megadetector 5.0.6__py3-none-any.whl → 5.0.7__py3-none-any.whl

data_management/coco_to_yolo.py

@@ -89,10 +89,12 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
                  images_to_exclude=None,
                  path_replacement_char='#',
                  category_names_to_exclude=None,
-                 write_output=True):
+                 category_names_to_include=None,
+                 write_output=True,
+                 flatten_paths=True):
     """
-    Convert a COCO-formatted dataset to a YOLO-formatted dataset, flattening the dataset
-    (to a single folder) in the process.
+    Convert a COCO-formatted dataset to a YOLO-formatted dataset, optionally flattening the 
+    dataset to a single folder in the process.
     
     If the input and output folders are the same, writes .txt files to the input folder,
     and neither moves nor modifies images.
@@ -130,6 +132,9 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
         
     ## Validate input
     
+    if category_names_to_include is not None and category_names_to_exclude is not None:
+        raise ValueError('category_names_to_include and category_names_to_exclude are mutually exclusive')
+        
     if output_folder is None:
         output_folder = input_image_folder
     
@@ -138,12 +143,16 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
             
     if category_names_to_exclude is None:
         category_names_to_exclude = {}
-        
+            
     assert os.path.isdir(input_image_folder)
     assert os.path.isfile(input_file)
     os.makedirs(output_folder,exist_ok=True)
     
-    
+    if (output_folder == input_image_folder) and (overwrite_images) and \
+        (not create_image_and_label_folders) and (not flatten_paths):
+            print('Warning: output folder and input folder are the same, disabling overwrite_images')
+            overwrite_images = False
+            
     ## Read input data
     
     with open(input_file,'r') as f:
@@ -184,7 +193,11 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
     
     for category in data['categories']:
         coco_id_to_name[category['id']] = category['name']
-        if (category['name'] in category_names_to_exclude):
+        if (category_names_to_include is not None) and \
+            (category['name'] not in category_names_to_include):
+            coco_category_ids_to_exclude.add(category['id'])
+            continue
+        elif (category['name'] in category_names_to_exclude):
             coco_category_ids_to_exclude.add(category['id'])
             continue               
         assert category['id'] not in coco_id_to_yolo_id
@@ -228,9 +241,13 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
         tokens = os.path.splitext(im['file_name'])
         if tokens[1].lower() not in typical_image_extensions:
             print('Warning: unusual image file name {}'.format(im['file_name']))
-                  
-        image_name = tokens[0].replace('\\','/').replace('/',path_replacement_char) + \
-            '_' + str(i_image).zfill(6)
+                
+        if flatten_paths:
+            image_name = tokens[0].replace('\\','/').replace('/',path_replacement_char) + \
+                '_' + str(i_image).zfill(6)            
+        else:
+            image_name = tokens[0]
+
         assert image_name not in image_names, 'Image name collision for {}'.format(image_name)
         image_names.add(image_name)
         
@@ -293,12 +310,6 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
                 # This category isn't in our category list.  This typically corresponds to whole sets
                 # of images that were excluded from the YOLO set.
                 if ann['category_id'] in coco_category_ids_to_exclude:
-                    category_name = coco_id_to_name[ann['category_id']]
-                    if category_name not in category_exclusion_warnings_printed:
-                        category_exclusion_warnings_printed.add(category_name)
-                        print('Warning: ignoring category {} in image {}'.format(
-                            category_name,image_id),end='')
-                        print('...are you sure you didn\'t mean to exclude this image?')                        
                     continue
                 
                 yolo_category_id = coco_id_to_yolo_id[ann['category_id']]
@@ -407,30 +418,38 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
         with open(image_id_to_output_image_json_file,'w') as f:
             json.dump(image_id_to_output_image_name,f,indent=1)
     
-    if (write_output):
-    
+
+    if (output_folder == input_image_folder) and (not create_image_and_label_folders):
+        print('Creating annotation files (not copying images, input and output folder are the same)')
+    else:
         print('Copying images and creating annotation files')
+
+    if create_image_and_label_folders:
+        dest_image_folder = os.path.join(output_folder,'images')
+        dest_txt_folder = os.path.join(output_folder,'labels')
+    else:
+        dest_image_folder = output_folder
+        dest_txt_folder = output_folder
+        
+    source_image_to_dest_image = {}
     
