PCL: How does textureMeshwithMultipleCameras() handle occlusion?

559 views Asked by At

From the kinfu samples, i notice:

https://github.com/PointCloudLibrary/pcl/blob/master/gpu/kinfu_large_scale/tools/standalone_texture_mapping.cpp

That the textureMeshing function with multiple cameras seems to exist. However, there doesn't seem to be any input to the depth image. Hence, how is this function able to handle occlusion?

1

There are 1 answers

0
Deepfreeze On

See the function:

tm.textureMeshwithMultipleCameras(mesh, my_cams);

It has as inputs the mesh and the camera positions and directions. For explanation of the functions see:

https://github.com/PointCloudLibrary/pcl/blob/46cb8fe5589e88e36d79f9b8b8e5f4ff4fceb5de/surface/include/pcl/surface/texture_mapping.h

and

https://github.com/PointCloudLibrary/pcl/blob/46cb8fe5589e88e36d79f9b8b8e5f4ff4fceb5de/surface/include/pcl/surface/impl/texture_mapping.hpp

The occlusion functions use Octree.