A Monocular Camera as an Operation Logger for Motorized Mobility Scooters: Visual Odometry Method to Estimate...

Area of Science:

• Assistive Technology
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Motorized mobility scooters (MMSs) are crucial for older adults' independence.
• Operational errors in MMS use pose accident risks due to lack of licensing and complex existing sensing systems.
• Inertial Measurement Units (IMUs) for MMS operation monitoring are complex for clinical use.

Purpose of the Study:

• To develop a simplified sensing system for recording MMS operations.
• To evaluate the feasibility of using a single camera and image processing for estimating operational parameters.
• To enhance the usability of systems for assessing MMS driving skills.

Main Methods:

• A novel system using a single-lens camera and image processing was developed.
• Optical flow and template matching algorithms were employed to estimate steering and throttle angles.
• The system processed road surface images captured by a monocular camera.

Main Results:

• The camera-based system achieved comparable accuracy to IMU-based systems for steering angle estimation.
• Throttle angle estimation showed high accuracy during stable driving but was affected by MMS inertia during acceleration/deceleration.
• The simplified sensor setup significantly reduced system complexity.

Conclusions:

• The proposed camera-based system offers a fundamental computational technique for measuring MMS operations.
• This approach enhances the usability of recording systems for evaluating MMS driving skills in older adults.
• The system's simplicity facilitates wider clinical application and improved safety assessments.