Hello everyone!
I was wondering if it is possible to achieve SLAM using multiple monocular lenses. I am trying to implement 360º SLAM on a drone. However, adding multiple HAWK cameras would take up too much space. Thanks!
Hi @jim.fabian
No, you cannot use a single monocular camera, but you can use only a set of stereo cameras.
To better assist you, here is the reference documentation:
cuVSLAM does not just work with a single stereo camera. It is possible to use up to 32 cameras (=16 stereo cameras) as input. Using multiple cameras can significantly increase robustness, especially in featureless environments.
https://nvidia-isaac-ros.github.io/concepts/visual_slam/cuvslam/index.html
Hello @Raffaello,
Thank you for your response! I now have a deeper understanding of how this package works.
Shifting the topic to the .launch files, is it possible to implement my own dual monocular cameras that works as a stereo? Specifically, can I use two IMX577 Camera Modules with their own calibration and positioning?
The reason for my question is that I want to optimize the sensors on a drone while using Isaac ROS packages for a 360º SLAM and vision for obstacles detection (one layer of Nvblox would not be enough for drone navigation).
Thanks!
Hi @jim.fabian
Yes, the Isaac ROS VSLAM requires a synchronized stereo camera that meets the following camera requirements:
| Specification | Required Specification |
|---|---|
| Minimum target imager framerate | 30 Hertz |
| Maximum permissible jitter in imager framerate | +/- 2 milliseconds |
| Maximum expected offset between imagers within stereo camera | +/- 100 microseconds |
| Maximum expected offset between imagers across stereo camera | +/- 100 microseconds |
But if you want to achieve the same performance as listed in the documentation, you must use a camera from this specific list: Sensors Setup — isaac_ros_docs documentation
Best,
Raffaello
1 Like
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.