Carla Car Monocular Camera 60 FPS
Video dataset simulated by an RGB monocular camera attached to the front of a virtual car in Carla Simulator, with time-stamped sequences for SLAM and odometry.
36,000 time-stamped PNG image files (nanosecond timstamps), YAML configuration files, and text
MIT
Description
The dataset Carla Car Monocular Camera 60 FPS contains video footage in the form of PNG images captured at 60 frames per second by a monocular camera placed at the front of a vehicle in the simulated urban environment Carla (Unreal Engine 5.5).
The images are accompanied by timestamp and configuration files for ORB-SLAM3, allowing use for SLAM (Simultaneous Localization and Mapping), odometry, and image segmentation tasks.
What is this dataset for?
- Develop and test monocular SLAM and visual odometry algorithms
- Train segmentation models for images from embedded cameras
- Simulate the perception of an autonomous vehicle in a virtual urban environment
Can it be enriched or improved?
Yes, this dataset can be enriched by adding semantic annotations (more detailed segmentation labels), simulated depth data, or GPS/IMU metadata if available. Documentation and comments can be translated for better accessibility.
🔎 In summary
🧠 Recommended for
- Robotics researchers
- Computer vision
- Autonomous driving
🔧 Compatible tools
- ORB-SLAM3
- OpenCV
- ROS
- PyTorch
- TensorFlow
💡 Tip
Use accurate timstamps to synchronize images and optimize monocular SLAM algorithms.
Frequently Asked Questions
Can this dataset be used to train autonomous driving models?
Yes, it provides simulated images in urban environments that are ideal for visual perception in autonomous driving.
What data formats are available in this dataset?
Timestamp PNG images, timestamp text file, YAML configuration file for SLAM.
Does the MIT license allow commercial use?
Yes, the MIT license is very permissive and allows commercial use without restrictions.