-        if create_image_and_label_folders:
-            dest_image_folder = os.path.join(output_folder,'images')
-            dest_txt_folder = os.path.join(output_folder,'labels')
-        else:
-            dest_image_folder = output_folder
-            dest_txt_folder = output_folder
-            
-        # TODO: parallelize this loop
-        #
-        # output_info = images_to_copy[0]
-        for output_info in tqdm(images_to_copy):
-    
-            source_image = output_info['source_image']
-            dest_image_relative = output_info['dest_image_relative']
-            dest_txt_relative = output_info['dest_txt_relative']
-            
-            dest_image = os.path.join(dest_image_folder,dest_image_relative)
-            os.makedirs(os.path.dirname(dest_image),exist_ok=True)
+    # TODO: parallelize this loop
+    #
+    # output_info = images_to_copy[0]
+    for output_info in tqdm(images_to_copy):
+
+        source_image = output_info['source_image']
+        dest_image_relative = output_info['dest_image_relative']
+        dest_txt_relative = output_info['dest_txt_relative']
+        
+        dest_image = os.path.join(dest_image_folder,dest_image_relative)
+        dest_txt = os.path.join(dest_txt_folder,dest_txt_relative)
+        
+        source_image_to_dest_image[source_image] = dest_image
+        
+        if write_output:
             
-            dest_txt = os.path.join(dest_txt_folder,dest_txt_relative)
+            os.makedirs(os.path.dirname(dest_image),exist_ok=True)        
             os.makedirs(os.path.dirname(dest_txt),exist_ok=True)
             
             if not create_image_and_label_folders:
@@ -438,7 +457,7 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
             
             if (not os.path.isfile(dest_image)) or (overwrite_images):
                 shutil.copyfile(source_image,dest_image)
-            
+        
             bboxes = output_info['bboxes']        
             
             # Only write an annotation file if there are bounding boxes.  Images with 
@@ -458,15 +477,17 @@ def coco_to_yolo(input_image_folder,output_folder,input_file,
                         assert len(bbox) == 5
                         s = '{} {} {} {} {}'.format(bbox[0],bbox[1],bbox[2],bbox[3],bbox[4])
                         f.write(s + '\n')
-                
-        # ...for each image
+                        
+        # ...if we're actually writing output
         
-    # ...if we're actually writing output
-
-    return_info = {}
-    return_info['class_list_filename'] = class_list_filename
+    # ...for each image
+    
+    coco_to_yolo_info = {}
+    coco_to_yolo_info['class_list_filename'] = class_list_filename
+    coco_to_yolo_info['source_image_to_dest_image'] = source_image_to_dest_image
+    coco_to_yolo_info['coco_id_to_yolo_id'] = coco_id_to_yolo_id
     
-    return return_info
+    return coco_to_yolo_info
 
 # ...def coco_to_yolo(...)
 

data_management/databases/integrity_check_json_db.py

@@ -89,6 +89,9 @@ def integrity_check_json_db(jsonFile, options=None):
         
     print(options.__dict__)
     
+    if options.baseDir is None:
+        options.baseDir = ''
+        
     baseDir = options.baseDir
     
     
@@ -116,10 +119,9 @@ def integrity_check_json_db(jsonFile, options=None):
     annotations = data['annotations']
     categories = data['categories']
     # info = data['info']
-    assert 'info' in data
+    assert 'info' in data, 'No info struct in database'
     
-    if len(baseDir) > 0:
-        
+    if len(baseDir) > 0:        
         assert os.path.isdir(baseDir), 'Base directory {} does not exist'.format(baseDir)
         
         
@@ -139,8 +141,8 @@ def integrity_check_json_db(jsonFile, options=None):
         assert 'name' in cat
         assert 'id' in cat
         
-        assert isinstance(cat['id'],int), 'Illegal category ID type'
-        assert isinstance(cat['name'],str), 'Illegal category name type'
+        assert isinstance(cat['id'],int), 'Illegal category ID type: [{}]'.format(str(cat['id']))
+        assert isinstance(cat['name'],str), 'Illegal category name type [{}]'.format(str(cat['name']))
         
         catId = cat['id']
         catName = cat['name']

data_management/generate_crops_from_cct.py

@@ -164,4 +164,4 @@ if False:
     options.box_expansion = 25
     
     htmlOutputFile,db = visualize_db(db_path,output_dir,image_base_dir,options)
-    
+    

data_management/importers/animl_results_to_md_results.py

@@ -145,7 +145,7 @@ def main():
         '--output_file',
         type=str,
         default=None,
-        help='output .json file (defaults to input file appened with ".json")')
+        help='output .json file (defaults to input file appended with ".json")')
     
