Accurate pose estimation of known objects from LADAR scans helps automate on-site operations.
This book documents an experimental study that uses scans of I-beams to determine their pose, then compares the results to precise reference measurements. It focuses on turning point-cloud data into actionable position and orientation information.
- Learn two practical object-segmentation approaches: voxel binning and triangular irregular networks (TINs).
- See how principal components analysis creates bounding boxes that are matched against ideal I-beam specifications.
- Understand how an independent laser-based site measurement system validates pose results, using reference spheres for frame alignment.
- Explore how registration and transformation steps tie together the LADAR data with real-world measurements.
Ideal for readers interested in practical methods for 3D object recognition, pose estimation, and verification in construction contexts.