检索结果-南通市图书馆

Bags of tricks for learning depth and camera motion from monocular videos

Virtual Reality & Intelligent Hardware 2019年第5期1卷 500-510页

作者： Bowen DONG Lu SHENG School of Computer Science and Technology Harbin Institute of TechnologyHarbin 150001China College of Software Beihang UniversityBeijing 100191China

Background Based on the seminal work proposed by Zhou et al., much of the recent progress in learning monocular visual odometry, i.e., depth and camera motion from monocular videos, can be attributed to the tricks in the training procedure, such as data augmentation and learning objectives. Methods Herein, we categorize a collection of such tricks through the theoretical examination and empirical evaluation of their effects on the final accuracy of the visual odometry. Results/Conclusions By combining the aforementioned tricks, we were able to significantly improve a baseline model adapted from SfMLearner without additional inference costs. Furthermore, we analyzed the principles of these tricks and the reason for their success. Practical guidelines for future research are also presented.

关键词： Unsupervised learning monocular visual odometry

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

An odometry Estimation Method for Micro Indoor Flying Robot with Lightweight Sensors

An Odometry Estimation Method for Micro Indoor Flying Robot ...

引用

第25届中国控制与决策会议

作者： Zheng Fang Lei Zhang State Key Laboratory of Synthetical Automation for Process Industries Northeastern University

ISBN: (纸本)9781467355339

Micro-quadrotor usually equips with lightweight sensors due to its limited payload. In this paper, an odometry estimation method based on 2D laser scanner and monocular vision is proposed to solve the pose estimation problem of micro-quadrotor flying robot in indoor environments. This method firstly detects and matches image features of current and previous frames to get all feature correspondences to calculate the monocular odometry. Then, laser scans are transformed into image coordinate system using sensor calibration parameters. After that, the unknown translation scale of monocular odometry is determined by using triangulation. At the same time, line features of laser scans are extracted and the line angle histogram is computed. If the angle distribution is wide enough, which means it is likely to be in a structural environment, a scan matching algorithm is used to calculate the laser odometry. Finally, estimation results based on the two different sensors are fused by using weighted average method. Experiment results show that the proposed method can recover the trajectory of the robot robustly and accurately. The close loop error is less than 1% for a 64 m loop around indoor corridor.

关键词： monocular visual odometry Scan Matching Pose estimation Sensor Fusion

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

Line Segment Semi-direct Tracking Based on Stereo odometry

Line Segment Semi-direct Tracking Based on Stereo Odometry

引用

第三十九届中国控制会议

作者： Liangyu Liu Yanhong Liu Lei Yang Fangfang Zhang School of Electrical Engineering Zhengzhou University

The feature based visual SLAM self locates by detecting feature points in the ***,it is difficult to extract effective feature points in the lack of texture environment,which is easy to cause tracking *** order to improve the robustness of slam in a thousand texture environment,some researchers put forward the slam method which combines feature points and feature ***,the extraction and matching of feature lines in these methods will consume a lot of computing *** paper proposes a direct tracking algorithm for line segment features,LSD algorithm is first used to extract the initial feature line segment in the key frame;in the ordinary frame,the direct tracking of line segment is realized by sampling,tracking and fitting *** algorithm is applied to the visual odometer,and experiments are carried out on the data sets of EuRoC and *** results show that the algorithm can improve the running speed of the visual odometer to 1.5 times compared with other algorithms,while maintaining similar trajectory accuracy.

关键词： visual SLAM line segment features monocular visual odometry semi-direct Tracking

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

欢迎您,

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

在线全文

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：