RoiSet facilitates object detection models

• created distinct generator class methods for RoiSet
• RoiSet.from_bounding_boxes can create RoiSet with just YXWH coordinates and no segmentation, e.g. as output from object detection models
• can also inspect and quantify overlap between and within RoiSet(s), at both bounding box and segmentation level
• can generate RoiSet from polygon segmentation e.g. in COCO segmentation models

model_server/base/roiset.py

@@ -775,10 +775,17 @@ class RoiSet(object):
     def acc_obj_ids(self):
         return make_object_ids_from_df(self._df, self.acc_raw.shape_dict)
 
-    # TODO: add docstring
     # TODO: make this work with obj det dataset
     @staticmethod
     def deserialize(acc_raw: GenericImageDataAccessor, where: Path, prefix=''):
+        """
+        Create an RoiSet object from saved files and an image accessor
+        :param acc_raw: accessor to image that contains ROIs
+        :param where: path to directory containing RoiSet serialization files, namely dataframe.csv and a subdirectory
+            named tight_patch_masks
+        :param prefix: starting prefix of patch mask filenames
+        :return:
+        """
         df = pd.read_csv(where / 'dataframe' / (prefix + '.csv'))[['label', 'zi', 'y0', 'y1', 'x0', 'x1']]
 
         def _read_binary_mask(r):