pointcloudToMesh

C++ application to convert pcd file, ply file, txt file or xyz point cloud to MESH representation (Gp3)

---

Input file structure support

• .pcd
• .ply
• .txt
• .xyz

Output file structure

Mesh cloud:

• .ply

Example

Docker image

There is a docker image for this project stored in docker hub, here. This image is compiled with pcl-docker-1.9.1, Ubuntu 20.04 and the mesh project (5.85GBaprox.).

To use it you have to install docker-engine in your host machine:

Download the docker image

docker pull danieltobon43/pointcloud-to-mesh:1.0-ubuntu-20-04

Check downloaded image

docker images

Run a docker container

docker run --rm -it \
           --volume=/tmp/.X11-unix:/tmp/.X11-unix:rw \
           --volume=/tmp/.docker.xauth:/tmp/.docker.xauth:rw \
           --env="XAUTHORITY=/tmp/.docker.xauth" \
           --env="DISPLAY" \
           --name="mesh" \
           --cap-add sys_ptrace \
           -p 127.0.0.1:2222:22 \
           --user=pcl \
           --volume=[PATH TO YOUR PCD FOLDER]:/home/pcl/project/pcd \
           -t danieltobon43/pointcloud-to-mesh:1.0-ubuntu-20-04 pcd/[YOUR PCD FILENAME] [surface method] [normal method] [output dir]

The previous command will run a docker container with the danieltobon43/pointcloud-to-mesh:1.0-ubuntu-20-04 image and will share a .pcd file from the host machine ([PATH TO YOUR PCD FOLDER]) to the pcd folder in the container.

More information about this docker image can be found in the docker hub repository.

Compilation

• Set "YOUR OWN" PCL Build DIR in CMakeList.txt e.g: /opt/pcl-1.8.1/build and save it
• Create a "build" folder

in the main folder:

- cd build  
- cmake ../src/
- make

Test

./pointcloudToMESH <input cloud> <surface method estimation> <normal method estimation> <output dir>

surface method estimation:
    1 --> for poisson
    2 --> for gp3
    
normal method estimation.
    1 --> for normal estimation
    2 --> for mls normal estimation
    
 example:
 ./pointcloudToMESH /home/xXx/PCD-PLY_Files/cloud.txt 2 1 /home/xXx/Desktop    

You can modify the parameters in "create_mesh" method to get better results!

for GP3:

setKSearch(100);                         //It was 20
gp3.setSearchRadius(10);                 //It was 0.025
gp3.setMu(5);                            //It was 2.5
gp3.setMaximumNearestNeighbors(100);     //It was 100
gp3.setMaximumSurfaceAngle(M_PI/4);      //45 degrees   
gp3.setMinimumAngle(M_PI/18);            //10 degrees 
gp3.setMaximumAngle(M_PI/1.5);           //120 degrees     
gp3.setNormalConsistency(false); 

for POISSON:

poisson.setDepth(depth);//7
poisson.setInputCloud(cloud_with_normals);
poisson.setPointWeight(pointWeight);//2
poisson.setDegree(2);
poisson.setSamplesPerNode(samplePNode);//1.5
poisson.setScale(scale);//1.1
poisson.setIsoDivide(isoDivide);//8
poisson.setConfidence(confidence);
poisson.setOutputPolygons(outputPolygons);
poisson.setManifold(manifold);
poisson.setSolverDivide(solverDivide);//8
poisson.reconstruct(triangles);