Exécution de notre outil de synthèse sur un nouvel article
Vérifiez l’article ci-dessous, puis lancez sa synthèse.
Mmwave SAR-Based Road Debris Detection with Adaptive Fusion of Stereo Camera Data
Hyokbeen Lee (Seoul National University), Duk‐jin Kim (Seoul National University)
Résumé : Extensive research utilizing camera, radar, and LiDAR has been conducted on the road debris detection. In particular, the Synthetic Aperture Radar (SAR) system using millimeterwave (mmWave) radar offers the advantages of high resolution and reliable operation under adverse weather conditions. However, mmWave SAR images are affected by background clutter, making road debris detection challenging. In this study, we propose a novel framework for road debris detection using mmWave SAR system with a stereo camera. We performed reducing background clutter of the SAR image. And we calculated a Mean Square Error (MSE) map for each sensor using window images. To achieve optimal sensor fusion, we determined the weight of each sensor by minimizing the MSE with respect to the ground truth image. This framework results in a 0.023 reduction in the average MSE for total debris detection. Furthermore, we derived an equation that configures optimal weight using the entropy of the camera image. This process enables an adaptive and optimized sensor fusion based on environmental conditions, allowing for operation across various scenarios.