OpenVINS

https://docs.openvins.com/
Admin
OpenVINS is a research platform for visual-inertial odometry and SLAM, implementing a sliding window...
Tokens:99,893
Snippets:1,136
Trust Score:5.4
Update:4 months ago
Show doc for...
Tokens:
Raw
### Build and Run ROS 1 Simulation

Source: https://docs.openvins.com/dev-docker.html

Inside the Docker bash environment, navigate to the workspace, build the OpenVINS codebase using catkin, source the setup, and run an example simulation.

```bash
cd catkin_ws
catkin build
source devel/setup.bash
rosrun ov_eval plot_trajectories none src/open_vins/ov_data/sim/udel_gore.txt
roslaunch ov_msckf simulation.launch
```

--------------------------------

### Install Python 3.6 and Dependencies for m.css

Source: https://docs.openvins.com/dev-docs.html

Installs Python 3.6 and necessary packages (jinja2, Pygments) for building the documentation site with m.css. Also installs texlive base packages.

```bash
sudo add-apt-repository ppa:deadsnakes/ppa
sudo apt-get update
sudo apt-get install python3.6 python3.6-distutils
curl https://bootstrap.pypa.io/pip/3.6/get-pip.py | sudo python3.6
sudo -H pip3.6 install jinja2 Pygments
sudo apt install texlive-base texlive-latex-extra texlive-fonts-extra texlive-fonts-recommended
```

--------------------------------

### Install ROS2 and Dependencies

Source: https://docs.openvins.com/gs-installing.html

Installs ROS2, necessary packages, and development libraries for building OpenVINS.

```bash
sudo apt update && sudo apt install curl gnupg lsb-release
sudo curl -sSL https://raw.githubusercontent.com/ros/rosdistro/master/ros.key  -o /usr/share/keyrings/ros-archive-keyring.gpg
echo "deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/ros-archive-keyring.gpg] http://packages.ros.org/ros2/ubuntu $(lsb_release -cs) main" | sudo tee /etc/apt/sources.list.d/ros2.list > /dev/null
sudo apt-get update
export ROS2_DISTRO=galactic # dashing=18.04, galactic=20.04
sudo apt install ros-$ROS2_DISTRO-desktop
sudo apt-get install ros-$ROS2_DISTRO-ros2bag ros-$ROS2_DISTRO-rosbag2* # rosbag utilities (seems to be separate)
sudo apt-get install libeigen3-dev libboost-all-dev libceres-dev
```

--------------------------------

### Install Visualization Tools and Process Profiling Data

Source: https://docs.openvins.com/dev-profiling.html

Installs necessary visualization programs and then processes the callgrind output file into a format suitable for visualization, followed by opening the visualization.

```bash
// install viz programs
apt-get install python3 graphviz
apt-get install gir1.2-gtk-3.0 python3-gi python3-gi-cairo graphviz
pip install gprof2dot xdot
// actually process and then viz call file
gprof2dot --format callgrind --strip /tmp/callgrind.txt --output /tmp/callgrind.xdot
xdot /tmp/callgrind.xdot
```

--------------------------------

### Install System Dependencies (Ubuntu)

Source: https://docs.openvins.com/gs-installing-free.html

Installs essential libraries required for building OpenVINS on Ubuntu systems. Ensure you have OpenCV, Eigen3, and Ceres installed.

```bash
sudo apt-get install libeigen3-dev libboost-all-dev libceres-dev
```

--------------------------------

### Install ROS1 Noetic Desktop Full

Source: https://docs.openvins.com/gs-installing.html

Installs the ROS1 Noetic desktop-full package and catkin tools. This is for Ubuntu 20.04 systems. Adjust `ROS1_DISTRO` for other Ubuntu versions.

```bash
sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu $(lsb_release -sc) main" > /etc/apt/sources.list.d/ros-latest.list'

```

```bash
sudo apt-key adv --keyserver 'hkp://keyserver.ubuntu.com:80' --recv-key C1CF6E31E6BADE8868B172B4F42ED6FBAB17C654

```

```bash
sudo apt-get update

```

```bash
export ROS1_DISTRO=noetic # kinetic=16.04, melodic=18.04, noetic=20.04

```

```bash
sudo apt-get install ros-$ROS1_DISTRO-desktop-full

```

```bash
sudo apt-get install python-catkin-tools # ubuntu 16.04, 18.04

```

```bash
sudo apt-get install python3-catkin-tools python3-osrf-pycommon # ubuntu 20.04

```

```bash
sudo apt-get install libeigen3-dev libboost-all-dev libceres-dev

```

--------------------------------

### Install Valgrind and Launch ROS Node

Source: https://docs.openvins.com/dev-profiling.html

Installs the valgrind tool and launches a ROS node with valgrind enabled for profiling processing time. Ensure the launch file is edited to include the valgrind prefix.

```bash
sudo apt install valgrind
roslaunch ov_msckf pgeneva_serial_eth.launch
```

--------------------------------

### Install Docker CE on Ubuntu

Source: https://docs.openvins.com/dev-docker.html

Installs Docker Community Edition, Docker CLI, and containerd.io on Ubuntu systems.

```bash
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg
echo "deb [arch=amd64 signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
sudo apt-get update
sudo apt-get install docker-ce docker-ce-cli containerd.io
```

