<div dir="ltr"><div><div><div>Hi Howard,<br></div><div><br></div>Sorry for the delay. You understood the question correctly, I was interested in how PDAL's chipper works.<br><br></div>In this case, I just wanted to make sure that it was x/y bounding boxes, as that is what I need, as I need to difference with ground points, so I want them "nearby" as in, in the same chip as my vegetation points. I like the idea of the abstraction of n-dimensional chipping though.<br>
<br></div>My main use case at the moment is just moar lidar, but definitely interested in applications for point clouds for OpenDroneMap, i.e. point clouds from Structure from Motion approaches.<br><br>I think too, it could be interesting to abuse point cloud / PDAL for continuous return LiDAR, but I don't personally have an application for that yet... .<br>
<div><div><br></div><div>Best,<br>Steve<br><br><br></div><div><br></div></div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Mon, Jun 30, 2014 at 12:11 AM, Howard Butler <span dir="ltr"><<a href="mailto:howard@hobu.co" target="_blank">howard@hobu.co</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class=""><br>
On Jun 27, 2014, at 8:14 PM, Stephen Mather <<a href="mailto:stephen@smathermather.com">stephen@smathermather.com</a>> wrote:<br>
<br>
> Hi All,<br>
><br>
> So, patches in pointcloud-- they are multidimensional like the data organized in them, or are they organized in x/y rectangles independent of z, intensity, color, and other possible dimensions?<br>
<br>
</div>I think I understand what you're asking, and I'll take a crack at it.<br>
<br>
Patches are organized by their x/y bounding box. Patches *may* overlap extents, though the most efficient querying pattern is to construct your patches so they don't (currently how PDAL's chipper works).<br>
<br>
There's nothing automagic, but you could construct overlapping patches of different attribute sets. The common model is to simply chunk up the multidimensional array of data into something that adapts well to existing spatial indexes.<br>
<br>
Did you have something different in mind? Point cloud-like data that aren't point clouds? I'm interested in what prompts the question...<br>
<span class="HOEnZb"><font color="#888888"><br>
Howard</font></span></blockquote></div><br></div>