Coarse Alignment Methodology of Point Cloud Based on Camera Position/Orientation Estimation Model.

Area of Science:

• Geomatics Engineering
• Computer Vision
• Robotics

Background:

• Accurate 3D reconstruction relies on precise registration of point clouds.
• Coarse alignment is a critical prerequisite for effective point cloud registration.

Purpose of the Study:

• To develop and evaluate a novel methodology for the coarse alignment of LiDAR point clouds.
• To estimate the position and orientation of LiDAR stations for improved registration.

Main Methods:

• Utilized the pinhole camera model and a position/orientation estimation algorithm.
• Employed LiDAR camera images for ground control points and reference station point clouds.
• Compared proposed methodology with semi-automatic registration using total station measurements.

Main Results:

• Mean LiDAR position error of 0.072 m, comparable to semi-automatic registration (0.070 m).
• Proposed method achieved 0.124 m mean registration accuracy, while semi-automatic reached 0.072 m.
• Refined alignment using the proposed method resulted in an average point-to-point distance of 0.0117 m.

Conclusions:

• The proposed methodology provides a viable approach for coarse alignment of LiDAR point clouds.
• Semi-automatic registration offers superior accuracy due to integrated coarse and refined alignment.
• Further refinement steps are necessary to match the accuracy of advanced semi-automatic techniques.