Active SLAM: A Review on Last Decade

Ahmed, Muhammad Farhan; Masood, Khayyam; Fremont, Vincent; Fantoni, Isabelle

doi:10.3390/s23198097

Cited by 15 publications

(7 citation statements)

References 122 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Mobile laser scanning using GeoSLAM instruments is recognized as an efficient method for close-range 3D data acquisition in forest environments, providing substantial time saving compared to TLS and easy-to-use operation compared to UAV-mounted sensors. However, data obtained through SLAM are known to exhibit noticeable noise in surface reconstruction [29,32], which must be addressed during data processing. Our study highlights that the noise is not symmetrically distributed along the reconstructed tree surface, contrary to the commonly assumed symmetric distribution in tree diameter estimation methods.…”

Section: Discussionmentioning

confidence: 99%

“…Another significant drawback of mobile scanning is the fact that point clouds obtained using SLAM-based scanners exhibit markedly higher levels of unavoidable stochastic noise [17,21,[29][30][31] caused by sensor position uncertainty [32], compared to TLS. However, SLAM technology can still provide accurate positioning of the scanner [23] which can outcompete GNSS (Global Navigation Satellite System) and IMU (Inertial Measurement Unit) positioning for forest mapping [33].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Noise Analysis for Unbiased Tree Diameter Estimation from Personal Laser Scanning Data

Kuželka,

Surový

2024

Remote Sensing

View full text Add to dashboard Cite

Personal laser scanning devices employing Simultaneous Localization and Mapping (SLAM) technology have rightfully gained traction in various applications, including forest mensuration and inventories. This study focuses the inherent stochastic noise in SLAM data. An analysis of noise distribution is performed in GeoSLAM ZEB Horizon for point clouds of trees of two species, Norway spruce and European beech, to mitigate bias in diameter estimates. The method involved evaluating residuals of individual 3D points concerning the real tree surface model based on TLS data. The results show that the noise is not symmetrical regarding the real surface, showing significant negative difference, and moreover, the difference from zero mean significantly differs between species, with an average of −0.40 cm for spruce and −0.44 cm for beech. Furthermore, the residuals show significant dependence on the return distance between the scanner and the target and the incidence angle. An experimental comparison of RANSAC circle fitting outcomes under various configurations showed unbiased diameter estimates with extending the inlier tolerance to 5 cm with 2.5 cm asymmetry. By showing the nonvalidity of the assumption of zero mean in diameter estimation methods, the results contribute to fill a gap in the methodology of data processing with the widely utilized instrument.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Noise Analysis for Unbiased Tree Diameter Estimation from Personal Laser Scanning Data

Kuželka,

Surový

2024

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Active SLAM aims to address the challenge of optimal exploration in unknown environments by proposing a path planning strategy that generates actions to reduce map and pose uncertainty [13]. Alongside the localization and mapping, the path planning module plays a vital role in the Active SLAM framework.…”

Section: Of 19mentioning

confidence: 99%

“…t − (t − 1)13.Set robot motion direction angle at p t−1 to 0• 14. ∆ϑ ← ω t−1 * ∆t 15. if ( 0 • + ∆ϑ ∈ N, perform16.…”

mentioning

confidence: 99%

Exploration- and Exploitation-Driven Deep Deterministic Policy Gradient for Active SLAM in Unknown Indoor Environments

Zhao,

Hwang

2024

Electronics

View full text Add to dashboard Cite

This study proposes a solution for Active Simultaneous Localization and Mapping (Active SLAM) of robots in unknown indoor environments using a combination of Deep Deterministic Policy Gradient (DDPG) path planning and the Cartographer algorithm. To enhance the convergence speed of the DDPG network and minimize collisions with obstacles, we devised a unique reward function that integrates exploration and exploitation strategies. The exploration strategy allows the robot to achieve the shortest running time and movement trajectory, enabling efficient traversal of unmapped environments. Moreover, the exploitation strategy introduces active closed loops to enhance map accuracy. We conducted experiments using the simulation platform Gazebo to validate our proposed model. The experimental results demonstrate that our model surpasses other Active SLAM methods in exploring and mapping unknown environments, achieving significant grid completeness of 98.7%.

show abstract

“…In the research field of intelligent robotics and autonomous navigation systems, Simultaneous Localization and Mapping (SLAM) technology plays a crucial role [ 1 , 2 ]. Notably, the work of M.W.…”

Section: Introductionmentioning

confidence: 99%

ADM-SLAM: Accurate and Fast Dynamic Visual SLAM with Adaptive Feature Point Extraction, Deeplabv3pro, and Multi-View Geometry

Huang,

Chen,

Zhang

et al. 2024

Sensors

View full text Add to dashboard Cite

Visual Simultaneous Localization and Mapping (V-SLAM) plays a crucial role in the development of intelligent robotics and autonomous navigation systems. However, it still faces significant challenges in handling highly dynamic environments. The prevalent method currently used for dynamic object recognition in the environment is deep learning. However, models such as Yolov5 and Mask R-CNN require significant computational resources, which limits their potential in real-time applications due to hardware and time constraints. To overcome this limitation, this paper proposes ADM-SLAM, a visual SLAM system designed for dynamic environments that builds upon the ORB-SLAM2. This system integrates efficient adaptive feature point homogenization extraction, lightweight deep learning semantic segmentation based on an improved DeepLabv3, and multi-view geometric segmentation. It optimizes keyframe extraction, segments potential dynamic objects using contextual information with the semantic segmentation network, and detects the motion states of dynamic objects using multi-view geometric methods, thereby eliminating dynamic interference points. The results indicate that ADM-SLAM outperforms ORB-SLAM2 in dynamic environments, especially in high-dynamic scenes, where it achieves up to a 97% reduction in Absolute Trajectory Error (ATE). In various highly dynamic test sequences, ADM-SLAM outperforms DS-SLAM and DynaSLAM in terms of real-time performance and accuracy, proving its excellent adaptability.

show abstract

Active SLAM: A Review on Last Decade

Cited by 15 publications

References 122 publications

Noise Analysis for Unbiased Tree Diameter Estimation from Personal Laser Scanning Data

Noise Analysis for Unbiased Tree Diameter Estimation from Personal Laser Scanning Data

Exploration- and Exploitation-Driven Deep Deterministic Policy Gradient for Active SLAM in Unknown Indoor Environments

ADM-SLAM: Accurate and Fast Dynamic Visual SLAM with Adaptive Feature Point Extraction, Deeplabv3pro, and Multi-View Geometry

Contact Info

Product

Resources

About