Modified ICP based Framework for 3D Localization and Mapping in GNSS-denied Environments using LiDAR

Abstract

Lidar odometry and mapping has emerged as fundamental technology in various fields, including accurate positioning and mapping, autonomous vehicles, robotics, and environmental monitoring. The ability to create accurate maps and calculate trajectories is essential for localization and navigation in complex environments, particularly in scenarios where GNSS signal are unavailable. In this context, we propose a novel method for simultaneous localization and mapping (SLAM) specifically designed for GNSS-denied environments and without the need for high-accuracy ranging or inertial measurement units (IMUs). Our approach aims to address the challenges posed by GNSS-denied environments and the cost constraints associated with using multiple sensors. By using only lidar sensors, specifically the Velodyne VLP 16 sensor, we offer a cost-effective solution. Our method utilizes a modified version of the Iterative Closest Point (ICP) algorithm to handle consecutive lidar scans seamlessly. In addition to our method, we conducted thorough accuracy assessments on the created map. The reason behind extensive accuracy evaluation is to ensure the quality of registration. By comparing alignment between the features of different frames within the map and determining the error between the registered features and the original, we gain a more comprehensive understanding of registration quality.