VectorMapper is a tool that helps to create a vector map from point cloud data, which is compatible to ADAS Map. The vector map represents a set of features inherent to the road, such as lanes, stop lines, traffic lights, and intersections. These pieces of information are particularly leveraged by Autoware, popular open-source software for self-driving, to enhance capabilities of path planning, object detection, traffic light recognition, and other critical tasks.
PCMapper is a tool that eases building a 3D point cloud map. Using as input a previously recorded ROSBAG file that may include range information captured by a LiDAR sensor, PCMapper will generate the file as output that contains a 3D point cloud map. The file formats based on PCD (Point Cloud Data), which is a standard format provided by PCL (Point Cloud Library). The generated map can be used, for instance, to localize a self-driving vehicle in the region captured.
MapExtender is a tool that extends a 3D point cloud map by adding new areas to the base map data. Both the new and the base map data need to be the PCD files. This is enabled by finding overlapped areas of the two PCD files so that they can be connected consistently with each other.
MapFilter is a tool that supports downsampling a dense 3D point cloud map using voxel grid filtering. This will reduce the computational load of localization and the data size of high-resolution maps generated by PCMapper. We recommend highly that the 3D point cloud map be downsampled for the purpose of self-driving.