Enhanced Real-Time State Estimation Using Multi-Sensor Fusion Based on Kalman Filtering

The study introduces a Kalman filter-based multi-sensor fusion method designed to integrate LiDAR, camera, and ultrasonic sensor data for improved state estimation accuracy and reliability. By explicitly modeling sensor-specific noise characteristics within a unified probabilistic framework, the proposed method effectively balances prediction and measurement uncertainties. Experimental results validate the method's superior performance, demonstrating significant reductions in measurement errors and enhancements in real-time stability compared to single-sensor approaches. This research offers practical implications for enhancing perception systems in autonomous driving and robotic navigation, with recommendations for extending the approach to higher-dimensional state estimation in future work.