--------------------------------

### Install NVIDIA Container Toolkit and Verify GPU Access

Source: https://docs.openvins.com/dev-docker.html

Installs NVIDIA Docker 2 for GPU support and verifies the installation by running nvidia-smi within a CUDA container.

```bash
distribution=$(. /etc/os-release;echo $ID$VERSION_ID) \
   && curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add - \
   && curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list
sudo apt-get update
sudo apt-get install -y nvidia-docker2
sudo systemctl restart docker
sudo docker run --rm --gpus all nvidia/cuda:11.0-base nvidia-smi
```

--------------------------------

### Build and Install OpenVINS (ROS-Free)

Source: https://docs.openvins.com/gs-installing-free.html

Clones the OpenVINS repository and builds the ov_msckf library in a ROS-free configuration. This process includes setting up the build directory, configuring CMake, compiling, and installing the library.

```bash
cd ~/github/
git clone https://github.com/rpng/open_vins/
cd open_vins/ov_msckf/
mkdir build && cd build
cmake -DENABLE_ROS=OFF ..
make -j4
sudo make install
```

--------------------------------

### Install Latest Clang Compiler

Source: https://docs.openvins.com/dev-profiling.html

Installs the latest Clang compiler version using an automated script and sets the CC and CXX environment variables. Ensure you have sudo privileges.

```bash
sudo bash -c "$(wget -O - https://apt.llvm.org/llvm.sh)"
export CC=/usr/bin/clang-11
export CXX=/usr/bin/clang++-11
```

--------------------------------

### Install Doxygen and Texlive

Source: https://docs.openvins.com/dev-docs.html

Installs doxygen and texlive with all packages required for building the PDF reference manual. Ensure you have doxygen version 1.9.4 or newer to avoid ghostscript errors.

```bash
sudo apt-get update
sudo apt-get install doxygen texlive-full
```

--------------------------------

### Build and Run ROS 2 Simulation

Source: https://docs.openvins.com/dev-docker.html

Inside the Docker bash environment, build the OpenVINS codebase using colcon for ROS 2, source the setup, and run an example simulation with the specified configuration.

```bash
cd catkin_ws
colcon build --event-handlers console_cohesion+
source install/setup.bash
ros2 run ov_eval plot_trajectories none src/open_vins/ov_data/sim/udel_gore.txt
ros2 run ov_msckf run_simulation src/open_vins/config/rpng_sim/estimator_config.yaml
```

--------------------------------

### Example Compile Time Analysis Results

Source: https://docs.openvins.com/dev-profiling.html

An example of the output generated in timing_results.txt after analyzing the build trace. It shows time summaries for compilation, parsing, codegen, and lists files and functions that took the longest to compile.

```text
Analyzing build trace from 'capture_file.bin'...
     Time summary:
Compilation (86 times):
  Parsing (frontend):          313.9 s
  Codegen & opts (backend):    222.9 s

     Files that took longest to parse (compiler frontend):
 13139 ms: /build//ov_msckf/CMakeFiles/ov_msckf_lib.dir/src/update/UpdaterSLAM.cpp.o
 12843 ms: /build//ov_msckf/CMakeFiles/run_serial_msckf.dir/src/ros_serial_msckf.cpp.o
 ...
 
     Functions that took longest to compile:
  1639 ms: main (/src/open_vins/ov_eval/src/error_comparison.cpp)
  1337 ms: ov_core::BsplineSE3::get_acceleration(double, Eigen::Matrix<double, ... (/src/open_vins/ov_core/src/sim/BsplineSE3.cpp)
  1156 ms: ov_eval::ResultSimulation::plot_state(bool, double) (/src/open_vins/ov_eval/src/calc/ResultSimulation.cpp)
  ...

     Expensive headers: 
 27505 ms: /src/open_vins/ov_core/src/track/TrackBase.h (included 12 times, avg 2292 ms), included via:
   TrackKLT.cpp.o TrackKLT.h  (4372 ms)
   TrackBase.cpp.o  (4297 ms)
   TrackSIM.cpp.o TrackSIM.h  (4252 ms)
   ...

```

--------------------------------

### Install rosbag2 and Plugins

Source: https://docs.openvins.com/dev-ros1-to-ros2.html

Install the necessary rosbag2 packages and plugins for ROS2 bag manipulation. This command is for ROS-$ROS2_DISTRO systems.

```bash
sudo apt-get install ros-$ROS2_DISTRO-rosbag2 ros-$ROS2_DISTRO-rosbag2*
sudo apt install ros-$ROS2_DISTRO-rosbag2-bag-v2-plugins
```

--------------------------------

### ROS 1 Launch and Run Commands

Source: https://docs.openvins.com/gs-tutorial.html

Commands to set up and run the OpenVINS system in ROS 1. This includes starting the ROS master, sourcing the environment, launching the estimator, and playing a ROS bag file.

```bash
roscore # term 0
```

```bash
source devel/setup.bash # term 1
```

```bash
roslaunch ov_msckf subscribe.launch config:=euroc_mav
```

```bash
rviz -d ov_msckf/launch/display.rviz # term 2
```

```bash
rosbag play V1_01_easy.bag # term 3
```

--------------------------------

### Install Ceres Solver from Source

Source: https://docs.openvins.com/gs-installing.html

Installs Ceres Solver from source, including necessary dependencies. This method is recommended when system packages are insufficient or a newer version is required. It addresses potential Eigen version mismatches.

```bash
sudo apt-get install -y cmake libgoogle-glog-dev libgflags-dev libatlas-base-dev libeigen3-dev libsuitesparse-dev
CERES_VERSION="2.0.0"
git clone https://ceres-solver.googlesource.com/ceres-solver
cd ceres-solver
git checkout tags/${CERES_VERSION}
mkdir build && cd build
cmake ..
make
sudo make install
```

