@ -261,13 +264,22 @@ def ext(self):
     @property
     def data(self):
         if callable(self._data):
-            return self._data()
-        return self._data
+            data = self._data()
+        else:
+            data = self._data
+
+        if self._size is None and data is not None:
+            self._size = data.shape[:2]
+        return data
 
     @property
     def has_data(self):
         return self._data is not None
 
+    @property
+    def has_size(self):
+        return self._size is not None or isinstance(self._data, np.ndarray)
+
     @property
     def size(self):
         if self._size is None:

tests/test_dataset.py

@@ -564,6 +564,25 @@ def test_flushes_changes_on_save(self):
 
             self.assertFalse(dataset.is_modified)
 
+    def test_does_not_load_images_on_saving(self):
+        # Issue https://github.com/openvinotoolkit/datumaro/issues/177
+        # Missing image metadata (size etc.) can lead to image loading on
+        # dataset save without image saving
+
+        called = False
+        def test_loader():
+            nonlocal called
+            called = True
+
+        dataset = Dataset.from_iterable([
+            DatasetItem(1, image=test_loader)
+        ])
+
+        with TestDir() as test_dir:
+            dataset.save(test_dir)
+
+        self.assertFalse(called)
+
 
 class DatasetItemTest(TestCase):
     def test_ctor_requires_id(self):