You can write a simple node which subscribes to the 2 RGB feeds, uses blob matching (look up Stereo matching in OpenCV or Stereo Binary SGBM) to create the stereo image, and publishes it.
Alternately, I think OpenCV contains the functionality for image_to_pointcloud conversion.
Shouldn't this cause the first while loop to break immediately because the first point in the global_plan would be the one that is closest to the bot?
That looks like the intent to me, but I can't tell where global_plan is coming from or if it's guaranteed to be sorted by distance. Not quite sure what the point is here, but if you had something like a ...
From what I understand, the fixed frame is one where objects are assumed to be stationary with time. Maybe an example will make this clear:
Assume a robot with a lidar sensor moving in your world. It spits out a laserscan in the /lidar frame. The goal pose might be in an /odom frame. Now, if you encounter objects in the /lidar frame, your navigation stack ...
ZeroSim is built with Unity. It's not necessarily "easy" but arguably easier then Gazebo. Definitely creating and iterating environments in ZeroSim is much easier then Gazebo.
Disclosure: I am a ZeroSim developer.
Does anyone have any idea where this is coming from
Looks like line 640 of rospy.impl.tcpros_base:
logger.debug("[%s]: writing header", self.name)
and why it's getting logged constantly?
The particular message is logged with a call to logger.debug which is putting this message in your log file. It's getting written every time the write_header ...