Fit4CAD: A point cloud benchmark for fitting simple geometric primitives in CAD objects

We propose Fit4CAD, a benchmark for the evaluation and comparison of methods forfitting simple geometric primitives in point clouds representing CAD objects. Thisbenchmark is meant to help both method developers and those who want to identifythe best performing tools. The Fit4CAD dataset is composed by 225 high quality pointclouds, each of which has been obtained by sampling a CAD object. The way theseelements were created by using existing platforms and datasets makes the benchmarkeasily expandable. The dataset is already split into a training set and a test set. To assessperformance and accuracy of the different primitive fitting methods, various measuresare defined. To demonstrate the effective use of Fit4CAD, we have tested it on twomethods belonging to two different categories of approaches to the primitive fittingproblem: a clustering method based on a primitive growing framework and a parametricmethod based on the Hough transform.