Updated LiDAR algorithm pipeline from LBL research

docs/examples/extreme_weather/Austin_hail_damage_example.ipynb

@@ -65,9 +65,7 @@
   {
    "cell_type": "code",
    "execution_count": 3,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
@@ -132,9 +130,7 @@
   {
    "cell_type": "code",
    "execution_count": 5,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stderr",
@@ -715,9 +711,9 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "mesh_test3",
+   "display_name": "Python 3 (ipykernel)",
    "language": "python",
-   "name": "mesh_test3"
+   "name": "python3"
   },
   "language_info": {
    "codemirror_mode": {
@@ -729,7 +725,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.1"
+   "version": "3.11.10"
   }
  },
  "nbformat": 4,

docs/examples/panel_detection/Sol-Searcher_example.ipynb

@@ -40,9 +40,7 @@
   {
    "cell_type": "code",
    "execution_count": 2,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
@@ -201,9 +199,7 @@
   {
    "cell_type": "code",
    "execution_count": 4,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stderr",
@@ -329,9 +325,7 @@
   {
    "cell_type": "code",
    "execution_count": 7,
-   "metadata": {
-    "scrolled": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "solar_install_address_list = list()\n",
@@ -1200,9 +1194,9 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "panel_seg_extreme_weather",
+   "display_name": "Python 3 (ipykernel)",
    "language": "python",
-   "name": "panel_seg_extreme_weather"
+   "name": "python3"
   },
   "language_info": {
    "codemirror_mode": {
@@ -1214,7 +1208,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.5"
+   "version": "3.11.10"
   }
  },
  "nbformat": 4,

docs/index_api.rst

@@ -9,41 +9,6 @@ Panel Detection
 ---------------
 Generates satellite images and runs DL and CV routines on images to get array azimuth and mounting type/configuration.
 
-.. autosummary::
-   :toctree: generated/
-   :caption: Panel Detection
-
-   panel_detection.PanelDetection
-   panel_detection.PanelDetection.generateSatelliteImage
-   panel_detection.PanelDetection.classifyMountingConfiguration
-   panel_detection.PanelDetection.diceCoeff
-   panel_detection.PanelDetection.diceCoeffLoss
-   panel_detection.PanelDetection.testBatch
-   panel_detection.PanelDetection.testSingle
-   panel_detection.PanelDetection.hasPanels
-   panel_detection.PanelDetection.detectAzimuth
-   panel_detection.PanelDetection.cropPanels
-   panel_detection.PanelDetection.plotEdgeAz
-   panel_detection.PanelDetection.clusterPanels
-   panel_detection.PanelDetection.runSiteAnalysisPipeline
-
-Panel Training
----------------
-Deep learning model training and development tools.
-
-.. autosummary::
-   :toctree: generated/
-   :caption: Panel Training
-
-   panel_train.TrainPanelSegmentationModel
-   panel_train.TrainPanelSegmentationModel.loadImagesToNumpyArray
-   panel_train.TrainPanelSegmentationModel.diceCoeff
-   panel_train.TrainPanelSegmentationModel.diceCoeffLoss
-   panel_train.TrainPanelSegmentationModel.trainSegmentation
-   panel_train.TrainPanelSegmentationModel.trainPanelClassifier
-   panel_train.TrainPanelSegmentationModel.trainMountingConfigClassifier
-   panel_train.TrainPanelSegmentationModel.trainingStatistics
-
 
 Utilities
 ---------
@@ -64,6 +29,8 @@ Helper functions and utilities.
    utils.binaryMaskToPolygon
    utils.convertMaskToLatLonPolygon
    utils.convertPolygonToGeojson
+   utils.detectAzimuth
+   utils.plotEdgeAz
 
