Monocular camera/IMU/GNSS integration for ground vehicle navigation in challenging GNSS environments

Area of Science:

• Robotics and Autonomous Systems
• Geomatics Engineering
• Sensor Fusion

Background:

• Commercially available Inertial Measurement Units (IMUs), video cameras, and Global Navigation Satellite Systems (GNSS) are increasingly integrated into automotive systems.
• GNSS offers global positioning but suffers from signal attenuation, reflections, and blockages, leading to navigation difficulties.
• IMUs provide high-bandwidth navigation solutions independent of external signals but accumulate drift errors over time; cameras offer potential in challenging GNSS environments.

Purpose of the Study:

• To develop an integrated camera/IMU/GNSS system for ground vehicle navigation in challenging environments.
• To leverage existing onboard automotive technologies for improved navigation accuracy and reliability.

Main Methods:

• Development of an integrated system combining camera, IMU, and GNSS sensors.
• Implementation of an extended Kalman filter (EKF) for sensor fusion.
• Validation using a live dataset collected in an operational traffic environment.

Main Results:

• The proposed integrated system demonstrated accurate navigation estimations.
• Experimental results indicate potential performance advantages over tightly coupled GNSS/IMU integration, especially in environments with sparse GNSS observations.
• The system effectively utilizes available onboard sensors for enhanced navigation.

Conclusions:

• The integrated camera/IMU/GNSS system offers a viable solution for robust ground vehicle navigation.
• The EKF-based fusion approach effectively mitigates individual sensor limitations.
• This integrated system shows promise for outperforming traditional methods in GNSS-denied or degraded environments.