With the rapid development of the logistics industry,vehicle target tracking in warehousing and distribution environments faces problems such as complex backgrounds,occlusion,and lighting changes.To solve these problems,this paper proposes a vehicle target tracking method based on optimized twin networks.This method enhances the ability to extract target features by introducing attention mechanisms on the basis of twin networks,using improved feature extraction techniques to adapt to complex storage environments,and combining Kalman filters to improve tracking stability and smoothness.The experimental results show that this method outperforms the compared algorithms in various performance indicators.In terms of accuracy,recall,and F1 score,they reached high levels of 94.2%,93.8%,and 94.0%,respectively.Especially in complex fields and extreme conditions such as narrow tunnels and elevated warehouses,this method demonstrates excellent tracking performance and robustness.
关键词
孪生网络/仓储配送/可视化监控/车辆目标跟踪
Key words
twin network/warehousing and distribution/visual monitoring/vehicle target tracking
维普
万方数据