Lidar-Based Semantic Mapping of Indoor Environments

基于激光雷达的室内环境语义映射

• 批准号: 558367-2020
• 负责人: Kelly, Jonathan
• 金额: $ 2.48万
• 依托单位: University of Toronto
• 依托单位国家: 加拿大
• 项目类别: Alliance Grants
• 财政年份: 2021
• 资助国家: 加拿大
• 起止时间: 2021-01-01 至 2022-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=723200
• 关键词: Lidar; Based; Semantic; Mapping; Indoor

Light detection and ranging (or lidar) sensors have revolutionized the nascent field of autonomous driving. The ability of lidar to generate accurate point cloud maps of the surrounding environment is essential to maintain situational awareness. Although lidar has primarily been deployed for self-driving vehicles to date, there are many other potential application areas. In particular, lidar can play an essential role in large-scale indoor mapping - the rich lidar data stream can be used to identify and track both known and novel objects indoors. Semantic maps, which contain information about individual object instances, are useful for inventory management in warehouses, for example.The Alliance research program will develop breakthrough capabilities in the area of lidar-based semantic mapping for warehousing and intralogistics applications. Intralogistics (which combines warehouse automation and logistics) is a $15 billion industry today and is expected reach $27 billion in size globally by 2025. In partnership with LeddarTech, Inc., the proposed Alliance research program will leverage machine learning to extract unique material and object signatures from lidar data streams in real time, improving the accuracy of semantic maps in cluttered indoor scenes. Enhanced semantic mapping capabilities will enable LeddarTech to enter new markets and to grow its offerings for existing markets, including the AV sector.

光检测和测距（或激光雷达）传感器彻底改变了新兴的自动驾驶领域。激光雷达生成周围环境的准确点云地图的能力对于保持态势感知至关重要。尽管迄今为止激光雷达主要应用于自动驾驶车辆，但还有许多其他潜在的应用领域。特别是，激光雷达可以在大规模室内测绘中发挥重要作用——丰富的激光雷达数据流可用于识别和跟踪室内已知和新物体。例如，语义地图包含有关单个对象实例的信息，对于仓库的库存管理非常有用。联盟研究计划将在基于激光雷达的语义地图领域开发突破性功能，用于仓储和内部物流应用。内部物流（结合了仓库自动化和物流）目前是一个价值 150 亿美元的行业，预计到 2025 年全球规模将达到 270 亿美元。与 LeddarTech, Inc. 合作，拟议的联盟研究计划将利用机器学习来提取独特的材料和物体来自激光雷达数据流的实时签名，提高了杂乱室内场景中语义地图的准确性。增强的语义映射功能将使 LeddarTech 能够进入新市场并为现有市场（包括 AV 领域）增加其产品。