A graph-based segmentation technique has been tailored to segment airborne LiDAR points which, unlike images, are irregularly distributed. In our method, every LiDAR point is labeled as a node and interconnected as a graph extended to its neighborhood, defined in a 4-D feature space: the spatial coordinates ($x,y,z$) and the reflection intensity. The interconnections between pairs of neighboring nodes are weighted based on the distance in the feature space. The segmentation consists of an iterative process of classification of nodes into homogeneous groups based on their similarity. This approach is intended to be part of a complete system for the classification of structures from LiDAR point clouds in applications needing fast response times. In this sense, a study of the performance/accuracy trade-off has been performed, extracting some conclusions about the benefits of the proposed solution. In addition, an interlaced graph-based approach is proposed to increase the reliability in general purpose segmentations.

Road extraction from Light Detection and Ranging (LiDAR) has become a hot topic over recent years. Nevertheless, it is still challenging to perform this task in a fully automatic way. Experiments are often carried out over small datasets with a focus on urban areas and it is unclear how these methods perform in less urbanized sites. Furthermore, some methods require the manual input of critical parameters, such as an intensity threshold. Aiming to address these issues, this paper proposes a method for the automatic extraction of road points suitable for different landscapes. Road points are identified using pipeline filtering based on a set of constraints defined on the intensity, curvature, local density, and area. We focus especially on the intensity constraint, as it is the key factor to distinguish between road and ground points. The optimal intensity threshold is established automatically by an improved version of the skewness balancing algorithm. Evaluation was conducted on ten study sites with different degrees of urbanization. Road points were successfully extracted in all of them with an overall completeness of 93%, a correctness of 83%, and a quality of 78%. These results are competitive with the state-of-the-art.

Over the last two decades, a wide range of applications have been developed from Light Detection and Ranging (LiDAR) point clouds. Most LiDAR-derived products require the distinction between ground and non-ground points. Because of this, ground filtering its being one of the most studied topics in the literature and robust methods are nowadays available. However, these methods have been designed to work with offline data and they are generally not well suited for real-time scenarios. Aiming to address this issue, this paper proposes an efficient method for ground filtering of airborne LiDAR data based on scan-line processing. In our proposal, an iterative 1-D spline interpolation is performed in each scan line sequentially. The final spline knots of a scan line are taken into account for the next scan line, so that valuable 2-D information is also considered without compromising computational efficiency. Points are labelled into ground and non-ground by analysing their residuals to the final spline. When tested against synthetic ground truth, the method yields a mean kappa value of 88.59% and a mean total error of 0.50%. Experiments with real data also show satisfactory results under visual inspection. Performance tests on a workstation show that the method can process up to 1 million points per second. The original implementation was ported into a low-cost development board to demonstrate its feasibility to run in embedded systems, where throughput was improved by using programmable logic hardware acceleration. Analysis shows that real-time filtering is possible in a high-end board prototype, as it can process the amount of points per second that current lightweight scanners acquire with low-energy consumption.