This repository contains an implementation of stereo visual odometry using images from the KITTI Odometry dataset. This project uses stereo depth estimation using the open-cv python package.
You can download the following data files from the KITTI dataset
- odometry_data_set(grayscale) ~ 22GB
- odometry_data_set(velodyne_laser_data) ~ 80GB
- odometry_data_set(calibration_files) ~ 1MB
- odometry_ground_truth_poses ~ 4 MB
download and place them in your data folder as shown in the directory structure section of the readme file.
The list of dependencies for the project can be found on the environment.yml file
To recreate the environment used to develop this project, you can create a conda environment using the environment.yml file provided:
conda env create -f environment.ymlThis will create a new environment named stereo-visual-odometry-env with all the necessary packages installed. You can then activate the environment using the following command:
conda activate stereo-visual-odometry-envOnce the environment is activated, you can run the project and use the packages listed above.
Stereo_Visual_Odometry
├── dataset
│ ├── data_odometry_calib
│ ├── data_odometry_gray
│ ├── data_odometry_poses
│ └── data_odometry_velodyne
├── environment.yml
├── Images
├── README.md
└── Stereo Visual Odometry.ipynb