LST-SLAM: A Stereo Thermal SLAM System for Kilometer-Scale Dynamic Environments

Zeyu Jiang; Kuan Xu; Changhao Chen

cs.ROFeb 24, 2026

LST-SLAM: A Stereo Thermal SLAM System for Kilometer-Scale Dynamic Environments

Authors: Zeyu Jiang, Kuan Xu, Changhao Chen

Organizations: PEAK-Lab, The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, 511453, China · School of Electrical and Electronic Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798

Abstract

Thermal cameras offer strong potential for robot perception under challenging illumination and weather conditions. However, thermal Simultaneous Localization and Mapping (SLAM) remains difficult due to unreliable feature extraction, unstable motion tracking, and inconsistent global pose and map construction, particularly in dynamic large-scale outdoor environments. To address these challenges, we propose LST-SLAM, a novel large-scale stereo thermal SLAM system that achieves robust performance in complex, dynamic scenes. Our approach combines self-supervised thermal feature learning, stereo dual-level motion tracking, and geometric pose optimization. We also introduce a semantic-geometric hybrid constraint that suppresses potentially dynamic features lacking strong inter-frame geometric consistency. Furthermore, we develop an online incremental bag-of-words model for loop closure detection, coupled with global pose optimization to mitigate accumulated drift. Extensive experiments on kilometer-scale dynamic thermal datasets show that LST-SLAM significantly outperforms recent representative SLAM systems, including AirSLAM and DROID-SLAM, in both robustness and accuracy.

Figures & tables

Fig. 1: LST-SLAM enables robust localization and mapping in kilometer-scale dynamic thermal scenes. The system leverages self-supervised thermal features and a dual-level tracking strategy, while an incremental online BoW supports loop closure detection and global pose optimization.

Fig. 2: LST-SLAM system pipeline. Stereo inputs are processed with self-supervised feature point and dynamic filtering networks to obtain robust thermal features, which drive tracking, mapping, and local optimization. Binarized descriptors build an incremental BoW (iBoW) for loop closure and global pose optimization.

Fig. 3: Top: Architecture and details of the STP network. Bottom: Schematic diagram of adaptive self-supervised training in STP networks under thermal modality.

Fig. 4: Qualitative results of feature extraction and tracking on the M2DGR dataset. Each row visualizes feature matches over continuous video segments sampled at larger frame intervals. The three sequences were chosen to represent different motion directions.

Fig. 5: The STP network significantly outperforms other methods and substantially enhances the SuperPoint behavior in terms of the number of matches and inlier points between different frames.

Sequence	Morning				Daytime				Nighttime				Average
	Campus		Residential		Campus		Residential		Campus		Residential
Length	7037m		3090m		7071m		3151m		6970m		3620m		5157m
Metric	$t_{\mathrm{apm}}$	$CR$	$t_{\mathrm{apm}}$	$CR$	$t_{\mathrm{apm}}$	$CR$	$t_{\mathrm{apm}}$	$CR$	$t_{\mathrm{apm}}$	$CR$	$t_{\mathrm{apm}}$	$CR$	$t_{\mathrm{apm}}$	$CR$
Without Loop-Closing
SVO 2.0 [ 34 ]	–	0%	–	0%	–	0%	–	0%	–	0%	–	0%	–	0%
DytanVO [ 35 ]	.0266	100%	.0439	100%	.0443	100%	.0559	100%	.0569	100%	.0538	100%	.0469	100%

TABLE I: Performance comparison of different odometry and SLAM systems on the MS 2 dataset across various sequences. The average is calculated over all sequences. The best results are in bold .

Fig. 6: The qualitative results. Left: Daytime Residential. Right: Morning Campus. We visualize the trajectories of the top five most accurate methods and the ground truth trajectory for each scene.

