Skip to content

gseg-ethz/supervoxel

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Supervoxel segmentation tool

This python library enables the segmentation of a point cloud into boundary preserving super voxels. Boundary preservation is achieved by incorporating cosine distance between the normal vectors in the objective function. This source code is based on the ISPRS paper by Lin et al. (2018) and the provided source code. Currently, we only provide a python wrapper for a single function, but other functions can be added by adapting the supervoxel.i file.


class supervoxel.computeSupervoxel(input_file, k_neighbors, resolution, save_file)

Parameters: 

- input_file (string): path to the file that should be segmented into supervoxels

- k_neighbors (int): number of nearest neighbors used for the normal vector estimation

- resolution (float): desired diameter of the supervoxels indirectly
  controls the number of supervoxels (see paper for more information)

- save_file (string): path where the results should be saved if "None" data will not be saved (the folder must exist already) 

Returns:

- Function saves "ASCII" files to the defined path (save_file). First three
  columns are the coordinates of the points, columns 3-6 are randomly assigned
  colors (based on label), label denoting the supervoxel (all points with the 
  same label belong to one supervoxel) 

[1] Lin, Y., Wang, C., Zhai, D., Li, W., & Li, J. (2018). Toward better boundary preserved supervoxel segmentation for 3D point clouds. ISPRS Journal of Photogrammetry and Remote Sensing, 143, 39–47. https://doi.org/10.1016/j.isprsjprs.2018.05.004

About

No description, website, or topics provided.

Resources

License

Stars

1 star

Watchers

2 watching

Forks

Packages

 
 
 

Contributors

Languages