一种动态场景下语义分割优化的ORB_SLAM2

王召东; 郭晨

doi:10.16411/j.cnki.issn1006-7736.2018.04.018

大连海事大学学报 >

2018 , Vol. 44 >Issue 4: 121 - 126

DOI: https://doi.org/10.16411/j.cnki.issn1006-7736.2018.04.018

控制

一种动态场景下语义分割优化的ORB_SLAM2

王召东 ,
郭晨

展开

(大连海事大学船舶电气工程学院，辽宁大连 116026)

王召东（1995—），男，硕士生，E-mail：342792314@qq.com.

收稿日期: 2018-06-19

修回日期: 2018-07-21

网络出版日期: 2018-07-21

基金资助

国家自然科学基金资助项目（51579024; 61374114）；中央高校基本科研业务费专项资金资助项目（3132016311）.

收起

An improved ORB_SLAM2 in dynamic scene with semantic segmentation

WANG Zhao-dong ,
GUO Chen

Expand

（Marine Electrical Engineering College,Dalian Maritime University, Dalian 116026, China）

Received date: 2018-06-19

Revised date: 2018-07-21

Online published: 2018-07-21

Fold

摘要

为提高ORB_SLAM2在动态环境下位姿估计的准确性，提出使用语义分割剔除分布在人上的移动特征点进而提高位姿准确性的方法.该方法在对输入图像提取ORB特征点的同时，对图像进行语义分割，获得人在图像中的像素点位置，再将这些分布在人上面的特征点进行剔除，并使用剔除后相对稳定的特征点进行位姿估计.利用改进方法在TUM公共数据集上进行测试，结果表明，改进后的系统能够明显降低动态环境下位姿估计的绝对误差和相对漂移，证明了该方法较原始的ORB_SLAM2系统能够明显提高动态环境下位姿估计的准确性.

关键词： 语义分割; 动态场景; 位姿估计; ORB_SLAM2

本文引用格式

王召东 , 郭晨 . 一种动态场景下语义分割优化的ORB_SLAM2[J]. 大连海事大学学报, 2018 , 44(4) : 121 -126 . DOI: 10.16411/j.cnki.issn1006-7736.2018.04.018

Abstract

In order to improve the accuracy of ORB_SLAM2 in dynamic scene, this paper proposed a method of using semantic segmentation to eliminate the mobile feature points distributed on the human body to improve the accuracy of pose. This method extracted ORB feature points from the input image, semantically segmented the image to obtain the position of the pixels in the image, and then eliminated these feature points distributed on the top of the human body, and estimated the pose using the relatively stable feature points after eliminating. The improved method was tested on the TUM data set, and results show that it can reduce absolute error and relative drift under dynamic environment, which proves that the method is more accurate in pose estimation than the traditional in the dynamic scene.

Key words： semantic segmentation; dynamic scene; pose estimation; ORB_SLAM2

参考文献

[1]余杰. 基于ORB关键帧闭环检测算法的SLAM方法研究[D]. 浙江工业大学, 2016.
Yu Jie. Realization of SLAM Based on Improved ORB Key Frame Detection and Matching[D]. Zhejiang University of Technology, 2016.
[2]Cheng M M, Zheng S, Lin W Y, et al. ImageSpirit:Verbal Guided Image Parsing[J]. Acm Transactions on Graphics, 2014, 34(1):1-11.
[3]Jiang F, Gu Q, Hao HZ, Li N, Guo YW, Chen DX. Survey on content-based image segmentation methods. Ruan Jian Xue Bao/Journal of Software, 2017,28(1):160-183 (in Chinese). http://www.jos.org.cn/1000-9825/5136.htm
姜枫,顾庆,郝慧珍,李娜,郭延文,陈道蓄.基于内容的图像分割方法综述.软件学报,2017,28(1):160?183.
http:// www.jos.org.cn/1000-9825/5136.htm
[4]Salehi A, Gay-Bellile V, Bourgeois S, et al. Improving Constrained Bundle Adjustment Through Semantic Scene Labeling[J]. 2016.
[5]An L, Zhang X, Gao H, et al. Semantic segmentation–aided visual odometry for urban autonomous driving[J]. 2017, 14(5):172988141773566.
[6]Papon J, Schoeler M. Semantic Pose Using Deep Networks Trained on Synthetic RGB-D[C]// IEEE International Conference on Computer Vision. IEEE, 2016:774-782.
[7]高翔,张涛.视觉SLAM十四讲[M].北京:电子工业出版社,2017:26-20.
Xiang Gao, Tao Zhang. 视觉SLAM十四讲[M].Beijing: Publishing House of Electronics Industry, 2017:26-20.
[8]Mur-Artal R, Tardós J D. ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras[J]. IEEE Transactions on Robotics, 2017, 33(5):1255-1262
[9]Badrinarayanan V, Kendall A, Cipolla R. SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2015, PP(99):1-1.
[10]K. Simonyan and A. Zisserman, “Very deep convolutional networks forlarge-scale image recognition,” arXiv preprint arXiv:1409.1556, 2014.

Options

文章导航

模态框（Modal）标题

摘要

本文引用格式

Abstract

参考文献