Method	Campus		Residential		Average
	$t_{\mathrm{apm}}$	$CR$	$t_{\mathrm{apm}}$	$CR$	$t_{\mathrm{apm}}$	$CR$
w/o STP	–	76.8%	.0184	100%	–	88.4%
w/o Loop	.0309	100%	.0352	100%	.0331	100%
w/o Seg	.0140	100%	.0159	100%	.0149	100%
w/o DT	.0127	100%	.0144	100%	.0136	100%
LST-SLAM	.0105	100%	.0099	100%	.0102	100%

TABLE II: Performance comparison of different ablation setups. We present the average results of six sequences in MS 2 datasets.

Dynamic SLAM

See all 12 papers

PL-LIT: A LiDAR-Inertial-Thermal SLAM Using Point-Line Features and Thermographic Mapping
Jun 28, 2026Jiawei Xia, Yixiao Feng, Yongliang Shi +4Thermal Imaging Simultaneous Localization and Mapping
Tsinghua University · The Hong Kong University of Science and Technology (Guangzhou) · University of California, Berkeley
TRGS-SLAM: IMU-Aided Gaussian Splatting SLAM for Blurry, Rolling Shutter, and Noisy Thermal Images
Mar 20, 2026Spencer Carmichael, Katherine A. SkinnerSimultaneous Localization and Mapping 3DGS SLAM
University of Michigan, Ann Arbor, MI USA
RoSe-SLAM: Robust Semantic-Aware Gaussian Splatting SLAM from Dynamic Monocular Videos
Aug 29, 2026Wenting Wang, Jiaxin Guo, Wenzhen Dong +3Dynamic SLAM Visual SLAM
Department of Mechanical and Automation Engineering, The Chinese University of Hong Kong, Hong Kong SAR. · Department of Mechanical, Aerospace and Civil Engineering, The University of Manchester, Manchester, UK. · Centre for Perceptual and Interactive Intelligence (CPII) Limited, Hong Kong SAR.

Jun 28, 2026J/K move · Enter open · S save

PL-LIT: A LiDAR-Inertial-Thermal SLAM Using Point-Line Features and Thermographic Mapping

Jiawei Xia, Yixiao Feng, Yongliang Shi, Chao Gao+3

Tsinghua University · The Hong Kong University of Science and Technology (Guangzhou) · University of California, Berkeley

Thermal Imaging Simultaneous Localization and Mapping Visual SLAM Multimodal Sensor Fusion

Open paper AskPDF arXiv

Thermal imaging is resilient to adverse conditions, such as intense illumination, low-light operation, and fog, and can therefore mitigate odometry degradation when visible-spectrum imagery becomes unreliable. Nevertheless, most thermal cameras employ automatic gain control (AGC), and thermal images often present low global contrast despite containing informative edge structures. These characteristics undermine brightness constancy and cause conventional optical flow tracking-based odometry pipelines that fundamentally rely on the brightness constancy assumption across consecutive frames. To address these issues, we propose a general LiDAR-Inertial-Thermal SLAM system that accommodates both visible-light and thermal cameras. PL-LIT combines an online photometric calibration module with a deep neural network for point-line feature extraction, enabling more stable and repeatable thermal tracking. For state estimation, we design a tightly coupled LiDAR-Inertial-Thermal formulation within an Error-State Iterated Kalman Filter (ESIKF). We further introduce a line-feature constraint scheme ensuring the reliability of geometric constraints across varying thermal appearances. In addition, PL-LIT builds a probabilistic thermal-intensity voxel map, which supports real-time thermal anomaly detection. Extensive experiments demonstrate that PL-LIT exhibits generality and robustness in visible-light environments, achieves state-of-the-art performance on long-range thermal infrared datasets, and provides practical safety inspection functionality based on thermographic mapping.

Abstract

Dynamic SLAM

Explore similar work

PL-LIT: A LiDAR-Inertial-Thermal SLAM Using Point-Line Features and Thermographic Mapping

TRGS-SLAM: IMU-Aided Gaussian Splatting SLAM for Blurry, Rolling Shutter, and Noisy Thermal Images

RoSe-SLAM: Robust Semantic-Aware Gaussian Splatting SLAM from Dynamic Monocular Videos