Incorporate isaac_ros_compression into perception_auv for bagging data

[kluge7]

Currently we just bag rosbags at normal size, this is not ideal since video streams can end up being multiple GBs large. An idea could be to incorporate isaac_ros_compression for reducing the image size when bagging data. If we are working on the AGX-ORIN then check out the release-3.2 branch. The idea is that we incorporate this into our relevant mission launch files, and then we rosbag the CompressedImage.msg topic (https://github.com/ros2/common_interfaces/blob/humble/sensor_msgs/msg/CompressedImage.msg) which we then can later on decode and upload to teams or NAS. H264 will lose information, so when we encode and decode we wont get an exact copy of the original data, so test if this is acceptable enough. It is also possible to create a H265 encoder and decoder node since that will be better than H264, can also research and explore FFV1 since that is lossless video codec.

1. Fork the isaac_ros_compression repository into vortexntnu organization
2. Explore how it works, you need to test it on our Jetson Orin AGX platform or another platform which has isaac_ros installed.
3. Decide on if you want to incorporate H264 (easy), or implement H265 or FFV1 (hard) ros nodes which utilize the correct hardware (need to implement both decoding and encoding)
4. Then incorporate them into our mission launch files in perception_auv, so that they compress an image topic of interest (most likely it will be the raw image data that is most interesting).
5. Add docs for how it works, and how it can be decoded.

Contacts

@kluge7
@jorgenfj