arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2009.02672cs.CVcs.AIcs.RO
收藏

深度视觉里程计的方法、挑战和应用:走向复杂和新兴领域

Approaches, Challenges, and Applications for Deep Visual Odometry: Toward to Complicated and Emerging Areas

Chongqing University(重庆大学);Chongqing Normal University(重庆师范大学);Changan Auto Company(长安汽车公司)

Ke Wang, Sai Ma, Junlan Chen, Fan Ren

摘要:

视觉里程计(VO)是一种较为流行的相对定位方法,其定位精度和精度都在不断提高,但在复杂环境下,其定位精度和精度往往会受到影响。与经典的基于几何的方法相比,基于深度学习的方法可以自动学习有效和鲁棒的表示,例如深度,光流,特征,自我运动等,不需要显式计算。然而,仍然缺乏对基于深度学习的VO(Deep VO)的最新进展的全面回顾。因此,本文旨在深入了解深度学习如何能够受益和优化VO系统。我们首先筛选出一些条件,包括准确性,效率,可扩展性,动态性,实用性和可扩展性,并采用它们作为标准。然后,以所提出的标准为统一度量,从深度估计、特征提取与匹配、姿态估计等方面详细评估和讨论了深度学习如何提高虚拟现实的性能。总结了Deep VO研究的复杂领域和新兴领域,如移动机器人、医疗机器人、增强现实和虚拟现实等。通过文献分解、分析和比较,提出了Deep VO研究中的若干开放性问题,并提出了该领域未来的研究方向。

英文摘要:

Visual odometry (VO) is a prevalent way to deal with the relative localization problem, which is becoming increasingly mature and accurate, but it tends to be fragile under challenging environments. Comparing with classical geometry-based methods, deep learning-based methods can automatically learn effective and robust representations, such as depth, optical flow, feature, ego-motion, etc., from data without explicit computation. Nevertheless, there still lacks a thorough review of the recent advances of deep learning-based VO (Deep VO). Therefore, this paper aims to gain a deep insight on how deep learning can profit and optimize the VO systems. We first screen out a number of qualifications including accuracy, efficiency, scalability, dynamicity, practicability, and extensibility, and employ them as the criteria. Then, using the offered criteria as the uniform measurements, we detailedly evaluate and discuss how deep learning improves the performance of VO from the aspects of depth estimation, feature extraction and matching, pose estimation. We also summarize the complicated and emerging areas of Deep VO, such as mobile robots, medical robots, augmented reality and virtual reality, etc. Through the literature decomposition, analysis, and comparison, we finally put forward a number of open issues and raise some future research directions in this field.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