摘要: |
视觉/惯性组合导航技术是自主导航领域的重要研究方向之一。首先介绍了视觉/惯性组合导航技术的发展概况,然后从纯视觉导航(里程计、同步定位与构图)以及组合导航(滤波、非线性优化)2个层次介绍了传统的基于视觉几何与运动学模型的视觉/惯性组合导航方法,还介绍了近年来发展迅猛的基于机器学习的视觉/惯性组合导航方法。最后,简要介绍了视觉/惯性组合导航技术的典型应用及未来发展趋势。 |
关键词: 视觉里程计 同步定位与构图 惯性导航 组合导航 |
DOI: |
|
基金项目:国家自然科学基金面上项目(61773394) |
|
A Survey of Intelligence Science and Technology Integrated Navigation Technology |
ZHANG Li-lian,QU Hao,MAO Jun,HU Xiao-ping |
(College of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, China) |
Abstract: |
Visual-inertial integrated navigation technology is an important research direction in autonomous navigation. This paper first introduces the development of visual-inertial integrated navigation technology, then introduces the traditional visual-inertial integrated method based on visual geometry and kinematics model from the viewpoint of pure visual navigation (odometry, simultaneous localization and mapping) and integrated navigation (filtering, nonlinear optimization), then introduces the visual-inertial integrated navigation method based on machine learning which has developed rapidly in recent years. Finally, the typical applications and the future trend of visual-inertial integrated navigation technology are briefly introduced. |
Key words: Visual odometry Simultaneous localization and mapping Inertial navigation Integrated navigation |