ABSTRACT
The active collision avoidance and pre-warning system prove highly effective in minimizing ship-bridge collisions. Existing research mainly leans on single-source data for assessing the risk of such collisions, yet single-source data inherently harbor limitations. In this study, we propose an active collision avoidance and pre-warning decision method based on vision and radar data fusion. Two risk identification models are designed to fit different environmental conditions. The method is tested on Youdungang bridge and shows fair risk identification ability. The future holds promise for delving deeper into the potential of multi-source data fusion within this domain.
