基于机器视觉的交通检测技术及应用研究
A Study on Traffic Detection Technology and Application Based on Machine Vision

传统交通检测方法存在检测效率低、计算成本高、准确性差等不足，为此，本文提出实时RT-DETR目标检测与BoTSORT目标跟踪相结合的机器学习算法，并通过实验证明本文所采用的技术在车辆检测精度上表现优异，能够有效识别并跟踪公路上行驶的车辆目标。进一步地，本文对交通冲突与车流量之间的关系进行分析，发现交通冲突频率与车流量之间存在显著的正相关关系。研究结果表明本文方法有助于提高交通检测效率和精度，可为交通主动管控提供支撑。
Traditional traffic detection has shortcomings such as low detection efficiency, high computational cost, and poor accuracy. Therefore, this paper proposed a machine learning algorithm combining RT-DETR object detection and BoTSORT object tracking, and conducted experimental studies to prove that the employed technology performs excellently in detection and tracking of vehicles driving on highways. Furthermore, this paper analyzed the relationship between traffic conflicts and traffic volume, and found a significant positive correlation between the frequency of traffic conflicts and the increase in traffic volume. The results indicate that the method proposed in this paper helps to improve the efficiency and accuracy of traffic detection and tracking, which can provide support for active traffic control.