undo process_data change

nerfstudio/process_data/colmap_utils.py

@@ -13,12 +13,12 @@
 # limitations under the License.
 
 """
-Tools supporting the execution of COLMAP and preparation of COLMAP-based datasets for nerfstudio training.
+Tools supporting the execution of COLMAP and preparation of COLMAP-based datasets for nerstudio training.
 """
 
 import json
 from pathlib import Path
-from typing import Any, Dict, Literal, Optional
+from typing import Any, Dict, Optional
 
 import appdirs
 import cv2
@@ -27,6 +27,7 @@ import requests
 import torch
 from rich.console import Console
 from rich.progress import track
+from typing_extensions import Literal
 
 # TODO(1480) use pycolmap instead of colmap_parsing_utils
 # import pycolmap
@@ -114,7 +115,9 @@ def run_colmap(
     colmap_version = get_colmap_version(colmap_cmd)
 
     colmap_database_path = colmap_dir / "database.db"
-    colmap_database_path.unlink(missing_ok=True)
+    if colmap_database_path.exists():
+        # Can't use missing_ok argument because of Python 3.7 compatibility.
+        colmap_database_path.unlink()
 
     # Feature extraction
     feature_extractor_cmd = [

nerfstudio/process_data/hloc_utils.py

@@ -21,9 +21,9 @@ Requires hloc module from : https://github.com/cvg/Hierarchical-Localization
 
 import sys
 from pathlib import Path
-from typing import Literal
 
 from rich.console import Console
+from typing_extensions import Literal
 
 from nerfstudio.process_data.process_data_utils import CameraModel
 

nerfstudio/process_data/metashape_utils.py

@@ -61,17 +61,12 @@ def metashape_to_json(  # pylint: disable=too-many-statements
     if sensors is None:
         raise ValueError("No sensors found")
 
-    calibrated_sensors = [
-        sensor for sensor in sensors if sensor.get("type") == "spherical" or sensor.find("calibration")
-    ]
+    calibrated_sensors = [sensor for sensor in sensors if sensor.find("calibration")]
     if not calibrated_sensors:
         raise ValueError("No calibrated sensor found in Metashape XML")
     sensor_type = [s.get("type") for s in calibrated_sensors]
     if sensor_type.count(sensor_type[0]) != len(sensor_type):
-        raise ValueError(
-            "All Metashape sensors do not have the same sensor type. "
-            "nerfstudio does not support per-frame camera_model."
-        )
+        raise ValueError("All Metashape sensors do not have the same sensor type")
     data = {}
     if sensor_type[0] == "frame":
         data["camera_model"] = CAMERA_MODELS["perspective"].value
@@ -92,54 +87,46 @@ def metashape_to_json(  # pylint: disable=too-many-statements
         s["h"] = int(resolution.get("height"))  # type: ignore
 
         calib = sensor.find("calibration")
-        if calib is None:
-            assert sensor_type[0] == "spherical", "Only spherical sensors should have no intrinsics"
-            s["fl_x"] = s["w"] / 2.0
-            s["fl_y"] = s["h"]
-            s["cx"] = s["w"] / 2.0
-            s["cy"] = s["h"] / 2.0
-        else:
-            f = calib.find("f")
-            assert f is not None, "Focal length not found in Metashape xml"
+        f = calib.find("f")
+        if f is not None:
             s["fl_x"] = s["fl_y"] = float(f.text)  # type: ignore
-            s["cx"] = _find_param(calib, "cx") + s["w"] / 2.0  # type: ignore
-            s["cy"] = _find_param(calib, "cy") + s["h"] / 2.0  # type: ignore
+        s["cx"] = _find_param(calib, "cx") + s["w"] / 2.0  # type: ignore
+        s["cy"] = _find_param(calib, "cy") + s["h"] / 2.0  # type: ignore
 
