Search

The Americans with Disabilities Act (ADA) requires that the design, construction, and maintenance of curb ramps provide accessibility to persons with disabilities. Light detection and ranging (lidar) can efficiently collect high-accuracy and high-resolution three-dimensional (3D) geometric data, which can be leveraged for ADA compliance assessment in a virtual environment to help guide and improve field practice. Most existing studies do not consider many aspects of the field inspection process including local variations in slope measurements and the physical instrument characteristics, warranting a more detailed accuracy evaluation. Accordingly, a workflow for virtual ADA compliance inspection of curb ramps using lidar data is proposed and was applied to a controlled experiment using full-size ramp specimens to optimize parameters and compare methods with rigorous accuracy assessments. The proposed workflow consists of data preprocessing, digital elevation model generation, and digital inclinometer simulation. In the digital inclinometer simulation, we developed and implemented four different approaches including surface normal, immediate neighbor linear regression, scaled neighbor linear regression, and touching point extraction. Among these approaches, the novel touching point extraction method considers both the roughness of the surface and the scale in which a digital inclinometer measures the slope. The accuracy and effectiveness of the framework were demonstrated, resulting in a root-mean square error of 0.18% for the mean slope, which is on par with the accuracy of a digital inclinometer (0.2%, 1−σ). An example virtual experiment is provided using the proposed approach to identify the optimal spacing of measurement samples to determine the slope of a curb ramp for ADA compliance assessment. [ABSTRACT FROM AUTHOR]

Dense three-dimensional (3D) point clouds collected from rapidly evolving data acquisition techniques such as light detection and ranging (lidar) and structure from motion (SfM) multiview stereo (MVS) photogrammetry contain detailed geometric information of a scene suitable for a wide variety of applications. Among the many processes within a typical point cloud processing workflow, segmentation is often a crucial step to group points with similar attributes to support more advanced modeling and analysis. Segmenting large point cloud data sets (i.e., hundreds of millions to billions of points) can be extremely time consuming and tedious with current tools, which primarily rely on significant manual effort. While many automated methods have been proposed, the practicality, scalability, and versatility of these approaches remain a bottleneck stifling processing of large data sets. To overcome these challenges, this paper introduces a novel, generalized segmentation framework called Vo-Norvana, which incorporates a new voxelization technique, a normal variation analysis considering the positioning uncertainty of the point cloud, and a custom region growing process for clustering. The proposed framework was tested with several large-volume data sets collected in diverse scene types using several data acquisition platforms including terrestrial lidar, mobile lidar, airborne lidar, and drone-based SfM-MVS photogrammetry. In evaluating the accuracy of models generated from Vo-Norvana against manual segmentation, the average error of the position, orientation, and dimensions are 2.7 mm, 0.083°, and 0.9 mm, respectively. Over 0.2 million points per second and 36 thousand voxels per second can be achieved when segmenting an airborne lidar data set containing over 639 million points to about 1 million segments. [ABSTRACT FROM AUTHOR]