Autonomous Exploration of Unknown Indoor Environments for High-Quality Mapping Using Feature-Based RGB-D SLAM.

Area of Science:

• Robotics
• Computer Vision
• Artificial Intelligence

Background:

• Simultaneous Localization and Mapping (SLAM) is vital for autonomous mobile robots in unknown indoor environments.
• Existing SLAM strategies prioritize full coverage and minimal exploration time, often neglecting mapping quality.
• Mapping quality is critical for 3D modeling, particularly with low-cost sensors in complex indoor settings.

Purpose of the Study:

• To propose a novel exploration algorithm for autonomous mobile robots that optimizes both exploration time and mapping quality.
• To integrate mapping quality factors into the exploration strategy for feature-based RGB-D SLAM.
• To enhance the efficiency and accuracy of indoor mapping using low-cost sensors.

Main Methods:

• Utilized feature-based RGB-D SLAM for its low computational cost and real-time dense reconstruction capabilities.
• Developed novel exploration strategies that incorporate mapping quality metrics specific to the RGB-D SLAM system.
• Employed a Voronoi path planner for efficient navigation, maximizing obstacle clearance and ensuring fixed paths.
• Evaluated three distinct exploration strategies across three real-world environments, considering varying texture levels.

Main Results:

• Achieved significant improvements in mapping quality compared to baseline frontier-based exploration.
• Demonstrated a notable reduction in exploration time with the proposed strategies.
• Observed particularly strong performance enhancements in low-texture environments.

Conclusions:

• The proposed exploration algorithm effectively balances exploration time and mapping quality in indoor SLAM.
• The integration of mapping quality factors enhances the performance of RGB-D SLAM systems.
• This approach offers a promising solution for high-quality, efficient indoor mapping with autonomous mobile robots, especially in challenging conditions.