--------------------------------

### Configure CMakeLists.txt for OpenVINS

Source: https://docs.openvins.com/gs-installing-free.html

Sets up include directories and library paths in your project's CMakeLists.txt to link against the installed OpenVINS library. Ensure the paths to headers and libraries are correct.

```cmake
# headers: /usr/local/include/open_vins/
# library: /usr/local/lib/libov_msckf_lib.so
set(OPENVINS_INCLUDE_DIR /usr/local/include/open_vins/)
set(OPENVINS_LIBRARIES ov_msckf_lib)
```

--------------------------------

### Get True Parameters

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Provides access to the true simulation parameters, including calibration and settings.

```cpp
auto get_true_parameters() -> InertialInitializerOptions
```

--------------------------------

### Clone OpenVINS and Build Docker Image

Source: https://docs.openvins.com/dev-docker.html

Prepare a ROS workspace, clone the OpenVINS repository, and build the Docker image. Ensure you select the correct version tag for your ROS and Ubuntu setup.

```bash
mkdir -p ~/workspace/catkin_ws_ov/src
cd ~/workspace/catkin_ws_ov/src
git clone https://github.com/rpng/open_vins.git
cd open_vins
export VERSION=ros1_20_04 # which docker file version you want (ROS1 vs ROS2 and ubuntu version)
docker build -t ov_$VERSION -f Dockerfile_$VERSION .
```

--------------------------------

### Build OpenCV from Source

Source: https://docs.openvins.com/gs-installing.html

Clone OpenCV and its contrib modules, then build and install them. Ensure the contrib modules are included during the CMake configuration step.

```bash
git clone https://github.com/opencv/opencv/
git clone https://github.com/opencv/opencv_contrib/
mkdir opencv/build/ 
cd opencv/build/ 
cmake -DOPENCV_EXTRA_MODULES_PATH=../../opencv_contrib/modules ..
make -j8
sudo make install
```

--------------------------------

### Configure ROS1 and ROS2 Aliases for Bash

Source: https://docs.openvins.com/gs-installing.html

Sets up bash aliases for sourcing ROS1 and ROS2 environments, and for sourcing the install directory within a workspace.

```bash
echo "alias ros2_source=\"source /opt/ros/$ROS2_DISTRO/setup.bash\"" >> ~/.bashrc
echo "alias source_install=\"source install/setup.bash\"" >> ~/.bashrc
source ~/.bashrc
```

--------------------------------

### Example Output of Error Comparison

Source: https://docs.openvins.com/eval-error.html

Illustrates the console output generated by the error_comparison script, showing dataset processing, algorithm evaluation, and the final ATE and RPE LaTeX tables.

```text
[ INFO] [1583425216.054023187]: [COMP]: 2895 poses in V1_01_easy.txt => length of 58.35 meters
[ INFO] [1583425216.092355692]: [COMP]: 16702 poses in V1_02_medium.txt => length of 75.89 meters
[ INFO] [1583425216.133532429]: [COMP]: 20932 poses in V1_03_difficult.txt => length of 78.98 meters
[ INFO] [1583425216.179616651]: [COMP]: 22401 poses in V2_01_easy.txt => length of 36.50 meters
[ INFO] [1583425216.225299463]: [COMP]: 23091 poses in V2_02_medium.txt => length of 83.23 meters
[ INFO] [1583425216.225660364]: ======================================
[ INFO] [1583425223.560550101]: [COMP]: processing mono_ov_vio algorithm
[ INFO] [1583425223.560632706]: [COMP]: processing mono_ov_vio algorithm => V1_01_easy dataset
[ INFO] [1583425224.236769465]: [COMP]: processing mono_ov_vio algorithm => V1_02_medium dataset
[ INFO] [1583425224.855489521]: [COMP]: processing mono_ov_vio algorithm => V1_03_difficult dataset
[ INFO] [1583425225.659183593]: [COMP]: processing mono_ov_vio algorithm => V2_01_easy dataset
[ INFO] [1583425226.442217424]: [COMP]: processing mono_ov_vio algorithm => V2_02_medium dataset
[ INFO] [1583425227.366004330]: ======================================
....
[ INFO] [1583425261.724448413]: ============================================
[ INFO] [1583425261.724469372]: ATE LATEX TABLE
[ INFO] [1583425261.724481841]: ============================================
 & \textbf{V1_01_easy} & \textbf{V1_02_medium} & \textbf{V1_03_difficult}
 & \textbf{V2_01_easy} & \textbf{V2_02_medium} & \textbf{Average} \\hline
mono_ov_slam & 0.699 / 0.058 & 1.675 / 0.076 & 2.542 / 0.063 & 0.773 / 0.124 & 1.538 / 0.074 & 1.445 / 0.079 \ 
mono_ov_vio & 0.642 / 0.076 & 1.766 / 0.096 & 2.391 / 0.344 & 1.164 / 0.121 & 1.248 / 0.106 & 1.442 / 0.148 \ 
....
[ INFO] [1583425261.724647970]: ============================================
[ INFO] [1583425261.724655060]: ============================================
[ INFO] [1583425261.724661046]: RPE LATEX TABLE
[ INFO] [1583425261.724666910]: ============================================
 & \textbf{8m} & \textbf{16m} & \textbf{24m} & \textbf{32m} & \textbf{40m} & \textbf{48m} \\hline
mono_ov_slam & 0.661 / 0.074 & 0.802 / 0.086 & 0.979 / 0.097 & 1.061 / 0.105 & 1.145 / 0.120 & 1.289 / 0.122 \ 
mono_ov_vio & 0.826 / 0.094 & 1.039 / 0.106 & 1.215 / 0.111 & 1.283 / 0.132 & 1.342 / 0.151 & 1.425 / 0.184 \ 
....
[ INFO] [1583425262.514587296]: ============================================

