Improve Documentation - Fix code links and improve tutorial layout

docs/tutorial/reconstruction_system/system_overview.rst

@@ -48,7 +48,7 @@ Getting the example code
 
 .. code-block:: sh
 
-    # Activate your conda enviroment, where you have installed open3d pip package.
+    # Activate your conda environment, where you have installed open3d pip package.
     # Clone the Open3D github repository and go to the example.
     cd examples/python/reconstruction_system/
 
@@ -69,7 +69,7 @@ Running the example with default dataset.
     python run_system.py --make --register --refine --integrate
 
 Changing the default dataset.
-One may change the default dataset to other avaialble datasets. 
+One may change the default dataset to other available datasets.
 Currently the following datasets are available:
 
 1. Lounge (keyword: ``lounge``) (Default)

docs/tutorial/t_reconstruction_system/index.rst

@@ -25,7 +25,7 @@ Getting the example code
 
 .. code-block:: sh
 
-    # Activate your conda enviroment, where you have installed open3d pip package.
+    # Activate your conda environment, where you have installed open3d pip package.
     # Clone the Open3D github repository and go to the example.
     cd examples/python/t_reconstruction_system/
 
@@ -40,7 +40,7 @@ Running the example with default dataset.
     # which is ``lounge`` dataset from stanford. 
     python dense_slam_gui.py 
 
-It is recommended to use CUDA if avaialble.
+It is recommended to use CUDA if available.
 
 .. code-block:: sh
 
@@ -49,7 +49,7 @@ It is recommended to use CUDA if avaialble.
     python dense_slam_gui.py --device 'cuda:0'
 
 Changing the default dataset.
-One may change the default dataset to other avaialble datasets. 
+One may change the default dataset to other available datasets.
 Currently the following datasets are available:
 
 1. Lounge (keyword: ``lounge``) (Default)

docs/tutorial/t_reconstruction_system/voxel_block_grid.rst

@@ -14,21 +14,21 @@ A voxel block grid can be constructed by:
 
 .. literalinclude:: ../../../examples/python/t_reconstruction_system/integrate.py
    :language: python
-   :lineno-start: 56
-   :lines: 27,57-74
+   :lineno-start: 27
+   :lines: 8,28-45
 
-In this example, the multi-value hash map has key shape shape ``(3,)`` and dtype ``int32``. The hash map values are organized as a structure of array (SoA). The hash map values include:
+In this example, the multi-value hash map has key of shape ``(3,)`` and dtype ``float32``. The hash map values are organized as a structure of array (SoA). The hash map values include:
 
 By default it contains:
 
-- Truncated Signed Distance Function (TSDF) of element shape ``(8, 8, 8, 1)``
-- Weight of element shape ``(8, 8, 8, 1)``
-- (Optionally) RGB color of element shape ``(8, 8, 8, 3)``
+- Truncated Signed Distance Function (TSDF) of element shape ``(16, 16, 16, 1)``
+- Weight of element shape ``(16, 16, 16, 1)``
+- (Optionally) RGB color of element shape ``(16, 16, 16, 3)``
 
-where ``8`` stands for the local voxel block grid resolution.
+where ``16`` stands for the local voxel block grid resolution.
 
 By convention, we use ``3.0 / 512`` as the voxel resolution. This spatial resolution is equivalent to representing a ``3m x 3m x 3m`` (m = meter) room with a dense ``512 x 512 x 512`` voxel grid.
 
-The voxel block grid is optimized to run fast on GPU. On CPU the it runs slower. Empirically, we reserve ``100000`` such blocks for a living room-scale scene to avoid frequent rehashing.
+The voxel block grid is optimized to run fast on GPU. On CPU, it runs slower. Empirically, we reserve ``50000`` such blocks for a living room-scale scene to avoid frequent rehashing.
 
-You can always customize your own properties, e.g., ``intensity`` of element shape ``(8, 8, 8, 1)`` in ``float32``, ``label`` of element shape ``(8, 8, 8, 1)`` in ``int32``, etc. To know how to process the data, please refer to :ref:`customized_integration`.
+You can always customize your own properties, e.g., ``intensity`` of element shape ``(16, 16, 16, 1)`` in ``float32``, ``label`` of element shape ``(16, 16, 16, 1)`` in ``int32``, etc. To know how to process the data, please refer to :ref:`customized_integration`.

docs/tutorial/visualization/cpu_rendering.rst

@@ -46,7 +46,7 @@ Here are the different ways to do that:
 
     # In Python code
     import os
-    os.environ['EGL_PLATFORM'] = 'surfaceless'   # Ubunu 20.04+
+    os.environ['EGL_PLATFORM'] = 'surfaceless'   # Ubuntu 20.04+
     os.environ['OPEN3D_CPU_RENDERING'] = 'true'  # Ubuntu 18.04
     import open3d as o3d
 

docs/tutorial/visualization/interactive_visualization.rst

@@ -3,12 +3,14 @@
 Interactive visualization
 -------------------------------------
 
