Table 2 Summary of recent robust SLAM systems
From: A survey: which features are required for dynamic visual simultaneous localization and mapping?
References | System properties | Implementation details | Practical consideration | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
Backbone | CT | Env | MS | HE | P/S | BI | OH | LC | HL | |
Low-level based SLAM (Robust SLAMÂ section) | ||||||||||
 Point-based or pixel-patch-based SLAM | ||||||||||
  Yang et al. [20] | ORB-SLAM2 [21] | D | I | RE | – | – | – | – | – | – |
  Du et al. [22] | ORB-SLAM2 | D | I | E + RE | – | √ | – | – | √ | – |
  Zhang et al. [23] | – | D | I | OF + DI | – | √ | √ | – | – | – |
  Tan et al. [24] | PTAM [6] | M | I | RE | – | – | – | √ | – | – |
 Point-line-based SLAM | ||||||||||
  Zhang et al. [25] | – | D | I | 3DE | – | √ | – | – | √ | √ |
Using high-level feature as semantic priors in low-level feature-based SLAM (Using high-level features as semantic priors for low-level-feature-based SLAMÂ section) | ||||||||||
 Point-based SLAM | ||||||||||
  Bescos et al. [26] | ORB-SLAM2 | M, S, D | I, O | SI + DI | S [27] | – | √ | √ | √ | – |
  Yu et al. [28] | ORB-SLAM2 | D | I | SI + E | S [29] | – | – | – | – | – |
  Cui and Ma [30] | ORB-SLAM2 | D | I | SI + E | S [29] | – | – | – | – | – |
  Han and Xi [31] | ORB-SLAM2 | D | I | SI + OF | S [32] | – | – | – | – | – |
  Long et al. [33] | ORB-SLAM2 | D | I, O | SI + DI | S [32] | – | √ | – | – | – |
  Ai et al. [34] | ORB-SLAM2 | S, D | I, O | SI | O [35] | √ | – | – | √ | – |
  Xiao et al. [36] | ORB-SLAM2 | M | I, O | SI + RE | O [37] | √ | – | – | √ | – |
  Brasch et al. [38] | ORB-SLAM [39] | M | O | SI + T | S [40] | √ | – | – | √ | – |
 Point-line-based SLAM | ||||||||||
  Zhang et al. [41] | – | D | I | SI + DI + E* | O [42] | – | – | – | – | √ |
Using high-level features in object SLAM (Using high-level features in object SLAMÂ section) | ||||||||||
 Yang and Scherer [14] | – | M | I, O | E | O [43] | – | – | – | – | √ |