An Indoor Wayfinding System Based on Geometric Features Aided Graph SLAM for the Visually Impaired

Area of Science:

• Robotics
• Computer Vision
• Human-Computer Interaction

Background:

• Visually impaired individuals face significant challenges with indoor navigation.
• Existing indoor wayfinding systems often lack the required accuracy and efficiency.
• Accurate pose estimation is crucial for effective spatial awareness and navigation.

Purpose of the Study:

• To develop a robust 6-degree of freedom (DOF) pose estimation (PE) method.
• To create an indoor wayfinding system for the visually impaired based on the PE method.
• To reduce accumulated pose error and enhance navigation accuracy in indoor environments.

Main Methods:

• Implemented a two-graph simultaneous localization and mapping (SLAM) approach.
• Integrated floor plane extraction for 6-DOF SLAM to minimize roll, pitch, and Z errors.
• Incorporated wall line extraction for 3-DOF SLAM to reduce X, Y, and yaw errors.

Main Results:

• The proposed PE method significantly reduces 6-DOF pose error compared to state-of-the-art planar SLAM methods.
• The system achieves higher pose accuracy with reduced computational time.
• Experimental results confirm the effectiveness of the PE method and the wayfinding system.

Conclusions:

• The developed PE method offers a more accurate and computationally efficient solution for indoor localization.
• The wayfinding system effectively assists visually impaired users in indoor navigation tasks.
• This technology has the potential to substantially improve independence and accessibility for the visually impaired.