[pgpointcloud] RLE and SIGBITS heuristics

[Rémi Cura]

Here are some facts about pgpointcloud compression

5.2 Billion las points (usual suspects):
ratio 4.36 (compared to .las files)

2.15 Billion ply points (21 attributes) :
ratio 2.86 (compared to binary .ply files, already using a pivot point to
reduce digits)

Pgpointcloud is maybe not good at compressing float
70 Million stereo point cloud with XYZ+color, 2. ratio

The system can write from 1 Million to 0.2 Million pts/sec to a client,
depending on the point type

I'm in the process to measure compression/decompression time on those data


I tried to get the best ratio (thanks to recent work of @strk on uint64_t)

Cheers,
Rémi-C

2015-04-15 17:32 GMT+02:00 Sandro Santilli <strk at keybit.net>:

> On Wed, Apr 15, 2015 at 10:04:47AM -0500, Howard Butler wrote:
>
> > https://github.com/verma/laz-perf
>
> [...]
>
> > Maybe it would be possible to incorporate lazperf into pgpointcloud
> > as a compression type? See [1] for how we are using it.
>
> I guess it would be possible, yes.
>
> It'd need be callable from C though, either via an API exposed
> by laz-perf directly or by custom adapter.
> That might defeat the "remove virtuals overhead" goal.
>
> Do you have tools in PDAL to compare compression/decompression speed
> and compressed format size between the various supported schemes here ?
> https://github.com/PDAL/PDAL/blob/master/include/pdal/Compression.hpp#L113
>
> --strk;
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/pgpointcloud/attachments/20150415/7f8813d8/attachment.html>