 LiDAR
 -----

panel_segmentation.egg-info/PKG-INFO

@@ -0,0 +1,51 @@
+Metadata-Version: 2.4
+Name: panel_segmentation
+Version: 0.0.0
+Summary: A package to segment solar panels from a satellite image and perform automated metadata extraction.
+Author-email: Kirsten Perry <kirsten.perry@nrel.gov>, Quyen Nguyen <nguyquye@oregonstate.edu>, Kevin Anderson <kevin.anderson@snl.gov>, Christopher Campos <chris.acampos@yahoo.com>, Ayobami Edun <aedun@ufl.edu>
+License: MIT
+Project-URL: Homepage, https://github.com/NREL/Panel-Segmentation
+Keywords: photovoltaic,solar,analytics,analysis,performance,PV,satellite
+Requires-Python: >=3.8
+License-File: LICENSE
+Requires-Dist: opencv-python-headless==4.6.0.66
+Requires-Dist: numpy>=1.18.1
+Requires-Dist: scikit_image>=0.16.2
+Requires-Dist: matplotlib>=3.1.3
+Requires-Dist: requests>=2.22.0
+Requires-Dist: tensorflow>=2.2.0
+Requires-Dist: pandas>=1.1.4
+Requires-Dist: Pillow>=8.0.1
+Requires-Dist: scikit_learn>=0.23.2
+Requires-Dist: h5py>=2.10.0
+Requires-Dist: detecto>=1.2.1
+Requires-Dist: torch>=1.9.0
+Requires-Dist: torchvision>=0.10.0
+Requires-Dist: mmdet==3.3.0
+Requires-Dist: mmengine==0.10.7
+Requires-Dist: kornia
+Requires-Dist: scipy
+Requires-Dist: rasterio
+Requires-Dist: geopandas
+Requires-Dist: beautifulsoup4==4.12
+Requires-Dist: dash==2.6.0
+Requires-Dist: laspy[laszip]==2.5.4
+Requires-Dist: open3d==0.18.0
+Requires-Dist: pyarrow==17.0.0
+Requires-Dist: pyproj==3.5.0
+Requires-Dist: boto3>=1.37.38
+Requires-Dist: cfgrib>=0.9.15.0
+Requires-Dist: folium>=0.18.0
+Requires-Dist: simplekml>=1.3.6
+Requires-Dist: xarray>=2023.1.0
+Provides-Extra: doc
+Requires-Dist: sphinx==5.0; extra == "doc"
+Requires-Dist: jinja2<3.1; extra == "doc"
+Requires-Dist: pydata-sphinx-theme==0.14.0; extra == "doc"
+Requires-Dist: ipython; extra == "doc"
+Requires-Dist: nbsphinx; extra == "doc"
+Requires-Dist: sphinx_copybutton; extra == "doc"
+Provides-Extra: test
+Requires-Dist: pytest>=5.3.5; extra == "test"
+Provides-Extra: optional
+Dynamic: license-file

panel_segmentation.egg-info/SOURCES.txt

@@ -0,0 +1,30 @@
+LICENSE
+MANIFEST.in
+README.md
+pyproject.toml
+versioneer.py
+panel_segmentation/__init__.py
+panel_segmentation/_version.py
+panel_segmentation/utils.py
+panel_segmentation.egg-info/PKG-INFO
+panel_segmentation.egg-info/SOURCES.txt
+panel_segmentation.egg-info/dependency_links.txt
+panel_segmentation.egg-info/not-zip-safe
+panel_segmentation.egg-info/requires.txt
+panel_segmentation.egg-info/top_level.txt
+panel_segmentation/lidar/__init__.py
+panel_segmentation/lidar/pcd_data.py
+panel_segmentation/lidar/plane_segmentation.py
+panel_segmentation/lidar/usgs_lidar_api.py
+panel_segmentation/models/hail_config.py
+panel_segmentation/models/hail_model.pth
+panel_segmentation/models/post_hurricane_config.py
+panel_segmentation/models/post_hurricane_model.pth
+panel_segmentation/models/pre_hurricane_config.py
+panel_segmentation/models/pre_hurricane_model.pth
+panel_segmentation/models/sol_searcher_config.py
+panel_segmentation/models/sol_searcher_model.pth
+panel_segmentation/tests/README.md
+panel_segmentation/tests/__init__.py
+panel_segmentation/tests/test_panel_detection.py
+panel_segmentation/tests/test_utils.py

