Goal

The capstone project "Application of LiDAR Technology for Advanced Object Modeling in 3D Printing," was completed as part of the Tech Capstone 2024. The project's goal was to bridge the gap between raw LiDAR data and 3D structures that could be printed by using LiDAR technology to generate sophisticated object models. The primary objective was to use point cloud data recorded from LIDAR to generate extremely accurate and detailed 3D models by combining LiDAR technology with 3D printing. To create accurate 3D printed models of a variety of things, including machinery parts and urban infrastructure, our goal was to acquire high-resolution point cloud data and process it efficiently.

Methodology

Project involved leveraging LiDAR technology to create precise 3D models of urban infrastructure and machinery components for 3D printing applications. Initially, we utilized Ouster LiDAR but transitioned to Livox LiDAR and iPhone LiDAR for improved data quality and application-specific needs. Data collection involved capturing point clouds using these sensors, optimizing settings for accuracy. We implemented the FAST-LIO algorithm for real-time odometry and mapping on an Ubuntu 20.04 system equipped with necessary libraries and drivers. Point cloud data was processed using Open3D, involving steps like bounding box extraction, ground plane removal, outlier removal, voxel downsampling, and surface reconstruction. The processed data was then refined in MeshLab and converted to STL format for 3D printing using the Ultimaker printer, successfully demonstrating the utility of our software solution in creating detailed 3D models.

Future development

We can get better results by choosing the right range LIDAR and recording more adequate data. In the future, The algorithm can be further improved for 3D reconstruction from point clouds. Moreover, a comprehensive GUI software can be developed to streamline the process more smoothly.