Distance Estimation by Fusion of Monocular Ranging and MMW Radar Information

A single sensor often fails to provide comprehensive and reliable information in complex environments. However, multi-sensor fusion leverages the complementary strengths of different sensors, fully utilizing their measurement capabilities and effectively mitigating the limitations of individual sensors. In practical applications, sensor data is frequently influenced by environmental noise, resulting in increased measurement errors. To overcome this challenge, this paper presents a monocular distance measurement model based on object detection algorithms, which is combined with millimeter-wave radar for sensor fusion, aiming to enhance measurement accuracy and system robustness. The proposed method first constructs a geometric model for monocular distance measurement based on object detection algorithms. Subsequently, a dynamic error model is introduced to address the issue of sensor errors varying with distance. Finally, Kalman filtering is applied to fuse data from multiple sensors. Simulation results demonstrate that the proposed approach effectively reduces sensor measurement noise, significantly improves distance measurement accuracy, and meets the requirements of practical applications.