A Robust Strategy for Roadside Cooperative Perception Based on Multi-Sensor Fusion

A Practical Large-Scale Roadside Multi-View Multi-Sensor Spatial Synchronization Framework for Intelligent Transportation Systems

“…Existing spatial synchronization methods primarily rely on sequential one-to-one calibration, requiring extensive manual intervention [3], [4], [25]. For instance, In the method proposed in [5], a set of sensors, including one LiDAR and two cameras, undergoes internal calibration followed by an external calibration with another set of sensors, leading to significant accumulated errors. Similarly, approaches in [3], [6] employ Zhang's method [7] for camera calibration.…”

“…In order to make a clear comparison with the previous methods, we manually performed camera calibration and spatial synchronization, following the procedures described in [5]. Experimental comparisons were conducted as shown in Table VI, and it is evident that our approach exhibits significant advantages in terms of deployment efficiency and accuracy.…”

“…Currently, there is a substantial research gap in the synchronization of multiple views and sensors in large-scale roadside deployments. Most existing studies focuses on experimental setups, as exemplified by Zheng et al's work [5], which primarily explores controlled scenarios [6], or single-view scenarios as discussed by Du et al [4]. Roadside spatial synchronization typically involves the CST, as shown in Fig.…”