```

--------------------------------

### compute_comparison_indices_length

Source: https://docs.openvins.com/classov__eval_1_1ResultTrajectory.html

Gets the indices at the end of subtractories of a given length when starting at each index.

```APIDOC
## compute_comparison_indices_length

### Description
Gets the indices at the end of subtractories of a given length when starting at each index. For each starting pose, find the end pose index which is the desired distance away.

### Parameters
- **distances** (std::vector<double>&) - Total distance travelled from start at each index
- **distance** (double) - Distance of subtrajectory
- **max_dist_diff** (double) - Maximum error between current trajectory length and the desired

### Returns
std::vector<int> - End indices for each subtrajectory
```

--------------------------------

### Build Documentation Site with m.css

Source: https://docs.openvins.com/dev-docs.html

Builds the documentation website using the m.css scripts. Requires cloning the m.css repository and running the doxygen.py script with the path to the Doxyfile.

```bash
cd m.css/documentation/
python3.6 doxygen.py <path_to_Doxyfile-mcss>
python3.6 doxygen.py ~/catkin_ws/src/open_vins/Doxyfile-mcss
```

--------------------------------

### Alias ROS1 and Workspace Sourcing (Bashrc)

Source: https://docs.openvins.com/gs-installing.html

Creates bash aliases for sourcing ROS1 environment variables and local workspace setup. Useful when both ROS1 and ROS2 are installed.

```bash
echo "alias ros1_source=\"source /opt/ros/$ROS1_DISTRO/setup.bash\"" >> ~/.bashrc

```

```bash
echo "alias source_devel=\"source devel/setup.bash\"" >> ~/.bashrc

```

```bash
source ~/.bashrc
```

--------------------------------

### initialize Method

Source: https://docs.openvins.com/classov__init_1_1DynamicInitializer.html

Attempts to initialize the visual-inertial system, recovering initial conditions and system covariance.

```APIDOC
## initialize Method

### Description
Try to get the initialized system. This method attempts to recover the initial conditions and covariance of the visual-inertial system.

### Parameters
* **timestamp** (double&) - Output parameter for the Timestamp we have initialized the state at (last imu state).
* **covariance** (Eigen::MatrixXd&) - Output parameter for the Calculated covariance of the returned state.
* **order** (std::vector<std::shared_ptr<ov_type::Type>>&) - Output parameter for the Order of the covariance matrix.
* **_imu** (std::shared_ptr<ov_type::IMU>&) - Pointer to the "active" IMU state (q_GtoI, p_IinG, v_IinG, bg, ba).
* **_clones_IMU** (std::map<double, std::shared_ptr<ov_type::PoseJPL>>&) - Map between imaging times and clone poses (q_GtoIi, p_IiinG).
* **_features_SLAM** (std::unordered_map<size_t, std::shared_ptr<ov_type::Landmark>>&) - Our current set of SLAM features (3d positions).

### Returns
* **bool** - True if we have successfully initialized our system.
```

--------------------------------

### setup_subscribers

Source: https://docs.openvins.com/classov__msckf_1_1ROS2Visualizer.html

Sets up ROS subscribers and their corresponding callbacks.

```APIDOC
## void setup_subscribers(std::shared_ptr<ov_core::YamlParser> parser)

### Description
Will setup ROS subscribers and callbacks.

### Parameters
* **parser** (std::shared_ptr<ov_core::YamlParser>) - Configuration file parser
```

--------------------------------

### SimulatorInit Constructor

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Initializes the SimulatorInit class, loading all configuration variables from the provided parameters.

```APIDOC
## SimulatorInit(InertialInitializerOptions& params_)

### Description
Default constructor, will load all configuration variables.

### Parameters
- **params_** (InertialInitializerOptions&) - Required - InertialInitializer parameters. Should have already been loaded from cmd.
```

--------------------------------

### Install Python Dependencies for Matplotlib and psutil

Source: https://docs.openvins.com/gs-installing.html

Installs Python development headers and libraries required for matplotlib-cpp visualization and psutil for system monitoring.

```bash
sudo apt-get install python2.7-dev python-matplotlib python-numpy python-psutil # for python2 systems
sudo apt-get install python3-dev python3-matplotlib python3-numpy python3-psutil python3-tk # for python3 systems
```

--------------------------------

### Example Output of error_singlerun

Source: https://docs.openvins.com/eval-error.html

This is an example of the detailed output generated by the 'error_singlerun' script, showing Absolute Trajectory Error and Relative Pose Error statistics.

```text
[ INFO] [1583422165.069376426]: [COMP]: 2813 poses in 1565371553_estimate => length of 57.36 meters
[ INFO] [1583422165.091423722]: ======================================
[ INFO] [1583422165.091438299]: Absolute Trajectory Error
[ INFO] [1583422165.091453099]: ======================================
[ INFO] [1583422165.091459679]: rmse_ori = 0.677 | rmse_pos = 0.055
[ INFO] [1583422165.091466321]: mean_ori = 0.606 | mean_pos = 0.051
[ INFO] [1583422165.091474211]: min_ori  = 0.044 | min_pos  = 0.001
[ INFO] [1583422165.091481730]: max_ori  = 1.856 | max_pos  = 0.121
[ INFO] [1583422165.091481730]: std_ori  = 0.302 | std_pos  = 0.021
[ INFO] [1583422165.127869924]: ======================================
[ INFO] [1583422165.127891080]: Relative Pose Error
[ INFO] [1583422165.127898322]: ======================================
[ INFO] [1583422165.127908551]: seg 8 - median_ori = 0.566 | median_pos = 0.068 (2484 samples)
[ INFO] [1583422165.127919603]: seg 16 - median_ori = 0.791 | median_pos = 0.060 (2280 samples)
[ INFO] [1583422165.127927646]: seg 24 - median_ori = 0.736 | median_pos = 0.070 (2108 samples)
[ INFO] [1583422165.127934904]: seg 32 - median_ori = 0.715 | median_pos = 0.071 (1943 samples)
[ INFO] [1583422165.127942178]: seg 40 - median_ori = 0.540 | median_pos = 0.063 (1792 samples)
```