     if len(sys.argv[1:]) == 0:
         parser.print_help()
@@ -157,4 +157,4 @@ def main():
     
 if __name__ == '__main__':
     main()
-    
+    

data_management/importers/noaa_seals_2019.py

@@ -2,7 +2,7 @@
 #
 # noaa_seals_2019.py
 #
-# Prepare the NOAA Artic Seals 2019 metadata for LILA.
+# Prepare the NOAA Arctic Seals 2019 metadata for LILA.
 #
 ########
 

data_management/importers/zamba_results_to_md_results.py

@@ -166,7 +166,7 @@ def main():
         '--output_file',
         type=str,
         default=None,
-        help='output .json file (defaults to input file appened with ".json")')
+        help='output .json file (defaults to input file appended with ".json")')
     
     if len(sys.argv[1:]) == 0:
         parser.print_help()
@@ -178,4 +178,4 @@ def main():
     
 if __name__ == '__main__':
     main()
-    
+    

data_management/labelme_to_coco.py

@@ -20,11 +20,19 @@ from tqdm import tqdm
 
 #%% Functions
 
-def labelme_to_coco(input_folder,output_file=None,category_id_to_category_name=None,
-                    empty_category_name='empty',empty_category_id=None,info_struct=None,
-                    relative_paths_to_include=None,relative_paths_to_exclude=None,
-                    use_folders_as_labels=False,recursive=True,no_json_handling='skip',
-                    validate_image_sizes=True):
+def labelme_to_coco(input_folder,
+                    output_file=None,
+                    category_id_to_category_name=None,
+                    empty_category_name='empty',
+                    empty_category_id=None,
+                    info_struct=None,
+                    relative_paths_to_include=None,
+                    relative_paths_to_exclude=None,
+                    use_folders_as_labels=False,
+                    recursive=True,
+                    no_json_handling='skip',
+                    validate_image_sizes=True,
+                    right_edge_quantization_threshold=None):
     """
     Find all images in [input_folder] that have corresponding .json files, and convert
     to a COCO .json file.
@@ -48,7 +56,13 @@ def labelme_to_coco(input_folder,output_file=None,category_id_to_category_name=N
         
     * 'skip': ignore image files with no corresponding .json files
     * 'empty': treat image files with no corresponding .json files as empty
-    * 'error': throw an error when an image file has no corresponding .json file 
+    * 'error': throw an error when an image file has no corresponding .json file
+    
+    right_edge_quantization_threshold is an off-by-default hack to handle cases where 
+    boxes that really should be running off the right side of the image only extend like 99%
+    of the way there, due to what appears to be a slight bias inherent to MD.  If a box extends
+    within [right_edge_quantization_threshold] (a small number, from 0 to 1, but probably around 
+    0.02) of the right edge of the image, it will be extended to the far right edge.    
     """
     
     if category_id_to_category_name is None:
@@ -86,6 +100,8 @@ def labelme_to_coco(input_folder,output_file=None,category_id_to_category_name=N
     images = []
     annotations = []
     
+    n_edges_quantized = 0
+    
     # image_fn_relative = image_filenames_relative[0]
     for image_fn_relative in tqdm(image_filenames_relative):
         
@@ -182,6 +198,14 @@ def labelme_to_coco(input_folder,output_file=None,category_id_to_category_name=N
                 x1 = max(p0[0],p1[0])
                 y0 = min(p0[1],p1[1])
                 y1 = max(p0[1],p1[1])
+                
+                if right_edge_quantization_threshold is not None:                    
+                    x1_rel = x1 / (im['width'] - 1)
+                    right_edge_distance = 1.0 - x1_rel
+                    if right_edge_distance < right_edge_quantization_threshold:
+                        n_edges_quantized += 1
+                        x1 = im['width'] - 1
+                        
                 bbox = [x0,y0,abs(x1-x0),abs(y1-y0)]
                 ann = {}
                 ann['id'] = str(uuid.uuid1())
@@ -197,6 +221,10 @@ def labelme_to_coco(input_folder,output_file=None,category_id_to_category_name=N
                   
     # ..for each image                
     
+    if n_edges_quantized > 0:
+        print('Quantized the right edge in {} of {} images'.format(
+            n_edges_quantized,len(image_filenames_relative)))
+        
     output_dict = {}
     output_dict['images'] = images
     output_dict['annotations'] = annotations

data_management/labelme_to_yolo.py

@@ -226,4 +226,4 @@ if False:
 
 #%% Command-line driver
 
-# TODO
+# TODO