基于改进的Boosting算法的仓库监控区域目标跟踪研究

Research on Target Tracking of Warehouse Monitoring Area Based on Improved Boosting Algorithm

目前,智能视频监控系统在仓库管理中得到广泛应用,对监控系统有关的核心技术——目标检测及跟踪算法的研究则是实现智能化监控管理的重要基础性工作。研究了Boosting跟踪算法及其改进策略,结合现有的多种目标跟踪算法,通过对比分析算法的用时、跟踪运动目标数、提取前景数、漂移现象存在、是否丢失目标和需创建跟踪器个数这6个方面,选择跟踪效果及综合性能最佳的改进的Boosting算法作为仓库视频运动物体的跟踪算法,为智能化仓库管理实践提供重要参考依据。

At present,intelligent video surveillance systems are widely used in warehouse management.The research on core technology-target detection and tracking algorithms related to monitoring systems is an important basic work to realize intelligent monitoring and management.This paper mainly studies the boosting tracking algorithm and its improvement strategy.Combining the existing multiple target tracking algorithms,the experiment compares the time of the algorithm,the number of moving targets,the number of extraction foregrounds,the existence of drift phenomenon,the target lost,and the number of created trackers.The improved boosting algorithm with the best tracking performance and comprehensive performance is used as the tracking algorithm for warehouse video moving objects,which provides an important reference for intelligent warehouse management practice.