-This tutorial introduces user interaction features of the visualizer window.
+This tutorial introduces user interaction features of the visualizer window provided by:-
+
+#. ``open3d.visualization.draw_geometries_with_editing``
+#. ``open3d.visualization.VisualizerWithEditing``
 
 .. literalinclude:: ../../../examples/python/visualization/interactive_visualization.py
    :language: python
-   :lineno-start: 27
-   :lines: 27-
+   :start-at: # examples/python/visualization/interactive_visualization.py
    :linenos:
 
 This script executes two applications of user interaction: ``demo_crop_geometry`` and ``demo_manual_registration``.
@@ -20,8 +22,7 @@ Crop geometry
 
 .. literalinclude:: ../../../examples/python/visualization/interactive_visualization.py
    :language: python
-   :lineno-start: 37
-   :lines: 37-51
+   :pyobject: demo_crop_geometry
    :linenos:
 
 This function simply reads a point cloud and calls ``draw_geometries_with_editing``. This function provides vertex selection and cropping.
@@ -58,26 +59,27 @@ To finish selection mode, press ``F`` to switch to freeview mode.
 Manual registration
 `````````````````````````````````````````````
 
-Select correspondences
-=====================================
-
 The following script registers two point clouds using point-to-point ICP. It gets initial alignment via user interaction.
 
+Prepare data
+=====================================
+
 .. literalinclude:: ../../../examples/python/visualization/interactive_visualization.py
    :language: python
-   :lineno-start: 61
-   :lines: 61-76
+   :pyobject: prepare_data
    :linenos:
 
-The script reads two point clouds, and visualizes the point clouds before alignment.
+This function reads two point clouds, and visualizes the point clouds before performing manual alignment.
 
 .. image:: ../../_static/visualization/interactive_visualization/manual_icp_initial.png
     :width: 400px
 
+Select correspondences
+=====================================
+
 .. literalinclude:: ../../../examples/python/visualization/interactive_visualization.py
    :language: python
-   :lineno-start: 52
-   :lines: 52-60
+   :pyobject: pick_points
    :linenos:
 
 The function ``pick_points(pcd)`` makes an instance of ``VisualizerWithEditing``. To mimic ``draw_geometries``, it creates windows, adds the geometry, visualizes the geometry, and then terminates. A novel interface function from ``VisualizerWithEditing`` is ``get_picked_points()`` that returns the indices of user-picked vertices.
@@ -115,8 +117,7 @@ Registration using user correspondences
 
 .. literalinclude:: ../../../examples/python/visualization/interactive_visualization.py
    :language: python
-   :lineno-start: 77
-   :lines: 77-110
+   :pyobject: register_via_correspondences
    :linenos:
 
 The later part of the demo computes an initial transformation based on the user-provided correspondences. This script builds pairs of correspondences using ``Vector2iVector(corr)``. It utilizes ``TransformationEstimationPointToPoint.compute_transformation`` to compute the initial transformation from the correspondences. The initial transformation is refined using ``registration_icp``.

docs/tutorial/visualization/non_blocking_visualization.rst

@@ -28,6 +28,7 @@ This rendering loop can be readily customized. For example, a custom loop can be
 
     vis = Visualizer()
     vis.create_window()
+    vis.add_geometry(geometry)
     for i in range(icp_iteration):
         # do ICP single iteration
         # transform geometry using ICP
@@ -39,8 +40,7 @@ The full script implementing this idea is displayed below.
 
 .. literalinclude:: ../../../examples/python/visualization/non_blocking_visualization.py
    :language: python
-   :lineno-start: 27
-   :lines: 27-
+   :start-at: # examples/python/visualization/non_blocking_visualization.py
    :linenos:
 
 The following sections explain this script.
@@ -50,8 +50,7 @@ Prepare example data
 
 .. literalinclude:: ../../../examples/python/visualization/non_blocking_visualization.py
    :language: python
-   :lineno-start: 35
-   :lines: 35-46
+   :pyobject: prepare_data
    :linenos:
 
 This part reads two point clouds and downsamples them. The source point cloud is intentionally transformed for the misalignment. Both point clouds are flipped for better visualization.
@@ -61,9 +60,10 @@ Initialize Visualizer class
 
 .. literalinclude:: ../../../examples/python/visualization/non_blocking_visualization.py
    :language: python
-   :lineno-start: 47
-   :lines: 47-59
+   :start-after: def demo_non_blocking_visualization():
    :linenos:
+   :lineno-start: 30
+   :lines: 3-7
 
 These lines make an instance of the visualizer class, open a visualizer window, and add two geometries to the visualizer.
 
@@ -72,9 +72,10 @@ Transform geometry and visualize it
 
