Has anyone used the ping360 to build accurate maps? I’m wondering how to best qualify a bin as a water vs a non water hit. I’ve seen a couple different strategies: static threshold, max value, averaged-value threshold.
I have access to a really great EKF (max .5 m drift over 30 minutes) and I’m hoping to build a map given this, however the general scan quality as of right now is not necessarily well set up so some maps are blurry.
The missions are being run on a river so there’s some influence from rain and wind on a surface vehicle and the AUV I’m trying to get this working on.
I’m wondering how people best identify a “hit” as opposed to open water reliably or if anyone has thoughts. Thanks!
The Ping360 is not typically used for making maps, it provides useful situational awareness and object detection, but as it is scanning horizontally from the ROV, it does not provide an estimate of bottom depth. Therefore the map you create would not be a bathymetric one, of how deep it is where?
It seems your system has an external position source, like a DVL, USBL, or both? If this is the case, your map could show objects in the water column. @EliotBR has given some tips on interpreting the strength of returns before, but generally it’s going to take experimentation (and potentially calibration) to determine what works best! Another thread details making maps with the sonar in ROS.
I’m interested in what your goals are for the created maps, as they seem atypical in format! Is knowing the depth profile of the river not desired?
Thanks for the answer and linked pages. Yes, I’m interested in mapping things in the river. I am currently using the sensor with a blueboat so I was wondering what would be required to map a dock or a wall or a bridge fixture.
Later, I might be interested in using it with a BlueROV or something else but the maps I’m interested in aren’t necessarily point clouds or gaussian splats but more just at the surface.
I’ve experimented with a Ping360 on a BlueBoat with the head pointed in the direction of travel, and didn’t find it particularly useful. While you do have the advantage of position via GPS, you’d need to calculate where the beam is measuring a distance two based on its rotation angle within the head, compounded with the roll and pitch of the vehicle. The scan rate is also low enough in comparison with a typical survey speed of 1 m/s, so you’d be measuring a distance to the bottom in a zig zag as the boat drove along?
Are you mounting the Ping360 in the same way, or just putting it upside down suspended from the BlueBoat? It may need to be fairly deep? It should be possible to map things like wall boundaries, but I’d expect this would be a fairly coarse measurement compared to more typical sonars intended for use on the BlueBoat, like the Omniscan SideScan. What are your goals for your maps?
I have the Ping360 suspended beneath the blueboat maybe a foot and a half and am doing the rotation projections with pitch and roll. I think the distancing measurement would be nice if it didn’t have such a wide vertical angle. I think the echosounder is more of what I want but it doesn’t rotate at all.
The goal is to have something that helps with 1) obstacle avoidance (so a rotating pinger is desirable) and 2) river structure studies.
For these, I’m considering including the SideScan but I want to have the ping as well. I might consider using them both, I was really just hoping to see if anyone had used it for lidar-like mapping with a BlueBoat.
Have you considered lowering your scanning rate or decreasing angular resolution? I know 1 m/s can be demanding with the sonar propagation time but I’m trying to think of ways around it.
To me, the Ping360 is just not the tool for the job - a sonar achieving a lidar like beam in water is very difficult to achieve! I could see the output being useful for forward looking collision avoidance, if set to sector sweep maybe 30-45 degrees (or less?) to increase the update rate. Making sure you’re on the latest Ping360 firmware is a good idea to (use PingViewer to update via USB connection, direct to computer.) Ethernet gives the fastest data rates possible!
The Omniscan 3D is best structure mapping and imaging I’ve seen in a sonar package <10-$20k, especially for the BlueBoat. You can checkout some datasets from it here, in the same interface that is used to interface with the sonars. The Surveyor multibeam echosounder produces similar output, but has a much narrower beam width, and produces many fewer points per ping. It does handle farther range better, but I doubt you’re getting more than 50-70m depth in most rivers?