MATLAB

Guidelines for the MATLAB Implementation

This is the guideline and structual explanation of the MATLAB implementation of the EMS algorithm.

Installation

The code is written and tested on MATALB R2021a. Should be compatible with releases newer than R2017. The EMS algorithm is self-contained. To run hierarchical-EMS and demo scripts, the Computer Vision Toolbox is required. For compaitibility problems, you are more than welcomed to put up an issue.

File Structure

• The source code is located in /src, where EMS.m is the core algorithm for single superquadric recovery, and Hierarchical_EMS.m is the algorithm for multisuperquadric recovery.
• Utility functions for demonstrations (e.g. superquadric rendering, sampling, baseline algorithms) are located in /src/utilities.
• Please add /src to path before excuting the test scripts.
• Test scripts are located in /example_scripts.
• Demo point cloud files (.ply) are provided in /example_scripts/data.

Run Demo

Demo for single superquadric recovery

There are two test scripts you can play with: /example_scripts/example_single_superquadric_script.m and /example_scripts/random_example_script.

• With example_single_superquadric_script.m, you will load a .ply in /example_scripts/data/single_superquadric and recover the superquadric representation. The script will compare and visualize the results obtained by the baseline methods (Radial-LSQ, Numerical Stable) with ours.
• With /example_scripts/random_example_script, a point cloud of a random superquadric will be generated, and then recovered. In the script, you can specify the partial, outlier and noise levels.

Demo for multiple superquadric recovery

For multiple superquadric recovery, please run the demo script /example_scripts/multi_superquadric_script.m. You can load point clouds from /example_scripts/data/multi_superquadrics.