Velodyne VLP-16 is spitting out THOUSANDS of CSVs!

[_puddnpie]

My buddy's got this Velodyne Lidar VLP-16 that he's been trying to get some good data from, and he's kinda scratching his head with the output.

He's got it hooked up and running with the Veloview software, and it records frames just fine, creating those .pcap files. Awesome, right?

But the thing is, when he tries to export stuff from those .pcap files, Veloview is spitting out every single frame as its own individual CSV file! We're talking like thousands of CSVs for one recording session. It's a total mess to deal with.

I want to help him, is there some trick to merge all those CSVs into one giant file? Or even better, is there a way to process the .pcap directly into a more standard point cloud format like LAS or something similar, instead of getting buried in CSVs?

He mentioned something about txt2las from LAStools, but it still seems like a roundabout way if he has to convert each CSV first.

He's also got a Garmin GPS 19x HVS hooked up to the Velodyne, with PPS at 1 Hz, so it's got timestamps for every frame. He even tried messing around with Wireshark to see if he could do anything there, but no luck.

Anyone out there dealt with Velodyne .pcap files before? What's your go to workflow for handling all this data?

Any secret software or scripts we should know about to avoid this CSV explosion?

Seriously, any tips would be a lifesaver! Thanks a bunch

 

[paxsmith]

Instead of going CSV route just skip straight to converting the .pcap files into a standard point cloud format using something like PDAL or ROS with a Velodyne driver, it would avoid all the CSV clutter completely

 

[Berth]

Yeah, Veloview's CSV dump is brutal, dont export from Veloview and use something like the velodyne_decoder or pcl PCDWriter in ROS or PCL to go straight from .pcap to point cloud formats like .pcd or .las, its better and easier to work with, if you're stuck with CSVs already, Python can stitch them fast