|
To better understand and accurately determine the phenotypic parameters of scrub sands, a three-dimensional reconstruction methodology has been proposed that relies on ground-based LiDAR and voxel modeling. The first stage involves using FaroFocus to gather point cloud data of the scrub from various angles, optimizing the collection process through information fusion. Next, the point cloud data is refined through background removal and radius filtering, setting a radius threshold to isolate a single scrub sand mound point cloud. To enhance computational efficiency, the point cloud volume is managed through down-sampling and fine denoising procedures. Lastly, the voxel method is employed for 3D reconstruction. Key features such as plant height, crown width, surface area, and volume of the scrub sand mound are computed using the alpha-shape algorithm and the convex hull method. These methodologies are utilized to align the single scrub sand mound, thereby achieving a comprehensive reconstruction of the scrub sand mound model. The research findings indicate that the introduction of down-sampling and fine denoising processes significantly improves the efficiency of point cloud reconstruction. The resulting model is highly accurate and realistic, closely resembling the actual scrub sand mound. This reconstructed model offers a visual research foundation and technical support for the study of scrub sandpile characteristics.
|
