Software

• Written by Dong Tian @ MERL
• Modified by Maurice Quach @ L2S, CNRS, CentraleSupélec, Université Paris-Saclay

This software is written to compute geometric distortions in a point cloud, including point-to-point and point-to-plane.

If you use this code for research, please cite:

@inproceedings{Tian_2017,
	doi = {10.1109/icip.2017.8296925},
	url = {https://doi.org/10.1109%2Ficip.2017.8296925},
	year = 2017,
	month = {sep},
	publisher = {{IEEE}},
	author = {Dong Tian and Hideaki Ochimizu and Chen Feng and Robert Cohen and Anthony Vetro},
	title = {Geometric distortion metrics for point cloud compression},
	booktitle = {2017 {IEEE} International Conference on Image Processing ({ICIP})}
}

Modifications

• Changed Readme to Markdown format
• Changes to output format to facilitate parsing

Compiling notes

PCL required

Tested with PCL 1.8. May work with some older versions, but not tested yet.

Compile under Linux, follow the steps:

mkdir build
cd build

Simple compile

The executable pc_error would be produced under ./build folder.

cmake ..
make

Eclipse IDE

The executable pc_error_d would be produced under ./run folder.

cmake -G "Eclipse CDT4 - Unix Makefiles" -DCMAKE_BUILD_TYPE=Debug -DCMAKE_ECLIPSE_GENERATE_SOURCE_PROJECT=TRUE -DCMAKE_ECLIPSE_MAKE_ARGUMENTS=-j12 ..

Compile under Window, not tested yet

Calling examples

• ./pc_error -a cloudA.ply

  Just load the single point cloud, check its intrinsic resolutions.

• ./pc_error -a cloudA.ply -A cloudNormalsA.ply

  Estimate normals for cloudA.ply and save the point cloud with normals to cloudNormalsA.ply.

• ./pc_error -a cloudOrg.ply -b cloudDec.ply

  Compute the point-to-point and point-to-plane metrics. Only report RMS and corresponding PSNR's.

• ./pc_error -a cloudOrg.ply -b cloudDec.ply -d

  Compute the point-to-point and point-to-plane metrics. Report RMS, Hausdorff and corresponding PSNR's.

• ./pc_error -a cloudOrg.ply -b cloudDec.ply -f

  Compute the point-to-point and point-to-plane metrics. Force normal estimation even normals are provided in the input cloud. This feature is for experimental purposes.

More parameters for experimental purposes

-k [ --knn ] arg (=12)    Set KNN number of neighbor points for normal
                          estimation. Default normal estimation method
-t [ --rtimes ] arg (=0)  Guide to set radius for normal estimation,
                          times to NN points distances
-s [ --singlePass ]       Force running a single pass, where the loop
                          is over the original point cloud. Non-symmetric
                          reports
-1 [ --point2point ]      Force to report point-to-point metric only

Screen snapshot example

$ ./build/pc_error -a test/sphere100.pcd -b test/noisy_sphere100.pcd
infile1: test/sphere100.pcd
infile2: test/noisy_sphere100.pcd
knn = 12
force normal estimation: 0

Failed to find match for field 'rgb'.
Failed to find match for field 'normal_x'.
Failed to find match for field 'normal_y'.
Failed to find match for field 'normal_z'.
Failed to find match for field 'curvature'.
Reading file 1 done.
Failed to find match for field 'rgb'.
Failed to find match for field 'normal_x'.
Failed to find match for field 'normal_y'.
Failed to find match for field 'normal_z'.
Failed to find match for field 'curvature'.
Reading file 2 done.
Minimum and maximum NN distances (intrinsic resolutions): 0.00999999, 0.0217456
Point cloud sizes for org version, dec version, and the scaling ratio: 10000, 10000, 1

0. Preparing normals.
	KNN in use: 12
	Normal estimation begin..
	Normal estimation on original point cloud DONE! It takes 0 seconds (in CPU time).
	Converting normal vector DONE. It takes 0 seconds (in CPU time).

1. Use infile1 (A) as reference, loop over A, use normals on B. (A->B).
	Error computing takes 0 seconds (in CPU time).
	### A->B,rms1,p2point,0.000847707
	### A->B,rms1PSNR,p2point,28.1825
	### A->B,rms1,p2plane,0.000471006
	### A->B,rms1PSNR,p2plane,33.2869
2. Use infile2 (B) as reference, loop over B, use normals on A. (B->A).
	Error computing takes 0 seconds (in CPU time).
	### B->A,rms2,p2point,0.000847707
	### B->A,rms2PSNR,p2point,28.1825
	### B->A,rms2,p2plane, 0.000471006
	### B->A,rms2PSNR,p2plane,33.2869
3. Final (symmetric).
	### Symmetric,rmsF,p2point,0.000847707
	### Symmetric,rmsFPSNR,p2point,28.1825
	### Symmetric,rmsF,p2plane,0.000471006
	### Symmetric,rmsFPSNR,p2plane,33.2869
Job done! 0 seconds elapsed (excluding the time to load the point clouds).