Abstract
Acquiring 3D data with LiDAR systems involves scanning multiple scenes from different points of view. In actual systems, the ICP algorithm (Iterative Closest Point) is commonly used to register the acquired point clouds together to form a unique one. However, this method faces local minima issues and often needs a coarse initial alignment to converge to the optimum. This paper develops a new method for registration adapted to indoor environments and based on structure priors of such scenes. Our method works without odometric data or physical targets. The rotation and translation of the rigid transformation are computed separately, using, respectively, the Gaussian image of the point clouds and a correlation of histograms. To evaluate our algorithm on challenging registration cases, two datasets were acquired and are available for comparison with other methods online. The evaluation of our algorithm on four datasets against six existing methods shows that the proposed method is more robust against sampling and scene complexity. Moreover, the time performances enable a real-time implementation.
Highlights
The most popular algorithm in practical automatic pipelines is ICP (Iterative Closest Point) [5]
This paper develops a new method for registration adapted to indoor environments and based on structure priors of such scenes
With the evaluation on DS3-V, we show that the Structured Scene Feature-based Registration (SSFR) algorithm can solve the registration with noisy data without further preprocesses than a uniform filter
Summary
The most popular algorithm in practical automatic pipelines is ICP (Iterative Closest Point) [5]. The algorithm performs a computationally-expensive nearest neighbor search, which limits the number of points that can be processed within a reasonable time. As the cost function is not convex, the minimization can lead to a local minimum To address this issue, one can provide the algorithm with scans that are initially closely aligned, either adapting the acquisition phase with less motion between scans or finding a coarse alignment beforehand. One can provide the algorithm with scans that are initially closely aligned, either adapting the acquisition phase with less motion between scans or finding a coarse alignment beforehand This coarse alignment can be manual, given by odometry measurements or provided by potentially expensive algorithms that do not necessarily converge
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
