稳健性(进化)
激光雷达
计算机视觉
计算机科学
遥感
测距
惯性测量装置
人工智能
帧(网络)
关键帧
电信
地理
生物化学
基因
化学
出处
期刊:Applied Optics
[The Optical Society]
日期:2021-06-17
卷期号:60 (18): 5430-5430
被引量:4
摘要
Lidar (also known as light detection and ranging) and inertial measurement unit (IMU) have become the core components in simultaneous localization and mapping (SLAM) system. Aiming at the problems of low accuracy and poor robustness, we propose what we believe is a novel scheme, including a local map matching rule, a plane representation method, and a key frame filtering process. Our solution removes redundant data during the plane fitting process and achieves higher accuracy. Experimental results prove that our method outperforms other popular methods in terms of trajectory errors and robustness under different scenarios.
科研通智能强力驱动
Strongly Powered by AbleSci AI