panel_segmentation.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

panel_segmentation.egg-info/not-zip-safe

@@ -0,0 +1 @@
+

panel_segmentation.egg-info/requires.txt

@@ -0,0 +1,43 @@
+opencv-python-headless==4.6.0.66
+numpy>=1.18.1
+scikit_image>=0.16.2
+matplotlib>=3.1.3
+requests>=2.22.0
+tensorflow>=2.2.0
+pandas>=1.1.4
+Pillow>=8.0.1
+scikit_learn>=0.23.2
+h5py>=2.10.0
+detecto>=1.2.1
+torch>=1.9.0
+torchvision>=0.10.0
+mmdet==3.3.0
+mmengine==0.10.7
+kornia
+scipy
+rasterio
+geopandas
+beautifulsoup4==4.12
+dash==2.6.0
+laspy[laszip]==2.5.4
+open3d==0.18.0
+pyarrow==17.0.0
+pyproj==3.5.0
+boto3>=1.37.38
+cfgrib>=0.9.15.0
+folium>=0.18.0
+simplekml>=1.3.6
+xarray>=2023.1.0
+
+[doc]
+sphinx==5.0
+jinja2<3.1
+pydata-sphinx-theme==0.14.0
+ipython
+nbsphinx
+sphinx_copybutton
+
+[optional]
+
+[test]
+pytest>=5.3.5

panel_segmentation.egg-info/top_level.txt

@@ -0,0 +1 @@
+panel_segmentation

panel_segmentation/__init__.py

@@ -1,5 +1,3 @@
-from panel_segmentation.panel_detection import PanelDetection
-from panel_segmentation.panel_train import TrainPanelSegmentationModel
 from panel_segmentation import lidar
 
 from ._version import get_versions

panel_segmentation/lidar/pcd_data.py

@@ -83,38 +83,24 @@ def readLaz(self, chunk_size=1000000):
         # Ensure that the inputs are of the correct type
         if not isinstance(chunk_size, int):
             raise TypeError("chunk_size variable must be of type int.")
-        # Process LiDAR data in chunks since .laz files are large
-        with laspy.open(self.laz_file_path) as f:
-            final_point_record = laspy.PackedPointRecord.empty(
-                f.header.point_format)
-            # Get crs of LiDAR data
+        with laspy.open(self.laz_file_path) as f:            
             self.source_crs = f.header.parse_crs()
-            # Check if crs has multiple components
-            if self.source_crs.is_compound:
-                # Get horizontal crs
-                # Only need horizontal projection since Shapely polygon is 2D
-                horizontal_crs = self.source_crs.sub_crs_list[0]
-                dst_crs = f"EPSG:{horizontal_crs.to_epsg()}"
+            if self.source_crs is None:
+                # Some USGS LAZ tiles have no embedded CRS — assume UTM NAD83 as a
+                # fallback. 
+                print("No CRS found in %s -- falling back to EPSG:4269 (NAD83).",
+                    self.laz_file_path)
+                dst_crs = "EPSG:4269"
+            elif self.source_crs.is_compound:
+                # Compound CRS (e.g. horizontal + vertical) -- take the horizontal component
+                dst_crs = f"EPSG:{self.source_crs.sub_crs_list[0].to_epsg()}"
             else:
                 dst_crs = f"EPSG:{self.source_crs.to_epsg()}"
-            # Make coordinate transformer from EPSG:4326 to LiDAR crs
-            self.transformer = Transformer.from_crs(
-                "EPSG:4326", dst_crs, always_xy=True)
-            # Get [x, y] scale factors to match transformed lat long
-            # to array values
-            self.scales = f.header.scales
+
+            self.transformer = Transformer.from_crs("EPSG:4326", dst_crs, always_xy=True)
+            self.scales  = f.header.scales
             self.offsets = f.header.offsets
-            # Get the points in chunks
-            points_list = []
-            for points in f.chunk_iterator(chunk_size):
-                points_list.append(points.array)
-            # Combine all chunks
-            concatenated_points = np.concatenate(points_list)
-            # Combine the points together and create a new LasData object
-            final_point_record.array = concatenated_points
-            laz_data = laspy.LasData(header=f.header)
-            laz_data.points = final_point_record
-        return laz_data
+        return laspy.read(self.laz_file_path)
 
     def filterLaz(self, laz_data, classification_list=[1, 6],
                   lat_lon_bbox_size=20):