Hi, I have a set of point cloud data that contains over 25 million points that I captured using a FARO LiDAR scanner and imported into FARO Scene. I tried importing the e57 files from scene into Reality capture to create a mesh and noticed my point cloud data looked more sparse than it did on FARO scene. I checked the tie points registered in Reality Capture and the point count resulted in only 287,000 points being registered after importing the e57 files and having Reality Capture convert them to lsp? Do you know why the tie points do not align with the same amount of points registered in my original point cloud data scan? Are there settings that I need to change in Reality Capture that will get me to 25 million points?
Hello @UEuser97
The tie point isn’t the same as imported point cloud. These are the points created from the LSPs.
To get full model resolution you need to create the mesh.
Hello @otrhan. I believe I have a similar issue as @UEuser97.
I’ve been noticing all of my converted LSP data contains a fraction of points (10%) as my imported e57 files (exported from FARO SCENE). To me, it seems my e57 points are being down-sampled during import when converting to lsp.
When I compare generated meshes between Reality Capture (RC) and FARO SCENE, using the same point cloud dataset, the high resolution mesh from FARO SCENE looks a lot more detailed, while the RC mesh looks very low resolution and blobby looking . While investigating further, I noticed the blobby effect may be due to the very sparse points in the LSP files.
Hello @PeterV231
The converted LSP is like a helpful file to align the laser scan data with images. Otherwise the whole point cloud is imported and from that point cloud is the mesh created (but you don’t see that after import, you see only the sparse point cloud from LSPs, therefore you see the fraction of the points).
Regarding to the mesh, what are your reconstruction settings? Also, is it possible that you have the scan out of the intensity range 0.03 - 1 ?
Thanks @otrhan about the LSP explanation, it make more sense now.
As for my reconstruction settings, I’ve been using 0.003 for ‘Minimal distance distance between two points’ and a ‘Minimal intensity’ of 0.03. I never adjusted the ‘Minimal intensity’ before, so I keep it at its defaults value of 0.03.
Can you try to change the minimal intensity value to 0 and then compute the mesh again?
Also, it would be better to find out the intensity values of your scan. If the values are out of the mentioned range, those parts won’t be meshed properly.