Publication: Simultaneous localization and mapping in dynamic environments using improved semantic dynamic cluster elimination method
Loading...
Date
2024-08
Authors
Qamar Ul Islam
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Simultaneous Localization and Mapping (SLAM) is a computational problem in robotics and computer vision where an autonomous robot or device constructs a map of an unknown environment while simultaneously keeping track of its own location within that environment. The Improved Semantic Dynamic Cluster Elimination (ISDCE) method addresses the challenge of accurate 3D pose estimation in dynamic environments by focusing on key components such as dynamic object identification and removal, static-dynamic differentiation, and geometric consistency maintenance. These components work together to solve the problems posed by dynamic obstacles, occlusions, and variable lighting conditions, thereby enhancing the
accuracy of visual SLAM systems. ISDCE integrates with existing SLAM frameworks by fine-tuning its elements to ensure effective handling of dynamic changes, allowing for real-time adjustments and reducing errors. The information flow within ISDCE enables seamless operation and interaction between its components, improving localization and mapping precision. Qualitative and quantitative results show a significant reduction in trajectory errors, with quantitative evaluations indicating a reduction in trajectory error by 35% and an increase in mapping accuracy by 40%. This research contributes to the field by incorporating semantic understanding into SLAM, expanding its applications in the autonomous robot navigation for the complex environments.