-            s["k1"] = _find_param(calib, "k1")
-            s["k2"] = _find_param(calib, "k2")
-            s["k3"] = _find_param(calib, "k3")
-            s["k4"] = _find_param(calib, "k4")
-            s["p1"] = _find_param(calib, "p1")
-            s["p2"] = _find_param(calib, "p2")
+        s["k1"] = _find_param(calib, "k1")
+        s["k2"] = _find_param(calib, "k2")
+        s["k3"] = _find_param(calib, "k3")
+        s["k4"] = _find_param(calib, "k4")
+        s["p1"] = _find_param(calib, "p1")
+        s["p2"] = _find_param(calib, "p2")
 
         sensor_dict[sensor.get("id")] = s
 
     components = chunk.find("components")
-    if components is not None:
-        component_dict = {}
-        for component in components:
-            transform = component.find("transform")
-            if transform is not None:
-                rotation = transform.find("rotation")
-                if rotation is None:
-                    r = np.eye(3)
-                else:
-                    r = np.array([float(x) for x in rotation.text.split()]).reshape((3, 3))
-                translation = transform.find("translation")
-                if translation is None:
-                    t = np.zeros(3)
-                else:
-                    t = np.array([float(x) for x in translation.text.split()])
-                scale = transform.find("scale")
-                if scale is None:
-                    s = 1.0
-                else:
-                    s = float(scale.text)
+    component_dict = {}
+    for component in components:
+        transform = component.find("transform")
+        if transform is not None:
+            rotation = transform.find("rotation")
+            if rotation is None:
+                r = np.eye(3)
+            else:
+                r = np.array([float(x) for x in rotation.text.split()]).reshape((3, 3))
+            translation = transform.find("translation")
+            if translation is None:
+                t = np.zeros(3)
+            else:
+                t = np.array([float(x) for x in translation.text.split()])
+            scale = transform.find("scale")
+            if scale is None:
+                s = 1.0
+            else:
+                s = float(scale.text)
 
-                m = np.eye(4)
-                m[:3, :3] = r
-                m[:3, 3] = t / s
-                component_dict[component.get("id")] = m
+            m = np.eye(4)
+            m[:3, :3] = r
+            m[:3, 3] = t / s
+            component_dict[component.get("id")] = m
 
     frames = []
     cameras = chunk.find("cameras")
@@ -172,10 +159,9 @@ def metashape_to_json(  # pylint: disable=too-many-statements
             num_skipped += 1
             continue
         transform = np.array([float(x) for x in camera.find("transform").text.split()]).reshape((4, 4))
-        if components is not None:
-            component_id = camera.get("component_id")
-            if component_id in component_dict:
-                transform = component_dict[component_id] @ transform
+        component_id = camera.get("component_id")
+        if component_id in component_dict:
+            transform = component_dict[component_id] @ transform
         transform = transform[[2, 0, 1, 3], :]
         transform[:, 1:3] *= -1
         frame["transform_matrix"] = transform.tolist()

nerfstudio/process_data/process_data_utils.py

@@ -20,11 +20,12 @@ import shutil
 import sys
 from enum import Enum
 from pathlib import Path
-from typing import List, Literal, Optional, OrderedDict, Tuple
+from typing import List, Optional, Tuple
 
 import cv2
 import numpy as np
 from rich.console import Console
+from typing_extensions import Literal, OrderedDict
 
 from nerfstudio.utils.rich_utils import status
 from nerfstudio.utils.scripts import run_command
@@ -38,13 +39,12 @@ class CameraModel(Enum):
 
     OPENCV = "OPENCV"
     OPENCV_FISHEYE = "OPENCV_FISHEYE"
-    EQUIRECTANGULAR = "EQUIRECTANGULAR"
 
 
 CAMERA_MODELS = {
     "perspective": CameraModel.OPENCV,
     "fisheye": CameraModel.OPENCV_FISHEYE,
-    "equirectangular": CameraModel.EQUIRECTANGULAR,
+    "equirectangular": CameraModel.OPENCV,
 }