Automatic coarse co-registration of point clouds from diverse scan geometries: a test of detectors and descriptors

Francesco Pirotti,Alberto Guarnieri,Sebastiano Chiodini,Carlo Bettanini

DOI: https://doi.org/10.5194/isprs-annals-X-1-W1-2023-581-2023

2023-08-27

Abstract:Point clouds are collected nowadays from a plethora of sensors, some having higher accuracies and higher costs, some having lower accuracies but also lower costs. Not only there is a large choice for different sensors, but also these can be transported by different platforms, which can provide different scan geometries. In this work we test the extraction of four different keypoint detectors and three feature descriptors. We benchmark performance in terms of calculation time and we assess their performance in terms of accuracy in their ability in coarse automatic co-registration of two clouds that are collected with different sensors, platforms and scan geometries. One, which we define as having the higher accuracy, and thus will be used as reference, was surveyed via a UAV flight with a Riegl MiniVUX-3, the other on a bicycle with a Livox Horizon over a walking path with un-even ground.The novelty in this work consists in comparing several strategies for fast alignment of point clouds from very different surveying geometries, as the drone has a bird's eye view and the bicycle a ground-based view. An added challenge is related to the lower cost of the bicycle sensor ensemble that, together with the rough terrain, reasonably results in lower accuracy of the survey. The main idea is to use range images to capture a simplified version of the geometry of the surveyed area and then find the best features to match keypoints. Results show that NARF features detected more keypoints and resulted in a faster co-registration procedure in this scenariowhereas the accuracy of the co-registration is similar to all the combinations of keypoint detectors and features.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to achieve automatic coarse registration of point clouds under different scanning geometric conditions (for example, data collected from aerial drones and ground bicycles). Specifically, the author tested the performance of four different keypoint detectors and three feature descriptors in terms of computation time and accuracy to evaluate their effects when coarsely registering point clouds from two different sources. One of the point clouds comes from high - precision drone flight scans, and the other comes from low - cost bicycle - mounted scanning devices. Due to uneven terrain and lower device costs, the data quality of the latter is relatively low. The main challenge of the research lies in how to quickly and effectively align these two point clouds with significantly different perspectives and precisions.

The paper solves this problem by comparing several strategies, especially by using range images to capture the simplified geometry of the measured area and then finding the best keypoint matching method. The experimental results show that the NARF feature detects more keypoints and can achieve a faster coarse registration process in this scenario, although all combinations of keypoint detectors and feature descriptors perform similarly in terms of the accuracy of coarse registration.