This ROS node can perform the following operations:

1. Rectify fisheye images coming from the realsense t265 ros driver
2. Compute and publish disparity image based on the step 1
3. Compute and publish a pointcloud2 based on the step 2

• Assuming you correclty installed librealsense, get your specific camera parameters using rs-enumerate-devices -c
• If using low-power cpu, use lower resolution (see full, half, third in cfg folder). Remeber to adjust your parameters accordingly.

• Launch the t265 ros driver, be sure to publish images
• roslaunch t265_depth t265_depth.launch

• input_topic_left: topic name of the left camera
• input_topic_right: topic name of the right camera
• output_frame_id: name of the frame with respect to where publish the cloud/disparity image
• process_every_nth_frame: to save computation, skip frames
• param_file_path: path of your camera parameters
• scale: full, half of a third of the resolution (1/4 not usefull, image details lost)