--------------------------------

### sim_distance_threshold

Source: https://docs.openvins.com/structov__msckf_1_1VioManagerOptions.html

Distance threshold for starting simulation. Simulation will commence after moving this distance along the b-spline, preventing static starts when initializing from ground truth in simulation.

```APIDOC
## double ov_msckf::VioManagerOptions::sim_distance_threshold

### Description
We will start simulating after we have moved this much along the b-spline. This prevents static starts as we init from groundtruth in simulation.
```

--------------------------------

### initialize_with_gt

Source: https://docs.openvins.com/classov__msckf_1_1VioManager.html

Initializes the IMU state of the system using provided ground truth data.

```APIDOC
## initialize_with_gt

### Description
Given a state, this will initialize our IMU state.

### Method
void

### Parameters
#### Path Parameters
- **imustate** (Eigen::Matrix<double, 17, 1>) - Required - The ground truth IMU state to initialize with
```

--------------------------------

### Valgrind Memory Leak Report Example

Source: https://docs.openvins.com/dev-profiling.html

An example of a memory leak report generated by valgrind's memcheck tool, indicating bytes lost and the call stack leading to the leak.

```text
==5512== 1,578,860 (24 direct, 1,578,836 indirect) bytes in 1 blocks are definitely lost in loss record 6,585 of 6,589
==5512==    at 0x4C3017F: operator new(unsigned long) (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
....
==5512==    by 0x543F868: operator[] (unordered_map.h:973)
==5512==    by 0x543F868: ov_core::TrackKLT::feed_stereo(double, cv::Mat&, cv::Mat&, unsigned long, unsigned long) (TrackKLT.cpp:165)
==5512==    by 0x4EF8C52: ov_msckf::VioManager::feed_measurement_stereo(double, cv::Mat&, cv::Mat&, unsigned long, unsigned long) (VioManager.cpp:245)
==5512==    by 0x1238A9: main (ros_serial_msckf.cpp:247)
```

--------------------------------

### Recorder Constructor

Source: https://docs.openvins.com/classov__eval_1_1Recorder.html

Initializes the Recorder by opening a specified file for output. It automatically creates the directory path if it does not exist.

```APIDOC
## Recorder(std::string filename)

### Description
Default constructor that will open the specified file on disk. If the output directory does not exist, this will also create the directory path to this file.

### Parameters
#### Path Parameters
- **filename** (std::string) - Required - Desired file we want to "record" into
```

--------------------------------

### Run Gazebo GUI within Docker with GPU Support

Source: https://docs.openvins.com/dev-docker.html

Launches the Gazebo GUI from within a Docker container, enabling GPU acceleration and X11 forwarding for GUI display on the host.

```bash
docker run -it --net=host --gpus all \
    --env="NVIDIA_DRIVER_CAPABILITIES=all" \
    --env="DISPLAY" \
    --env="QT_X11_NO_MITSHM=1" \
    --volume="/tmp/.X11-unix:/tmp/.X11-unix:rw" \
    osrf/ros:noetic-desktop-full \
    bash -it -c "roslaunch gazebo_ros empty_world.launch"
```

