MSDM-SLAM

This repository represnets a 3D DNN-based Metric Semantic Dense Mapping pipeline and a Visual Inertial SLAM system that can be run on a ground mobile robot for the following tasks:

• An accurate automatic indoor dense 3D mapping with objects' semantic annotations.
• A robust and accurate SLAM system with enhanced version of ORB-SLAM3.
• Supporting 2D LiDAR mapping and navigation on occupancy grid.

One of the nodes in this pipeline is a ROS node for generating scaled metric depth estimation using MiDaS from https://github.com/isl-org/MiDaS and to generate point cloud of the estimated metric depth point cloud. Scaling process can be done using both of depth values from RGB-D camera or from features from Visual SLAM system.

To use that node PyTorch +1.7 should be there with CUDA +11.0.

Run Depth Only

roslaunch midas_cpp midas_cpp_xyzrgb.launch input_topic:="/input/image" gt_topic:="/ground_truth_depth" camera_info_in:="/image/camera_info"

Run Segmentation only

roslaunch midas_cpp midas_cpp_seg.launch input_topic:="/input/image"  image_input_original_topic:="/input/image_color" segmentation_topic:="/segmented"

Run Segmentation and Depth rescale with SLAM features

roslaunch midas_cpp midas_cpp_features_seg.launch input_topic:="/input/image"  camera_info_in:="/input/camera_info" map_topic:="/input/map_point" pose_topic:="/input/pose" image_input_original_topic:="/input/image_color" segmentation_topic:="/segmented"

ORB_SLAM3

Authors: Carlos Campos, Richard Elvira, Juan J. Gómez Rodríguez, José M. M. Montiel, Juan D. Tardos.

The Changelog describes the features of each version.

ORB-SLAM3 is the first real-time SLAM library able to perform Visual, Visual-Inertial and Multi-Map SLAM with monocular, stereo and RGB-D cameras, using pin-hole and fisheye lens models. In all sensor configurations, ORB-SLAM3 is as robust as the best systems available in the literature, and significantly more accurate.

We provide examples to run ORB-SLAM3 in the EuRoC dataset using stereo or monocular, with or without IMU, and in the TUM-VI dataset using fisheye stereo or monocular, with or without IMU. Videos of some example executions can be found at ORB-SLAM3 channel.

This software is based on ORB-SLAM2 developed by Raul Mur-Artal, Juan D. Tardos, J. M. M. Montiel and Dorian Galvez-Lopez (DBoW2).