@@ -13,17 +13,17 @@ Next figure shows the basic structure of filter nodes for both the 3D LIDAR and
The filters used for the 3D LIDAR are:
*[lidar_angle_filter](https://gitlab.iri.upc.edu/mobile_robotics/colrobtransp/3d_navigation/iri_point_cloud_angle_filter): it filters some angular sections of the LIDAR data (to avoid detection of the robot itself for example). Follow the link for a more detailed information.
*[lidar_angle_filter](https://gitlab.iri.upc.edu/labrobotica/ros/navigation/3d_navigation/iri_point_cloud_angle_filter): it filters some angular sections of the LIDAR data (to avoid detection of the robot itself for example). Follow the link for a more detailed information.
*[lidar_obstacle_detector](https://gitlab.iri.upc.edu/mobile_robotics/colrobtransp/3d_navigation/iri_lidar_obstacle_detection): it uses the pointcloud data to detect the slope (both positive and negative) of the surrounding environment and remove from the output pointcloud those regions that are traversable by the robot. It is also capable of detecting holes. Follow the link for a more detailed information.
*[lidar_obstacle_detector](https://gitlab.iri.upc.edu/labrobotica/ros/navigation/3d_navigation/iri_lidar_obstacle_detection): it uses the pointcloud data to detect the slope (both positive and negative) of the surrounding environment and remove from the output pointcloud those regions that are traversable by the robot. It is also capable of detecting holes. Follow the link for a more detailed information.
The filters used to pre-process the pointcloud from the depth camera before actual detection are:
*[pc_throttle](https://gitlab.iri.upc.edu/mobile_robotics/colrobtransp/3d_navigation/iri_average_point_cloud): it reduces the rate of the incoming pointcloud to a more suitable rate for the navigation stack to reduce the computational cost of the whole filter chain. Follow the link for a more detailed information.
*[pc_throttle](https://gitlab.iri.upc.edu/labrobotica/ros/navigation/3d_navigation/iri_average_point_cloud): it reduces the rate of the incoming pointcloud to a more suitable rate for the navigation stack to reduce the computational cost of the whole filter chain. Follow the link for a more detailed information.
*[average_point_cloud](https://gitlab.iri.upc.edu/mobile_robotics/colrobtransp/3d_navigation/iri_average_point_cloud): it averages a number of pointclouds using a moving window to reduce the noise. Follow the link for a more detailed information.
*[average_point_cloud](https://gitlab.iri.upc.edu/labrobotica/ros/navigation/3d_navigation/iri_average_point_cloud): it averages a number of pointclouds using a moving window to reduce the noise. Follow the link for a more detailed information.
*[voxel_grid](http://wiki.ros.org/pcl_ros/Tutorials/filters): it down-samples the input pointcloud to reduce the number of points to handle in the next filters. Follow the link for a more detailed information.
...
...
@@ -37,13 +37,13 @@ The filters used to pre-process the pointcloud from the depth camera before actu
The node used to detect ramps and positive obstacles using the depth camera is:
*[obstacle_normals_detection](https://gitlab.iri.upc.edu/mobile_robotics/colrobtransp/3d_navigation/iri_obstacle_detection_normals): it computes the normal of all the points using their neightbors to fit a plane, and use them to detect the slope to check whether it is traversable by the robot or not. Follow the link for a more detailed information.
*[obstacle_normals_detection](https://gitlab.iri.upc.edu/labrobotica/ros/navigation/3d_navigation/iri_obstacle_detection_normals): it computes the normal of all the points using their neightbors to fit a plane, and use them to detect the slope to check whether it is traversable by the robot or not. Follow the link for a more detailed information.
The node used to detect holes and negative obstacles using the depth camera is:
*[point_cloud_hole_detection](https://gitlab.iri.upc.edu/mobile_robotics/colrobtransp/3d_navigation/iri_point_cloud_hole_detection): it detects the number of points at ground level at a given region in front of the robot. If the number of points does not reach a certain threshold, a hole is detected. Follow the link for a more detailed information.
*[point_cloud_hole_detection](https://gitlab.iri.upc.edu/labrobotica/ros/navigation/3d_navigation/iri_point_cloud_hole_detection): it detects the number of points at ground level at a given region in front of the robot. If the number of points does not reach a certain threshold, a hole is detected. Follow the link for a more detailed information.
