Citation: Wang, F.; Zhang, C.; Zhang,
W.; Fang, C.; Xia, Y.; Liu, Y.; Dong, H.
Object-Based Reliable Visual
Navigation for Mobile Robot. Sensors
2022, 22, 2387. https://doi.org/
10.3390/s22062387
Academic Editors: Arturo de la
Escalera Hueso and Andrey V. Savkin
Received: 27 January 2022
Accepted: 16 March 2022
Published: 20 March 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Object-Based Reliable Visual Navigation for Mobile Robot
Fan Wang
1,2
, Chaofan Zhang
1,
* , Wen Zhang
1,
*, Cuiyun Fang
1,2
, Yingwei Xia
1
, Yong Liu
1
and Hao Dong
3
1
Anhui Institute of Optics and Fine Mechanics, Hefei Institutes of Physical Science, Chinese Academy of
Sciences, Hefei 230031, China; wanfan8@mail.ustc.edu.cn (F.W.); fangcy@mail.ustc.edu.cn (C.F.);
xiayw@aiofm.ac.cn (Y.X.); liuyong@aiofm.ac.cn (Y.L.)
2
Science Island Branch of Graduate School, University of Science and Technology of China, Hefei 230026, China
3
China National Tobacco Quality Supervision Test Center, Zhengzhou 450001, China; dongh@ztri.com.cn
* Correspondence: zcfan@aiofm.ac.cn (C.Z.); zhangwen@aiofm.ac.cn (W.Z.); Tel.: +86-187-5519-1725 (C.Z.);
+86-181-5607-2858 (W.Z.)
Abstract:
Visual navigation is of vital importance for autonomous mobile robots. Most existing
practical perception-aware based visual navigation methods generally require prior-constructed
precise metric maps, and learning-based methods rely on large training to improve their generality. To
improve the reliability of visual navigation, in this paper, we propose a novel object-level topological
visual navigation method. Firstly, a lightweight object-level topological semantic map is constructed
to release the dependence on the precise metric map, where the semantic associations between objects
are stored via graph memory and topological organization is performed. Then, we propose an object-
based heuristic graph search method to select the global topological path with the optimal and shortest
characteristics. Furthermore, to reduce the global cumulative error, a global path segmentation
strategy is proposed to divide the global topological path on the basis of active visual perception and
object guidance. Finally, to achieve adaptive smooth trajectory generation, a Bernstein polynomial-
based smooth trajectory refinement method is proposed by transforming trajectory generation into a
nonlinear planning problem, achieving smooth multi-segment continuous navigation. Experimental
results demonstrate the feasibility and efficiency of our method on both simulation and real-world
scenarios. The proposed method also obtains better navigation success rate (SR) and success weighted
by inverse path length (SPL) than the state-of-the-art methods.
Keywords:
topological path planning; visual navigation; object-level topological semantic map;
Bernstein polynomial
1. Introduction
Over the last few decades, autonomous mobile robots have gained increasing attention
for various applications, such as indoor service, surveillance missions, and search-and-
rescue. Safe and reliable autonomous navigation is of crucial importance for mobile robots
to execute their main tasks in complex environments. Vision-based navigation has become
a popular research area due to the richness and practicality of vision sensors [
1
]. Most
existing practical indoor vision navigation methods focused on path planning with a prior
precise metric map, such as an occupancy grid map [
2
] and dense map [
3
]. Generally, these
maps are constructed with Simultaneous Localization And Mapping (SLAM) algorithms,
which can perform well in a conditional ideal environment [
4
]. In spite of their remarkable
results, some challenging environments, such as unstructured indoor areas and dynamic
objects, pose great challenges for the performance of visual navigation methods. With the
development of deep learning, the learning-based visual navigation methods demonstrate
strong navigation performance to the above problems [
5
–
7
], while they need a large number
of training datasets to improve generalization capabilities. Therefore, it is necessary to
exploit a reliable and feasible visual navigation method.
Sensors 2022, 22, 2387. https://doi.org/10.3390/s22062387 https://www.mdpi.com/journal/sensors
