Radar visualization (nutonomy#16)

* Added filtering for invalid RADAR points

* Created separate classes for Radar and Lidar PointCloud

* Fixed issue nutonomy#15

* Fixed rare bug when point cloud is empty

* Made readme clearer

* Added radar visualization

* Uncommented opencv rendering in tutorial

* Added missign requirement for scipy

* Implemented multisweep visualization

* Reformatting

* Added radar examples to tutorial

* Advanced filtering options for radar points

* Made nsweeps optional, refactoring

* Updated notebook

* Removed outputs from Ipython notebook

* changed order of data class methods for readability

* changed to used the same sample for full tutorial

README.md

@@ -12,8 +12,9 @@ Welcome to the devkit of the [nuScenes](https://www.nuscenes.org) dataset.
 - [Setting up a new virtual environment](#setting-up-a-new-virtual-environment)
 
 ## Changelog
-- Oct. 4, 2018: Code to parse RADAR data released
-- Sep. 12, 2018: Devkit for teaser dataset released
+- Nov. 21, 2018: RADAR filtering and multi sweep aggregation.
+- Oct. 4, 2018: Code to parse RADAR data released.
+- Sep. 12, 2018: Devkit for teaser dataset released.
 
 ## Dataset download
 To download nuScenes you need to go to the [Download page](https://www.nuscenes.org/download), 
@@ -24,10 +25,10 @@ Please unpack the archives to the `/data/nuscenes` folder \*without\* overwritin
 Eventually you should have the following folder structure:
 ```
 /data/nuscenes
-    maps	-	Large image files (~500 Gigapixel) that depict the drivable surface and sidewalks in the scene
-    samples	-	Sensor data for keyframes
-    sweeps	-	Sensor data for intermediate frames
-    v0.1	-	JSON tables that include all the meta data and annotations
+    maps	-	Large image files (~500 Gigapixel) that depict the drivable surface and sidewalks in the scene.
+    samples	-	Sensor data for keyframes.
+    sweeps	-	Sensor data for intermediate frames.
+    v0.1	-	JSON tables that include all the meta data and annotations.
 ```
 If you want to use another folder, specify the `dataroot` parameter of the NuScenes class below.
 
@@ -36,10 +37,9 @@ Download the devkit to your home directory using:
 ```
 cd && git clone https://github.com/nutonomy/nuscenes-devkit.git
 ```
-The devkit is tested for Python 3.7. 
-We may add backward compatibility in future releases.
-To install the required packages, run the following command in your favourite virtual environment. If you need help in 
-installing Python 3.7 or in setting up a new virtual environment, you can look at [these instructions](#setting-up-a-new-virtual-environment):
+The devkit is tested for Python 3.7.
+To install Python 3.7 and set up a new virtual environment, you can look at [these instructions](#setting-up-a-new-virtual-environment).
+To install the required packages, run the following command in your favourite virtual environment:
 ```
 pip install -r requirements.txt
 ```

python-sdk/export/export_pointclouds_as_obj.py

@@ -13,14 +13,16 @@
 from pyquaternion import Quaternion
 from tqdm import tqdm
 
-from nuscenes_utils.data_classes import PointCloud
+from nuscenes_utils.data_classes import LidarPointCloud
 from nuscenes_utils.geometry_utils import view_points
 from nuscenes_utils.nuscenes import NuScenes
 
 
 def export_scene_pointcloud(nusc: NuScenes, out_path: str, scene_token: str, channel: str='LIDAR_TOP',
                             min_dist: float=3.0, max_dist: float=30.0, verbose: bool=True) -> None:
     """
+    Export fused point clouds of a scene to a Wavefront OBJ file.
+    This point-cloud can be viewed in your favorite 3D rendering tool, e.g. Meshlab or Maya.
     :param nusc: NuScenes instance.
     :param out_path: Output path to write the point-cloud to.
     :param scene_token: Unique identifier of scene to render.
@@ -59,7 +61,7 @@ def export_scene_pointcloud(nusc: NuScenes, out_path: str, scene_token: str, cha
             sample_rec = nusc.get('sample', sc_rec['sample_token'])
             lidar_token = sd_rec['token']
             lidar_rec = nusc.get('sample_data', lidar_token)
-            pc = PointCloud.from_file(osp.join(nusc.dataroot, lidar_rec['filename']))
+            pc = LidarPointCloud.from_file(osp.join(nusc.dataroot, lidar_rec['filename']))
 
             # Get point cloud colors.
             coloring = np.ones((3, pc.points.shape[1])) * -1
@@ -113,7 +115,7 @@ def pointcloud_color_from_image(nusc: NuScenes, pointsensor_token: str, camera_t
     cam = nusc.get('sample_data', camera_token)
     pointsensor = nusc.get('sample_data', pointsensor_token)
 
-    pc = PointCloud.from_file(osp.join(nusc.dataroot, pointsensor['filename']))
+    pc = LidarPointCloud.from_file(osp.join(nusc.dataroot, pointsensor['filename']))
     im = Image.open(osp.join(nusc.dataroot, cam['filename']))
 
     # Points live in the point sensor frame. So they need to be transformed via global to the image plane.