Iterative Hough Transform for Line Detection in 3D Point Clouds

Abstract

The Hough transform is a voting scheme for locating geometric objects in point clouds. This paper describes its application for detecting lines in three dimensional point clouds. For parameter quantization, a recently proposed method for Hough parameter space regularization is used. The voting process is done in an iterative way by selecting the line with the most votes and removing the corresponding points in each step. To overcome the inherent inaccuracies of the parameter space discretization, each line is estimated with an orthogonal least squares fit among the candidate points returned from the Hough transform. Source CodeThe reviewed C++ source code for this algorithm is available from the web page of this article 1 . Compilation and usage instruction are included in the README.txt file of the archive. Supplementary MaterialSix reference data sets are provided with the article. The data sets contain both synthetic data and experimental data of radioactive beams recorded in an active target time projection chamber (AT-TPC).

Automatic trajectory recognition in Active Target Time Projection Chambers data by means of hierarchical clustering

“…As in this case all trajectories are straight lines, data set III provides a nice test set for comparing our new algorithm with the Hough transform, which is a standard algorithm for finding lines in point clouds [10]. In order to achieve a fair comparison, we have first optimized the two parameters of the Hough transform, i.e.…”

“…This requires the shape of the tracks to be known a priori, and it leads to a voting scheme for parameter values, which can be either done exhaustively for all points (Hough transform [8]) or by random sampling (RANSAC [9]). In the absence of a magnetic field, the tracks are straight lines, which can easily be described parametrically, and both RANSAC and the Hough transform have been applied to this special case [10,11]. This is not generalizable to the case of unknown shapes like they occur in the presence of a magnetic field.…”

Low Overlapping Point Cloud Registration Using Line Features Detection

“…After the subsets ζ A ⊂ A and ζ B ⊂ B containing points corresponding to sharp features are identified, we proceed to find line features in them. We discretize the Hough parameter space by the method of [28] based on a tessellation of an icosahedron. This is followed by a modified version of Hough transform algorithm, which is applied iteratively and corrected by least squares error line fitting [28].…”

“…We discretize the Hough parameter space by the method of [28] based on a tessellation of an icosahedron. This is followed by a modified version of Hough transform algorithm, which is applied iteratively and corrected by least squares error line fitting [28]. These two improvements results in greater accuracy.…”

Parametric Reconstruction of Glass Fiber-reinforced Polymer Composites from X-ray Projection Data—A Simulation Study

“…From the filtered maxima of the NCC, an initial estimate of each fiber's direction and centroid is computed using the iterative Hough Transform algorithm for 3D line segments proposed by Dalitz et al [18]. This algorithm differs slightly from the traditional Hough transform.…”