GitPlanet

Cheap and reliable Node.js hosting starts at $3/month, and $1/month static HTML hosting

Created with love in Canada, visit hostnodejs.com today

Feel like to post an Ad? Learn Details
All Projects → bxh1 → VIDO-SLAM

bxh1 / VIDO-SLAM

Licence: other
VIDO-SLAM is a Visual Inertial SLAM system for dynamic environments, and it can also estimate dynamic objects motion and track objects.

Programming Languages

C++
36643 projects - #6 most used programming language
CMake
9771 projects

Labels

slamvisual-inertial-odometrysemantic-slamvido-slam

Projects that are alternatives of or similar to VIDO-SLAM

SALSA-Semantic-Assisted-SLAM
SALSA: Semantic Assisted Life-Long SLAM for Indoor Environments (16-833 SLAM Project at CMU)
Stars: ✭ 52 (-45.26%)
Mutual labels:  slam, semantic-slam
DSP-SLAM
[3DV 2021] DSP-SLAM: Object Oriented SLAM with Deep Shape Priors
Stars: ✭ 377 (+296.84%)
Mutual labels:  slam, semantic-slam
HybVIO
HybVIO visual-inertial odometry and SLAM system
Stars: ✭ 261 (+174.74%)
Mutual labels:  slam, visual-inertial-odometry
direct lidar odometry
Direct LiDAR Odometry: Fast Localization with Dense Point Clouds
Stars: ✭ 202 (+112.63%)
Mutual labels:  slam
bench ws
A catkin workspace to compare against different state-estimation algorithms namely VINS-Mono, VINS-Fusion, ORBSLAM3, Stereo-MSCKF, etc.
Stars: ✭ 15 (-84.21%)
Mutual labels:  slam
2019-UGRP-DPoom
2019 DGIST DPoom project under UGRP : SBC and RGB-D camera based full autonomous driving system for mobile robot with indoor SLAM
Stars: ✭ 35 (-63.16%)
Mutual labels:  slam
ndt localizer
This robot lcoalisation package for lidar-map based localisation using multi-sensor state estimation.
Stars: ✭ 32 (-66.32%)
Mutual labels:  slam
VINS-Mono
VINS-Mono中文注释
Stars: ✭ 149 (+56.84%)
Mutual labels:  slam
staticfusion
StaticFusion
Stars: ✭ 107 (+12.63%)
Mutual labels:  slam
awesome-lidar
😎 Awesome LIDAR list. The list includes LIDAR manufacturers, datasets, point cloud-processing algorithms, point cloud frameworks and simulators.
Stars: ✭ 217 (+128.42%)
Mutual labels:  slam
gmmloc
Implementation for IROS2020: "GMMLoc: Structure Consistent Visual Localization with Gaussian Mixture Model"
Stars: ✭ 91 (-4.21%)
Mutual labels:  slam
Rotation Coordinate Descent
(CVPR 2020 Oral) A fast global rotation averaging algorithm.
Stars: ✭ 44 (-53.68%)
Mutual labels:  slam
event-slam-accumulator-settings
Research on Event Accumulator Settings for Event-Based SLAM
Stars: ✭ 16 (-83.16%)
Mutual labels:  slam
vslam
Basic algorithms for vslam.
Stars: ✭ 44 (-53.68%)
Mutual labels:  slam
bht-ams-playerstage
Player/Stage SLAM
Stars: ✭ 35 (-63.16%)
Mutual labels:  slam
G2LTex
Code for CVPR 2018 paper --- Texture Mapping for 3D Reconstruction with RGB-D Sensor
Stars: ✭ 104 (+9.47%)
Mutual labels:  slam
OPVO
Sample code of BMVC 2017 paper: "Visual Odometry with Drift-Free Rotation Estimation Using Indoor Scene Regularities"
Stars: ✭ 40 (-57.89%)
Mutual labels:  slam
SLAM AND PATH PLANNING ALGORITHMS
This repository contains the solutions to all the exercises for the MOOC about SLAM and PATH-PLANNING algorithms given by professor Claus Brenner at Leibniz University. This repository also contains my personal notes, most of them in PDF format, and many vector graphics created by myself to illustrate the theoretical concepts. Hope you enjoy it! :)
Stars: ✭ 107 (+12.63%)
Mutual labels:  slam
Slam-Dunk-Android
Android implementation of "Fusion of inertial and visual measurements for rgb-d slam on mobile devices"
Stars: ✭ 25 (-73.68%)
Mutual labels:  slam
Awesome-SLAM-Papers
A curated list of SLAM-related papers / SLAM文章速递
Stars: ✭ 65 (-31.58%)
Mutual labels:  slam
View All Similar Projects ➔

VIDO-SLAM

VIDO-SLAM is a Visual-Inertial Dynamic Object SLAM System that is able to estimate the camera poses, perform Visual, Visual-Inertial SLAM with monocular camera, track dynamic objects. We provide demo to run the SLAM system in the Kaist Dataset using monocular camera, with or without IMU.

This software is based on VDO-SLAM , FlowNet, Mask RCNN. MonoDepth2, ORB_SLAM3

Features:

  • Integrate MonoDepth2, FlowNet and MaskRcnn with ROS that can run in SLAM system in real time
  • Visual-Inertial SLAM in real-world outdoor scenarios
  • Track and Estimate dynamic objects motion

1. Demo

2. Related Papers:

  • Jun Zhang, Mina Henein, Robert Mahony and Viorela Ila. VDO-SLAM: A Visual Dynamic Object-aware SLAM System. Submitted to The International Journal of Robotics Research. IJRR (Under Review).
  • Carlos Campos, J. M. M. Montiel and Juan D. Tardós, Inertial-Only Optimization for Visual-Inertial Initialization, ICRA 2020.

3. Dependences

  • ROS
  • PyTorch (Version: 1.4.0)
  • GPU (at least 8GB of memory)
  • OpenCV
  • Pangolin
  • g2o
  • Eigen

4. Building

Clone the repository:

https://github.com/bxh1/VIDO-SLAM.git

The directory structure is as follows

VIDO-SLAM
   ├── README.md
   ├── src
   ├── figure
   └── vido_slam

VIDO-SLAM/src/thirdparty/mask_rcnn must be build with python3 install.py install. This builds and installs the CUDA files needed by this network.

VIDO-SLAM/vido_slam must be build with cmake:

mkdir build
cd build && cmake .. && make -j4

A dynamic library will be built in this folder: vido_slam/lib/libvido_slam.so

VIDO-SLAM/src/thirdparty and VIDO-SLAM/src/realtime_demo are ROS package, build with catkin_make

cd VIDO-SLAM
catkin_make

5. Running the Demo

VIDO-SLAM has two modes: offline and online.

Offline mode needs to run three networks in advance to get data, and then run the following command:

cd VIDO-SLAM/vido_slam/build
./run_vido  path_to_VIDO-SLAM/src/config/kaist_config.yaml

Online mode runs three networks in real time through ROS, and sends the network results into vido-slam system:

cd VIDO-SLAM
source devel/setup.bash
roslaunch demo run_realtime_vido.launch

The VO and VIO switch options are in the config file:

6. TODO

Visual-Inertial optimization

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].