SHREC'21 track: Fitting simple geometric primitives on point clouds from CAD objects

The aim of this SHREC'21 track is to evaluate the quality of automatic algorithms for fitting geometric primitives on CAD models represented as point clouds. The goal is to identify the number and the type of simple geometric surface primitives that are present on the surface of the CAD models. As simple primitives we mean the classical surface primitives derived from constructive solid geometry, i.e., planes, spheres, ellipsoids, tori, cones and cylinders. The track is organized by IMATI-CNR. The final report of this SHREC'21 track will be submitted as a joint contribution to the international journal Computers & Graphics and will follow a two-stage review process. The paper will be authored by the track coordinators and all participants who submitted their results.

Dataset, ground truth and evaluation
A set of sample models that could be used for training the algorithms will be made available in mid-January, while the test set will be made available at the end January.

For each point cloud in the test set, the goal of the track is to recognize and locate the surface primitives on its surface. The participants are asked to list, for each model, the primitives found: their type, the points that are recognized to belong to it, and optionally the coefficients with respect to given representations.

Registration and other procedures
Each participant is requested to register to the track by sending an email to Chiara Romanengo (email: chiara.romanengo@ge.imati.cnr.it) and Andrea Raffo (email: andrea.raffo@ge.imati.cnr.it) with the subject "SHREC'21 track: Fitting simple geometric primitives on point clouds from CAD objects". Then, an answer will be sent to each participant with the instructions to:

• Download the set of sample models, which can be used as a training set by learning-based methods, from January 15th, 2021;
• Download the test set from January 30th, 2021;
• Provide by February 14th, 2021 the results of their runs on the test set, the program that generated the results (binary code, including appropriate user's instructions if needed) and a summary of the proposed method. For each task, the results of up to 3 runs may be submitted; each run may come from a different algorithm, or from a different parameter setting. Note that methods which do not rely on the learning-based paradigm can skip the training set and directly work on the test set.

We plan to submit the benchmark code and data to the replicability stamp

Further Information

• The evaluations will be done automatically;
• The track results will be combined into a joint paper to be submitted to Computers & Graphics;
• The description of the track and its results will be presented at the Eurographics Workshop on 3D Object Retrieval (September, 2021).