Robust Simultaneous Localization and Mapping via Loop Closure Optimization

通过环路闭合优化实现稳健的同步定位和映射

• 批准号: 538631-2019
• 负责人: Beltrame, Giovanni
• 金额: $ 1.82万
• 依托单位: École Polytechnique de Montréal
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2019
• 资助国家: 加拿大
• 起止时间: 2019-01-01 至 2020-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=681115
• 关键词: Robust; Simultaneous; Localization; Mapping; via

SLAM (Simultaneous Localization And Mapping) has been an extremely active and fast-evolving field for the past decades. It attempts to solve the fundamental problem in truly autonomous systems: how to extract information from data and fuse the information to create a map of the environment, which facilitates the subsequent decision making process. Despite the extensive works and vast advances, long-term operation remained an open problem due to the brittleness of current SLAM systems. Perceptual aliasing, where different places appear similar to the sensor, prevents the loop closure detection as well as map estimation from achieving reliable performance over long periods of time. The main goal of this project is to develop a pose graph optimization algorithm that is resilient to spurious loop closure detections. ARA Robotics, a Montreal-based company that specializes in aerial robotics and software design, is currently expanding its offering to cover SLAM solutions, and enhance the navigation and mapping capability of their robots, using the RTAB-MAP framework. ARA Robotics is interested in research on a robust pose graph optimization algorithm that enhances the performance of long-term SLAM operation. The project will contribute to a software product for SLAM that will contribute to the development of Canada's Unmanned Air Vehicle (UAV) industry and robotics industry in general. This research project will apply state-of-the-art work on SLAM to highly sought-after civilian applications for UAVs. This will have direct relevance to the Canadian economy.

SLAM(同步定位与测绘)在过去的几十年里一直是一个非常活跃和快速发展的领域。它试图解决真正自主系统中的根本问题：如何从数据中提取信息，并将信息融合以创建环境地图，从而为后续的决策过程提供便利。尽管进行了大量的工作和巨大的进步，但由于当前SLAM系统的脆性，长期运行仍然是一个悬而未决的问题。感知混叠使环路闭合检测和MAP估计在很长一段时间内无法获得可靠的性能，其中不同位置看起来与传感器相似。该项目的主要目标是开发一种对伪环闭合检测具有弹性的姿势图优化算法。Ara Robotics是一家总部位于蒙特利尔的公司，专门从事空中机器人和软件设计，目前正在扩大其产品范围，以涵盖SLAM解决方案，并使用RAB-MAP框架增强其机器人的导航和地图绘制能力。ARA Robotics感兴趣的是研究一种稳健的姿势图优化算法，以提高长期SLAM操作的性能。该项目将有助于SLAM的软件产品，这将有助于加拿大无人机(UAV)行业和整个机器人行业的发展。这一研究项目将把SLAM的最先进工作应用于备受欢迎的无人机民用应用。这将与加拿大经济直接相关。