--------------------------------

### Example Output of error_dataset script

Source: https://docs.openvins.com/eval-error.html

This is an example of the console output generated by the error_dataset script. It includes detailed metrics such as ATE (Absolute Trajectory Error) and RPE (Relative Pose Error) for each processed algorithm.

```text
[ INFO] [1583422654.333642977]: ======================================
[ INFO] [1583422654.333915102]: [COMP]: processing mono_ov_slam algorithm
[ INFO] [1583422654.362655719]: [TRAJ]: q_ESTtoGT = 0.000, 0.000, -0.129, 0.992 | p_ESTinGT = 0.978, 2.185, 0.932 | s = 1.00
....
[ INFO] [1583422654.996859432]: [TRAJ]: q_ESTtoGT = 0.000, 0.000, -0.137, 0.991 | p_ESTinGT = 0.928, 2.166, 0.957 | s = 1.00
[ INFO] [1583422655.041009388]:     ATE: mean_ori = 0.684 | mean_pos = 0.057
[ INFO] [1583422655.041031730]:     ATE: std_ori  = 0.14938 | std_pos  = 0.01309
[ INFO] [1583422655.041039552]:     ATE 2D: mean_ori = 0.552 | mean_pos = 0.053
[ INFO] [1583422655.041046362]:     ATE 2D: std_ori  = 0.17786 | std_pos  = 0.01421
[ INFO] [1583422655.044187033]:     RPE: seg 7 - mean_ori = 0.543 | mean_pos = 0.065 (25160 samples)
[ INFO] [1583422655.047047771]:     RPE: seg 14 - mean_ori = 0.593 | mean_pos = 0.060 (23470 samples)
[ INFO] [1583422655.049672955]:     RPE: seg 21 - mean_ori = 0.664 | mean_pos = 0.081 (22050 samples)
[ INFO] [1583422655.052090494]:     RPE: seg 28 - mean_ori = 0.732 | mean_pos = 0.083 (20520 samples)
[ INFO] [1583422655.054294322]:     RPE: seg 35 - mean_ori = 0.793 | mean_pos = 0.090 (18960 samples)
[ INFO] [1583422655.055676035]:     RMSE: mean_ori = 0.644 | mean_pos = 0.056
[ INFO] [1583422655.056987984]:     RMSE 2D: mean_ori = 0.516 | mean_pos = 0.052
[ INFO] [1583422655.058269163]:     NEES: mean_ori = 793.646 | mean_pos = 13.095
[ INFO] [1583422656.182660653]: ======================================
[ INFO] [1583422656.183065588]: [COMP]: processing mono_ov_vio algorithm
[ INFO] [1583422656.209545279]: [TRAJ]: q_ESTtoGT = 0.000, 0.000, -0.148, 0.989 | p_ESTinGT = 0.931, 2.169, 0.971 | s = 1.00
....
[ INFO] [1583422656.791523636]: [TRAJ]: q_ESTtoGT = 0.000, 0.000, -0.149, 0.989 | p_ESTinGT = 0.941, 2.163, 0.974 | s = 1.00
[ INFO] [1583422656.835407991]:     ATE: mean_ori = 0.639 | mean_pos = 0.076
[ INFO] [1583422656.835433475]:     ATE: std_ori  = 0.05800 | std_pos  = 0.00430
[ INFO] [1583422656.835446222]:     ATE 2D: mean_ori = 0.514 | mean_pos = 0.070
[ INFO] [1583422656.835457020]:     ATE 2D: std_ori  = 0.07102 | std_pos  = 0.00492
[ INFO] [1583422656.838656567]:     RPE: seg 7 - mean_ori = 0.614 | mean_pos = 0.092 (25160 samples)
[ INFO] [1583422656.841540191]:     RPE: seg 14 - mean_ori = 0.634 | mean_pos = 0.092 (23470 samples)
[ INFO] [1583422656.844219466]:     RPE: seg 21 - mean_ori = 0.632 | mean_pos = 0.115 (22050 samples)
[ INFO] [1583422656.846646272]:     RPE: seg 28 - mean_ori = 0.696 | mean_pos = 0.119 (20520 samples)
[ INFO] [1583422656.848862913]:     RPE: seg 35 - mean_ori = 0.663 | mean_pos = 0.154 (18960 samples)
[ INFO] [1583422656.850321777]:     RMSE: mean_ori = 0.600 | mean_pos = 0.067
[ INFO] [1583422656.851673985]:     RMSE 2D: mean_ori = 0.479 | mean_pos = 0.060
[ INFO] [1583422656.853037942]:     NEES: mean_ori = 625.447 | mean_pos = 10.629
[ INFO] [1583422658.194763413]: ======================================
....
```

--------------------------------

### ROS2Visualizer Constructor

Source: https://docs.openvins.com/classov__msckf_1_1ROS2Visualizer.html

Initializes the ROS2Visualizer with ROS node, VIO manager, and an optional simulator.

```APIDOC
## ROS2Visualizer(std::shared_ptr<rclcpp::Node> node, std::shared_ptr<VioManager> app, std::shared_ptr<Simulator> sim = nullptr)

### Description
Default constructor.

### Parameters
* **node** (std::shared_ptr<rclcpp::Node>) - ROS node pointer
* **app** (std::shared_ptr<VioManager>) - Core estimator manager
* **sim** (std::shared_ptr<Simulator>) - Simulator if we are simulating
```

--------------------------------

### current_timestamp

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Gets the current timestamp of the simulation.

```APIDOC
## double current_timestamp()

### Description
Gets the timestamp we have simulated up too.

### Returns
- **double** - Timestamp
```

--------------------------------

### ov_core::CamBase Constructor

Source: https://docs.openvins.com/classov__core_1_1CamBase.html

Initializes the base camera model with specified image dimensions.

```APIDOC
## ov_core::CamBase::CamBase(int width, int height)

### Description
Default constructor.

### Parameters
#### Path Parameters
- **width** (int) - Required - Width of the camera (raw pixels)
- **height** (int) - Required - Height of the camera (raw pixels)
```

--------------------------------

### Get Current Timestamp

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Retrieves the current timestamp of the simulation.

```cpp
auto current_timestamp() -> double
```

--------------------------------

### Perturb Parameters

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Gets a set of perturbed parameters for simulation.

```cpp
void perturb_parameters(InertialInitializerOptions& params_)
```

--------------------------------

### get_num_features

Source: https://docs.openvins.com/classov__core_1_1TrackBase.html

Gets the current number of active features being tracked.

```APIDOC
## auto get_num_features() -> int

### Description
Getter method for number of active features.

### Returns
- int: The number of active features.
```

--------------------------------

### initialized

Source: https://docs.openvins.com/classov__msckf_1_1VioManager.html

Checks if the VIO system has been initialized.

```APIDOC
## initialized

### Description
If we are initialized or not.

### Method
auto -> bool

### Returns
- bool - True if the system is initialized, false otherwise.
```

