An Overview on Visual SLAM: From Tradition to Semantic

Abstract

Visual SLAM (VSLAM) has been developing rapidly due to its advantages of low-cost sensors, the easy fusion of other sensors, and richer environmental information. Traditional visionbased SLAM research has made many achievements, but it may fail to achieve wished results in challenging environments. Deep learning has promoted the development of computer vision, and the combination of deep learning and SLAM has attracted more and more attention. Semantic information, as high-level environmental information, can enable robots to better understand the surrounding environment. This paper introduces the development of VSLAM technology from two aspects: traditional VSLAM and semantic VSLAM combined with deep learning. For traditional VSLAM, we summarize the advantages and disadvantages of indirect and direct methods in detail and give some classical VSLAM open-source algorithms. In addition, we focus on the development of semantic VSLAM based on deep learning. Starting with typical neural networks CNN and RNN, we summarize the improvement of neural networks for the VSLAM system in detail. Later, we focus on the help of target detection and semantic segmentation for VSLAM semantic information introduction. We believe that the development of the future intelligent era cannot be without the help of semantic technology. Introducing deep learning into the VSLAM system to provide semantic information can help robots better perceive the surrounding environment and provide people with higher-level help.

Visual SLAM: What Are the Current Trends and What to Expect?

“…However, they did not include other vision sensors, such as event camera-based ones, which will be discussed later in Section 4.1 . Chen et al [ 18 ] reviewed a wide range of traditional and semantic VSLAM publications. They divided the SLAM development era into classical , algorithmic-analysis , and robust-perception stages and introduced hot issues there.…”

“…They collected approaches that have been evaluated on the KITTI dataset, enabling them to have a brief description of the advantages and demerits of each system. Cheng et al [ 18 ] reviewed the VSLAM-based autonomous driving systems and raised the future development trends of such systems in a similar manuscript. Some other researchers surveyed VSLAM works with the ability to work in real-world conditions.…”

Integrating Sparse Learning-Based Feature Detectors into Simultaneous Localization and Mapping—A Benchmark Study

“…From a methodological point of view, localization is achieved by relying on Visual Odometry (VO) or Simultaneous Localization And Mapping (SLAM), as presented in numerous literature works e.g., Yousif et al [ 4 ], Agostinho et al [ 5 ] and Chen et al [ 6 ]. Most of these approaches leverage image features (i.e., keypoints) tracking across multiple frames to estimate the camera ego-motion.…”

YPD-SLAM: A Real-Time VSLAM System for Handling Dynamic Indoor Environments

“…In these scenarios, RGB-D cameras or LiDAR are often used as the primary sensors to capture the scene [ 1 , 2 ]. The Visual SLAM framework is now relatively mature and consists mainly of front-end feature extraction, back-end state estimation, loopback detection, and map building [ 3 ]. Some excellent SLAM algorithms, such as ORB-SLAM2 [ 4 ], HECTOR-SLAM [ 5 ], LSD-SLAM [ 6 ], etc., have been applied in some fields with more excellent results.…”