Merge pull request #476 from microsoft/PreRelease

Adding Sink overwrite argument for image saving

PW_FT_classification/src/utils/utils.py

@@ -5,7 +5,7 @@
 from PIL import Image 
 import numpy as np
 
-def save_crop_images(results, output_dir, original_csv_path):
+def save_crop_images(results, output_dir, original_csv_path, overwrite=False):
     """
     Save cropped images based on the detection bounding boxes.
 
@@ -16,6 +16,8 @@ def save_crop_images(results, output_dir, original_csv_path):
             Directory to save the cropped images.
         original_csv_path (str):
             Path to the original CSV file.
+        overwrite (bool):
+            Whether overwriting existing image folders. Default to False.
     Return:
         new_csv_path (str):
             Path to the new CSV file.
@@ -29,7 +31,7 @@ def save_crop_images(results, output_dir, original_csv_path):
     new_records = []
 
     os.makedirs(output_dir, exist_ok=True)
-    with sv.ImageSink(target_dir_path=output_dir, overwrite=False) as sink:
+    with sv.ImageSink(target_dir_path=output_dir, overwrite=overwrite) as sink:
         for entry in results:
             # Process the data if the name of the file is in the dataframe
             if os.path.basename(entry["img_id"]) in original_df['path'].values:

PytorchWildlife/utils/post_process.py

@@ -21,7 +21,7 @@
 
 
 # !!! Output paths need to be optimized !!!
-def save_detection_images(results, output_dir):
+def save_detection_images(results, output_dir, overwrite=False):
     """
     Save detected images with bounding boxes and labels annotated.
 
@@ -30,11 +30,13 @@ def save_detection_images(results, output_dir):
             Detection results containing image ID, detections, and labels.
         output_dir (str):
             Directory to save the annotated images.
+        overwrite (bool):
+            Whether overwriting existing image folders. Default to False.
     """
     box_annotator = sv.BoxAnnotator(thickness=4, text_thickness=4, text_scale=2)
     os.makedirs(output_dir, exist_ok=True)
 
-    with sv.ImageSink(target_dir_path=output_dir, overwrite=True) as sink:
+    with sv.ImageSink(target_dir_path=output_dir, overwrite=overwrite) as sink:
         if isinstance(results, list):
             for entry in results:
                 annotated_img = box_annotator.annotate(
@@ -57,7 +59,7 @@ def save_detection_images(results, output_dir):
 
 
 # !!! Output paths need to be optimized !!!
-def save_crop_images(results, output_dir):
+def save_crop_images(results, output_dir, overwrite=False):
     """
     Save cropped images based on the detection bounding boxes.
 
@@ -66,10 +68,12 @@ def save_crop_images(results, output_dir):
             Detection results containing image ID and detections.
         output_dir (str):
             Directory to save the cropped images.
+        overwrite (bool):
+            Whether overwriting existing image folders. Default to False.
     """
     assert isinstance(results, list)
     os.makedirs(output_dir, exist_ok=True)
-    with sv.ImageSink(target_dir_path=output_dir, overwrite=True) as sink:
+    with sv.ImageSink(target_dir_path=output_dir, overwrite=overwrite) as sink:
         for entry in results:
             for i, (xyxy, _, _, cat, _) in enumerate(entry["detections"]):
                 cropped_img = sv.crop_image(

demo/image_demo.py

@@ -43,7 +43,7 @@
 results = detection_model.single_image_detection(transform(img), img.shape, tgt_img_path)
 
 # Saving the detection results 
-pw_utils.save_detection_images(results, os.path.join(".","demo_output"))
+pw_utils.save_detection_images(results, os.path.join(".","demo_output"), overwrite=False)
 
 #%% Batch detection
 """ Batch-detection demo """
@@ -68,11 +68,11 @@
 
 #%% Output to annotated images
 # Saving the batch detection results as annotated images
-pw_utils.save_detection_images(results, "batch_output")
+pw_utils.save_detection_images(results, "batch_output", overwrite=False)
 
 #%% Output to cropped images
 # Saving the detected objects as cropped images
-pw_utils.save_crop_images(results, "crop_output")
+pw_utils.save_crop_images(results, "crop_output", overwrite=False)
 
 #%% Output to JSON results
 # Saving the detection results in JSON format

demo/image_detection_colabdemo.ipynb

@@ -925,7 +925,7 @@
         "transform = pw_trans.MegaDetector_v5_Transform(target_size=detection_model.IMAGE_SIZE,\n",
         "                                               stride=detection_model.STRIDE)\n",
         "results = detection_model.single_image_detection(transform(img), img.shape, temp_file_path)\n",
-        "pw_utils.save_detection_images(results, \"./demo_output\")"
+        "pw_utils.save_detection_images(results, \"./demo_output\", overwrite=False)"
       ]
     },
     {
@@ -997,7 +997,7 @@
       },
       "outputs": [],
       "source": [
-        "pw_utils.save_detection_images(results, \"batch_output\")"
+        "pw_utils.save_detection_images(results, \"batch_output\", overwrite=False)"
       ]
     },
     {
@@ -1020,7 +1020,7 @@
       },
       "outputs": [],
       "source": [
-        "pw_utils.save_crop_images(results, \"crop_output\")"
+        "pw_utils.save_crop_images(results, \"crop_output\", overwrite=False)"
       ]
     },
     {

demo/image_detection_demo.ipynb

@@ -88,7 +88,7 @@
     "transform = pw_trans.MegaDetector_v5_Transform(target_size=detection_model.IMAGE_SIZE,\n",
     "                                               stride=detection_model.STRIDE)\n",
     "results = detection_model.single_image_detection(transform(img), img.shape, tgt_img_path)\n",
-    "pw_utils.save_detection_images(results, os.path.join(\".\",\"demo_output\"))"
+    "pw_utils.save_detection_images(results, os.path.join(\".\",\"demo_output\"), overwrite=False)"
    ]
   },
   {
@@ -138,7 +138,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "pw_utils.save_detection_images(results, \"batch_output\")"
+    "pw_utils.save_detection_images(results, \"batch_output\", overwrite=False)"
    ]
   },
   {
@@ -157,7 +157,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "pw_utils.save_crop_images(results, \"crop_output\")"
+    "pw_utils.save_crop_images(results, \"crop_output\", overwrite=False)"
    ]
   },
   {

setup.py

@@ -4,7 +4,7 @@
         long_description = file.read()
 setup(
     name='PytorchWildlife',
-    version='1.0.2.11', 
+    version='1.0.2.12', 
     packages=find_packages(),
     url='https://github.com/microsoft/CameraTraps/',  
     license='MIT',

version.txt

@@ -1 +1 @@
-1.0.2.11
+1.0.2.12