--------------------------------

### ROS1Visualizer Constructor

Source: https://docs.openvins.com/classov__msckf_1_1ROS1Visualizer.html

Initializes the ROS1Visualizer with ROS node handle, VIO manager, and an optional simulator.

```APIDOC
## ROS1Visualizer(std::shared_ptr<ros::NodeHandle> nh, std::shared_ptr<VioManager> app, std::shared_ptr<Simulator> sim = nullptr)

### Description
Default constructor.

### Parameters
* **nh** (std::shared_ptr<ros::NodeHandle>) - ROS node handler
* **app** (std::shared_ptr<VioManager>) - Core estimator manager
* **sim** (std::shared_ptr<Simulator>) - Simulator if we are simulating
```

--------------------------------

### get_start_time

Source: https://docs.openvins.com/classov__core_1_1BsplineSE3.html

Returns the simulation start time from which the spline should be considered valid.

```APIDOC
## auto get_start_time() -> double

### Description
Returns the simulation start time that we should start simulating from.

### Returns
- **double** - The simulation start time.

### Method
`get_start_time`
```

--------------------------------

### successful

Source: https://docs.openvins.com/classov__core_1_1YamlParser.html

Checks if all parameters were successfully read from the configuration files.

```APIDOC
## successful

### Description
Check to see if all parameters were read succesfully.

### Returns
- **bool** - True if all parameters were found, false otherwise.
```

--------------------------------

### get_last_ids

Source: https://docs.openvins.com/classov__core_1_1TrackBase.html

Gets the IDs of active features in the last frame, per camera.

```APIDOC
## auto get_last_ids() -> std::unordered_map<size_t, std::vector<size_t>>

### Description
Getter method for active features in the last frame (ids per camera).

### Returns
- std::unordered_map<size_t, std::vector<size_t>>: A map where keys are camera IDs and values are vectors of feature IDs.
```

--------------------------------

### YamlParser Constructor

Source: https://docs.openvins.com/classov__core_1_1YamlParser.html

Initializes the YamlParser by loading configuration files from the specified path. It can optionally fail the program if the configuration file is not found.

```APIDOC
## YamlParser Constructor

### Description
Constructor that loads all three configuration files.

### Parameters
- **config_path** (string) - Required - Path to the YAML file we will parse.
- **fail_if_not_found** (bool) - Optional - If we should terminate the program if we can't open the config file. Defaults to true.
```

--------------------------------

### get_last_obs

Source: https://docs.openvins.com/classov__core_1_1TrackBase.html

Gets the observations (KeyPoints) of active features in the last frame, per camera.

```APIDOC
## auto get_last_obs() -> std::unordered_map<size_t, std::vector<cv::KeyPoint>>

### Description
Getter method for active features in the last frame (observations per camera).

### Returns
- std::unordered_map<size_t, std::vector<cv::KeyPoint>>: A map where keys are camera IDs and values are vectors of KeyPoints.
```

--------------------------------

### Get True Feature Map

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Returns the ground truth 3D map of features used in the simulation.

```cpp
auto get_map() -> std::unordered_map<size_t, Eigen::Vector3d>
```

--------------------------------

### InertialInitializerOptions::print_and_load_initializer

Source: https://docs.openvins.com/structov__init_1_1InertialInitializerOptions.html

Prints all initializer settings that have been loaded. This is useful for visually verifying that parameters were loaded correctly from ROS or command-line parsers.

```APIDOC
## void ov_init::InertialInitializerOptions::print_and_load_initializer(const std::shared_ptr<ov_core::YamlParser>& parser = nullptr)

### Description
This function will load print out all initializer settings loaded. This allows for visual checking that everything was loaded properly from ROS/CMD parsers.

### Parameters
* **parser** (std::shared_ptr<ov_core::YamlParser>) - If not null, this parser will be used to load our parameters.
```

--------------------------------

### Get Next Camera Measurement

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Retrieves the next camera measurement, including timestamp, camera IDs, and feature observations.

```cpp
auto get_next_cam(double& time_cam, std::vector<int>& camids, std::vector<std::vector<std::pair<size_t, Eigen::VectorXf>>>& feats) -> bool
```

--------------------------------

### config

Source: https://docs.openvins.com/classov__core_1_1FeatureInitializer.html

Retrieves the current configuration settings of the feature initializer.

```APIDOC
## config

### Description
Gets the current configuration of the feature initializer.

### Method
N/A (This is a class method)

### Returns
- **const FeatureInitializerOptions** - Const feature initializer config
```

--------------------------------

### Configure ROS2 Environment for Bash

Source: https://docs.openvins.com/gs-installing.html

Appends the ROS2 setup script to your bashrc for automatic environment loading in new terminals.

```bash
echo "source /opt/ros/$ROS2_DISTRO/setup.bash" >> ~/.bashrc
source ~/.bashrc
```

--------------------------------

### print_and_load_initializer

Source: https://docs.openvins.com/structov__init_1_1InertialInitializerOptions.html

Loads and prints all initializer settings. This function aids in visual verification of loaded settings from ROS/CMD parsers.

```APIDOC
## void print_and_load_initializer(const std::shared_ptr<ov_core::YamlParser>& parser = nullptr)

### Description
This function will load and print out all initializer settings. This allows for visual checking that everything was loaded properly from ROS/CMD parsers.

### Parameters
* **parser** (std::shared_ptr<ov_core::YamlParser>) - Optional - A shared pointer to a YamlParser object to load parameters from.
```