 .. literalinclude:: ../../../examples/python/visualization/non_blocking_visualization.py
    :language: python
-   :lineno-start: 59
-   :lines: 59-72
+   :start-after: def demo_non_blocking_visualization():
    :linenos:
+   :lineno-start: 30
+   :lines: 8-23
 
 This script calls ``registration_icp`` for every iteration. Note that it explicitly forces only one ICP iteration via ``ICPConvergenceCriteria(max_iteration = 1)``. This is a trick to retrieve a slight pose update from a single ICP iteration. After ICP, source geometry is transformed accordingly.
 

examples/python/visualization/interactive_visualization.py

@@ -37,6 +37,15 @@ def draw_registration_result(source, target, transformation):
     o3d.visualization.draw_geometries([source_temp, target_temp])
 
 
+def prepare_data():
+    pcd_data = o3d.data.DemoICPPointClouds()
+    source = o3d.io.read_point_cloud(pcd_data.paths[0])
+    target = o3d.io.read_point_cloud(pcd_data.paths[2])
+    print("Visualization of two point clouds before manual alignment")
+    draw_registration_result(source, target, np.identity(4))
+    return source, target
+
+
 def pick_points(pcd):
     print("")
     print(
@@ -53,36 +62,34 @@ def pick_points(pcd):
     return vis.get_picked_points()
 
 
-def demo_manual_registration():
-    print("Demo for manual ICP")
-    pcd_data = o3d.data.DemoICPPointClouds()
-    source = o3d.io.read_point_cloud(pcd_data.paths[0])
-    target = o3d.io.read_point_cloud(pcd_data.paths[2])
-    print("Visualization of two point clouds before manual alignment")
-    draw_registration_result(source, target, np.identity(4))
-
-    # pick points from two point clouds and builds correspondences
-    picked_id_source = pick_points(source)
-    picked_id_target = pick_points(target)
-    assert (len(picked_id_source) >= 3 and len(picked_id_target) >= 3)
-    assert (len(picked_id_source) == len(picked_id_target))
-    corr = np.zeros((len(picked_id_source), 2))
-    corr[:, 0] = picked_id_source
-    corr[:, 1] = picked_id_target
-
+def register_via_correspondences(source, target, source_points, target_points):
+    corr = np.zeros((len(source_points), 2))
+    corr[:, 0] = source_points
+    corr[:, 1] = target_points
     # estimate rough transformation using correspondences
     print("Compute a rough transform using the correspondences given by user")
     p2p = o3d.pipelines.registration.TransformationEstimationPointToPoint()
     trans_init = p2p.compute_transformation(source, target,
                                             o3d.utility.Vector2iVector(corr))
-
     # point-to-point ICP for refinement
     print("Perform point-to-point ICP refinement")
     threshold = 0.03  # 3cm distance threshold
     reg_p2p = o3d.pipelines.registration.registration_icp(
         source, target, threshold, trans_init,
         o3d.pipelines.registration.TransformationEstimationPointToPoint())
     draw_registration_result(source, target, reg_p2p.transformation)
+
+
+def demo_manual_registration():
+    print("Demo for manual ICP")
+    source, target = prepare_data()
+
+    # pick points from two point clouds and builds correspondences
+    source_points = pick_points(source)
+    target_points = pick_points(target)
+    assert (len(source_points) >= 3 and len(target_points) >= 3)
+    assert (len(source_points) == len(target_points))
+    register_via_correspondences(source, target, source_points, target_points)
     print("")
 
 

examples/python/visualization/non_blocking_visualization.py

@@ -10,22 +10,27 @@
 import open3d as o3d
 import numpy as np
 
-if __name__ == "__main__":
-    o3d.utility.set_verbosity_level(o3d.utility.VerbosityLevel.Debug)
+
+def prepare_data():
     pcd_data = o3d.data.DemoICPPointClouds()
     source_raw = o3d.io.read_point_cloud(pcd_data.paths[0])
     target_raw = o3d.io.read_point_cloud(pcd_data.paths[1])
-
     source = source_raw.voxel_down_sample(voxel_size=0.02)
     target = target_raw.voxel_down_sample(voxel_size=0.02)
+
     trans = [[0.862, 0.011, -0.507, 0.0], [-0.139, 0.967, -0.215, 0.7],
              [0.487, 0.255, 0.835, -1.4], [0.0, 0.0, 0.0, 1.0]]
     source.transform(trans)
-
     flip_transform = [[1, 0, 0, 0], [0, -1, 0, 0], [0, 0, -1, 0], [0, 0, 0, 1]]
     source.transform(flip_transform)
     target.transform(flip_transform)
+    return source, target
+
 
+def demo_non_blocking_visualization():
+    o3d.utility.set_verbosity_level(o3d.utility.VerbosityLevel.Debug)
+
+    source, target = prepare_data()
     vis = o3d.visualization.Visualizer()
     vis.create_window()
     vis.add_geometry(source)
@@ -46,4 +51,9 @@
         if save_image:
             vis.capture_screen_image("temp_%04d.jpg" % i)
     vis.destroy_window()
+
     o3d.utility.set_verbosity_level(o3d.utility.VerbosityLevel.Info)
+
+
+if __name__ == '__main__':
+    demo_non_blocking_visualization()