--------------------------------

### ROS 1 Launch File Configuration

Source: https://docs.openvins.com/gs-tutorial.html

This ROS 1 launch file configures the MSCKF estimation node, setting parameters like verbosity, configuration path, stereo usage, and maximum cameras. It's used to launch the `run_subscribe_msckf` node.

```xml
<launch>

    <!-- what config we are going to run (should match folder name) -->
    <arg name="verbosity"   default="INFO" /> <!-- ALL, DEBUG, INFO, WARNING, ERROR, SILENT -->
    <arg name="config"      default="euroc_mav" /> <!-- euroc_mav, tum_vi, rpng_aruco -->
    <arg name="config_path" default="$(find ov_msckf)/../config/$(arg config)/estimator_config.yaml" />

    <!-- MASTER NODE! -->
    <node name="run_subscribe_msckf" pkg="ov_msckf" type="run_subscribe_msckf" output="screen">
        <param name="verbosity"    type="string" value="$(arg verbosity)" />
        <param name="config_path"  type="string" value="$(arg config_path)" />
        <param name="use_stereo"   type="bool"   value="true" />
        <param name="max_cameras"  type="int"    value="2" />
    </node>

</launch>
```

--------------------------------

### initialize_with_gt

Source: https://docs.openvins.com/classov__msckf_1_1VioManager.html

Initializes the IMU state using a provided ground truth state. The state includes time, IMU to world rotation, position, velocity, gyroscope bias, and accelerometer bias.

```APIDOC
## initialize_with_gt

### Description
Given a state, this will initialize our IMU state.

### Method
void

### Parameters
#### Path Parameters
- **imustate** (Eigen::Matrix<double, 17, 1>) - Required - State in the MSCKF ordering: [time(sec),q_GtoI,p_IinG,v_IinG,b_gyro,b_accel]
```

--------------------------------

### Clone KAIST Datasets Repositories

Source: https://docs.openvins.com/gs-datasets.html

Clone the necessary repositories for processing KAIST dataset formats into ROS bags. Ensure you have git installed.

```bash
git clone https://github.com/irapkaist/irp_sen_msgs.git
git clone https://github.com/rpng/kaist2bag.git
```

--------------------------------

### StaticInitializer Constructor

Source: https://docs.openvins.com/classov__init_1_1StaticInitializer.html

Initializes the StaticInitializer. This constructor assumes the IMU starts from a standstill and uses inertial measurements to set up the initial orientation and biases.

```APIDOC
## StaticInitializer(InertialInitializerOptions& params_, std::shared_ptr<ov_core::FeatureDatabase> db, std::shared_ptr<std::vector<ov_core::ImuData>> imu_data_) explicit

### Description
Default constructor.

### Parameters
* **params_** (InertialInitializerOptions&) - Parameters loaded from either ROS or CMDLINE.
* **db** (std::shared_ptr<ov_core::FeatureDatabase>) - Feature tracker database with all features in it.
* **imu_data_** (std::shared_ptr<std::vector<ov_core::ImuData>>) - Shared pointer to our IMU vector of historical information.
```

--------------------------------

### Get Next IMU Reading

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Fetches the next available IMU reading, including time, angular velocity (wm), and linear acceleration (am).

```cpp
auto get_next_imu(double& time_imu, Eigen::Vector3d& wm, Eigen::Vector3d& am) -> bool
```

--------------------------------

### ov_core::TrackBase::TrackBase

Source: https://docs.openvins.com/classov__core_1_1TrackBase.html

Initializes the TrackBase with specified camera configurations, feature counts, and tracking modes.

```APIDOC
## ov_core::TrackBase::TrackBase

### Description
Public constructor with configuration variables.

### Parameters
- **cameras** (std::unordered_map<size_t, std::shared_ptr<CamBase>>) - camera calibration object which has all camera intrinsics in it
- **numfeats** (int) - number of features we want want to track (i.e. track 200 points from frame to frame)
- **numaruco** (int) - the max id of the arucotags, so we ensure that we start our non-auroc features above this value
- **stereo** (bool) - if we should do stereo feature tracking or binocular
- **histmethod** (HistogramMethod) - what type of histogram pre-processing should be done (histogram eq?)
```

--------------------------------

### ov_core::CamEqui::CamEqui

Source: https://docs.openvins.com/classov__core_1_1CamEqui.html

Default constructor for the CamEqui camera model. Initializes the camera with specified width and height.

```APIDOC
## ov_core::CamEqui::CamEqui(int width, int height)

### Description
Default constructor.

### Parameters
- **width** (int) - Required - Width of the camera (raw pixels)
- **height** (int) - Required - Height of the camera (raw pixels)
```

--------------------------------

### CpiBase Constructor

Source: https://docs.openvins.com/classov__core_1_1CpiBase.html

Initializes the CpiBase with specified noise densities for gyroscope and accelerometer, and an option for averaging IMU measurements.

```APIDOC
## CpiBase Constructor

### Description
Default constructor. Initializes the preintegration with gyroscope and accelerometer noise parameters.

### Parameters
- **sigma_w** (double) - gyroscope white noise density (rad/s/sqrt(hz))
- **sigma_wb** (double) - gyroscope random walk (rad/s^2/sqrt(hz))
- **sigma_a** (double) - accelerometer white noise density (m/s^2/sqrt(hz))
- **sigma_ab** (double) - accelerometer random walk (m/s^3/sqrt(hz))
- **imu_avg_** (bool) - Optional. If true, averages IMU measurements. Defaults to false.
```

--------------------------------

### Get Simulation State at Desired Time

Source: https://docs.openvins.com/classov__init_1_1SimulatorInit.html

Retrieves the simulation state at a specified timestep. The state includes time, rotation, position, velocity, and biases.

```cpp
auto get_state(double desired_time, Eigen::Matrix<double, 17, 1>& imustate) -> bool
```
Tokens:
Raw